Classification of breast cancer patients using a combination of clinical criteria and informative genesets

[0001] This application claims the benefit under 35 U.S.C. §119(e) of U.S. Provisional Patent Application No. 60/650,401, filed on Feb. 4, 2005, U.S. Provisional Patent Application No. 60/604,076, filed on Aug. 24, 2004, and U.S. Provisional Patent Application No. 60/550,810, filed on Mar. 5, 2004, each of which is incorporated by reference herein in its entirety.

[0002] The present invention relates to the use of both phenotypic and genotypic aspects of a condition, such as a disease, in order to identify discrete subsets of patients for which specific sets of informative genes are then identified. The invention also relates to the classification of individuals, such as breast cancer patients, into a subset of the condition on the basis of clinical parameters and the status of markers, for example, of genes expression patterns, and the prognosis of those individuals on the basis of markers informative for prognosis within the subset of the condition. The invention also relates to methods of determining a course of treatment or therapy to an individual having, or suspected of having, a condition, such as breast cancer. The invention further relates to methods of structuring a clinical trial, particularly using five breast cancer-specific patient subsets and prognosis-informative genes for each, and of identifying patient populations for clinical trials or for other condition-related, for example, breast cancer-related, research. Finally, the invention relates to computer implementations of the above methods.

[0003] The increased number of cancer cases reported in the United States, and, indeed, around the world, is a major concern. Currently there are only a handful of treatments available for specific types of cancer, and these provide no guarantee of success. In order to be most effective, these treatments require not only an early detection of the malignancy, but a reliable assessment of the severity of the malignancy.

[0004] The incidence of breast cancer, a leading cause of death in women, has been gradually increasing in the United States over the last thirty years. Its cumulative risk is relatively high; 1 in 8 women are expected to develop some type of breast cancer by age 85 in the United States. In fact, breast cancer is the most common cancer in women and the second most common cause of cancer death in the United States. In 1997, it was estimated that 181,000 new cases were reported in the U.S., and that 44,000 people would die of breast cancer (Parker et al., CA Cancer J. Clin. 47:5-27 (1997); Chu et al., J. Nat. Cancer Inst. 88:1571-1579 (1996)). While mechanism of tumorigenesis for most breast carcinomas is largely unknown, there are genetic factors that can predispose some women to developing breast cancer (Miki et al., Science, 266:66-71 (1994)).

[0005] Sporadic tumors, those not currently associated with a known germline mutation, constitute the majority of breast cancers. It is also likely that other, non-genetic factors also have a significant effect on the etiology of the disease. Regardless of the cancer's origin, breast cancer morbidity and mortality increases significantly if it is not detected early in its progression. Thus, considerable effort has focused on the early detection of cellular transformation and tumor formation in breast tissue.

[0006] A marker-based approach to tumor identification and characterization promises improved diagnostic and prognostic reliability. Typically, the diagnosis of breast cancer requires histopathological proof of the presence of the tumor. In addition to diagnosis, histopathological examinations also provide information about prognosis and selection of treatment regimens. Prognosis may also be established based upon clinical parameters such as tumor size, tumor grade, the age of the patient, and lymph node metastasis.

[0007] Diagnosis and/or prognosis may be determined to varying degrees of effectiveness by direct examination of the outside of the breast, or through mammography or other X-ray imaging methods (Jatoi, Am. J. Surg. 177:518-524 (1999)). The latter approach is not without considerable cost, however. Every time a mammogram is taken, the patient incurs a small risk of having a breast tumor induced by the ionizing properties of the radiation used during the test. In addition, the process is expensive and the subjective interpretations of a technician can lead to imprecision. For example, one study showed major clinical disagreements for about one-third of a set of mammograms that were interpreted individually by a surveyed group of radiologists. Moreover, many women find that undergoing a mammogram is a painful experience. Accordingly, the National Cancer Institute has not recommended mammograms for women under fifty years of age, since this group is not as likely to develop breast cancers as are older women. It is compelling to note, however, that while only about 22% of breast cancers occur in women under fifty, data suggests that breast cancer is more aggressive in pre-menopausal women.

[0008] In clinical practice, accurate diagnosis of various subtypes of breast cancer is important because treatment options, prognosis, and the likelihood of therapeutic response all vary broadly depending on the diagnosis. Accurate prognosis, or determination of distant metastasis-free survival could allow the oncologist to tailor the administration of adjuvant chemotherapy, with women having poorer prognoses being given the most aggressive treatment. Furthermore, accurate prediction of poor prognosis would greatly impact clinical trials for new breast cancer therapies, because potential study patients could then be stratified according to prognosis. Trials could then be limited to patients having poor prognosis, in turn making it easier to discern if an experimental therapy is efficacious.

[0009] To date, no set of satisfactory predictors for prognosis based on the clinical information alone has been identified. Many have observed that the ER status has a dominant signature in the breast tumor gene expression profiling. See West et al., Proc. Natl. Acad. Sci. U.S.A. 98:11462 (2001); van 't Veer et al., Nature 415:530 (2002); Sorlie et al., Proc. Natl. Acad. Sci. U.S.A. 100:8418 (2003); Perou et al. Nature 406:747 (2000); Gruvberger et al., Cancer Res. 61:5979 (2001); Sotiriou et al., Proc. Natl. Acad. Sci. U.S.A. 100:10393 (2003). It is generally accepted that there is some relationship between patient survival and ER status. van de Vijver et al., N. Engl. J. Med. 347:1999 (2002); Surowiak et al, Folia Histochem. Cytobiol. 39:143 (2001); Pichon et al. Br. J Cancer 73:1545 (1996); Collett et al., J. Clin. Pathol. 49:920 (1996). BRCA1 mutations are related to the familial cancer susceptibility. Biesecker et al., JAMA 269:1970 (1993); Easton et al., Cancer Surv. 18:95 (1993). Age is also considered to be a prognosis factor since young cancer patients tend to have poor tumors. Maggard et al., J. Surg. Res. 113:109 (2003). Lymph node status is a factor in deciding the treatment. Eifel et al., J. Natl. Cancer Inst. 93:979 (2001).

[0010] The discovery and characterization of BRCA1 and BRCA2 has recently expanded our knowledge of genetic factors which can contribute to familial breast cancer. Germ-line mutations within these two loci are associated with a 50 to 85% lifetime risk of breast and/or ovarian cancer (Casey, Curr. Opin. Oncol. 9:88-93 (1997); Marcus et al., Cancer 77:697-709 (1996)). Only about 5% to 10% of breast cancers, however, are associated with breast cancer susceptibility genes, BRCA1 and BRCA2. The cumulative lifetime risk of breast cancer for women who carry the mutant BRCA1 is predicted to be approximately 92%, while the cumulative lifetime risk for the non-carrier majority is estimated to be approximately 10%. BRCA1 is a tumor suppressor gene that is involved in DNA repair and cell cycle control, which are both important for the maintenance of genomic stability. More than 90% of all mutations reported so far result in a premature truncation of the protein product with abnormal or abolished function. The histology of breast cancer in BRCA1 mutation carriers differs from that in sporadic cases, but mutation analysis is the only way to find the carrier. Like BRCA1, BRCA2 is involved in the development of breast cancer, and like BRCA1 plays a role in DNA repair. However, unlike BRCA1, it is not involved in ovarian cancer.

[0011] Other genes have been linked to breast cancer, for example c-erb-2 (HER2) and p53 (Beenken et al., Ann. Surg. 233(5):630-638 (2001). Overexpression of c-erb-2 (HER2) and p53 have been correlated with poor prognosis (Rudolph et al., Hum. Pathol. 32(3):311-319 (2001), as has been aberrant expression products of mdnm2 (Lukas et al., Cancer Res. 61(7):3212-3219 (2001) and cyclin1 and p27 (Porter & Roberts, International Publication WO98/33450, published Aug. 6, 1998).

[0012] The detection of BRCA1 or BRCA2 mutations represents a step towards the design of therapies to better control and prevent the appearance of these tumors. Recently, many studies have used gene expression profiling to analyze various cancers, and those studies have provided new diagnosis and prognosis information in the molecular level. See Zajchowski et al., “Identification of Gene Expression Profiled that Predict the Aggressive Behavior of Breast Cancer Cells,” Cancer Res. 61:5168 (2001); West et al., “Predicting the Clinical Status of Human Breast Cancer by Using Gene Expression Profiles,” Proc. Natl. Acad. Sci. U.S.A. 98:11462 (2001); van 't Veer et al., “Gene Expression Profiling Predicts the Outcome of Breast Cancer,” Nature 415:530 (2002); Roberts et al., “Diagnosis and Prognosis of Breast Cancer Patients,” WO 02/103320; Sorlie et al., Proc. Natl. Acad. Sci. U.S.A. 100:8418 (2003); Perou et al., Nature 406:747 (2000); Khan et al., Cancer Res 58, 5009 (1998); Golub et al., Science 286, 531 (1999); DeRisi et al., Nat. Genet. 14:457 (1996); Alizadeh et al., Nature 403, 503 (2000). Methods for the identification of informative genesets for various cancers have also been described. See Roberts et al., “Diagnosis and Prognosis of Breast Cancer Patients,” WO 02/103320; Golub et al., U.S. Pat. No. 6,647,341.

[0013] Genesets have been identified that are informative for differentiating individuals having, or suspected of having, breast cancer based on estrogen receptor (ER) status, or BRCA1 mutation vs. sporadic (i.e., other than BRCA1-type) mutation status. See Roberts et al., WO 02/103320; van't Veer et al., Nature 415:530 (2001). Genesets have also been identified that enable the classification of sporadic tumor-type individuals as those who will likely have no metastases within five years of initial diagnosis (i.e., individuals with a good prognosis) or those who will likely have a metastasis within five years of initial diagnosis (i.e., those having a poor prognosis). Roberts, supra; van't Veer, supra.

[0014] Roberts et al. WO 02/103320 describes a 70-gene set, useful for the prognosis of breast cancer, which outperformed clinical measures of prognosis, and which showed good potential in selecting good outcome patients, thereby avoiding over-treatment. van de Vijver et al., N. Engl. J. Med. 347:1999 (2002). The expression of genes with most predictive value, however, were not homogeneous among poor patients, suggesting the need for improvement.

[0015] Although the patterns of gene expression as described in Roberts et al. were correlated with existing clinical indicators such as estrogen receptor and BRCA1 status, clinical measures were not incorporated. Furthermore, although the poor-outcome group in particular showed heterogeneity in expression pattern, the best classifier decision rule found during these studies was a fairly simple one based on the similarity of a patient profile to the average profile of a good-outcome training group.

[0016] It is evident that breast cancer is the result of more than one type of molecular event. Likewise, a variety of other conditions, such as other cancers; non-cancer diseases such as diabetes, autoimmune or neurodegenerative disorders, obesity; etc., are also the result of more than one molecular event. Moreover, an individual's response to exposure to particular environmental conditions, for example, exposure to natural or man-made agents, such as toxins, pollutants, drugs, food additives, etc., likely result from more than one molecular event. Thus, there exists a need for improved prognostic methods so that appropriate courses of prophylaxis and/or therapy may be provided. Genesets having improved prognostic power can be identified by first identifying discrete subsets of individuals based on genotypic or phenotypic characteristics relevant to the disease or condition, and then identifying genesets informative for prognosis within those subsets of patients. Individuals having the condition, or who are suspected of having the condition, such as breast cancer, would then be provided therapies appropriate to the molecular mechanisms underlying the condition. The present invention provides such methods for breast cancer, and for other cancers, diseases or conditions.

[0017] The present invention provides methods of identifying relevant subsets of conditions, and the identification of markers relevant to those subsets, for example, for prognosis of individuals classifiable into one of those subsets. The invention further provides sets of markers useful for the prognosis of individuals having breast cancer, wherein those patients have been classified according to one or more characteristics of breast cancer.

[0018] Thus, the present invention provides a method of identifying a set of informative genes or markers for a condition comprising a plurality of phenotypic or genotypic characteristics, comprising: (a) classifying each of a plurality of samples or individuals on the basis of one or more phenotypic or genotypic characteristics of said condition into a plurality of first classes; and (b) identifying within each of said first classes a first set of genes or markers informative for said condition, wherein said first set of genes or markers within each of said first classes is unique to said class relative to other first classes. In a specific embodiment, this method further comprises additionally classifying into a plurality of second classes said samples or individuals in at least one of said first classes on the basis of a phenotypic or genotypic characteristic different that that used in said classifying step (a); and identifying within at least one of said second classes a second set of informative genes or markers, wherein said second set of informative genes or markers within each of said second classes is unique to said second class relative to other first and second classes.

[0019] The invention further provides a method of identifying a set of informative genes or markers for a condition comprising a plurality of phenotypic or genotypic characteristics, comprising: (a) classifying each of a plurality of samples or individuals on the basis of one or more phenotypic or genotypic characteristics into a plurality of first classes; (b) classifying at least one of said first classes into a plurality of second classes on the basis of phenotypic or genotypic characteristic different than that used in said classifying step (a); and (c) identifying within at least one of said first classes or said second classes a set of genes or markers informative for said condition, wherein said second set of genes or markers is unique to said class relative to other first and second classes.

[0020] The invention further provides a method of identifying a set of informative genes or markers for a condition comprising a plurality of phenotypic or genotypic characteristics, comprising: (a) selecting a first characteristic from said plurality of phenotypic or genotypic characteristics; (b) identifying at least two first condition classes differentiable by said first characteristic; (c) selecting a plurality of individuals classifiable into at least one of said first condition classes; and (d) identifying in samples derived from each of said plurality of individuals a set of genes or markers informative for said condition within said at least one of said first condition classes.

[0021] The invention further provides a method of classifying an individual with a condition as having a good prognosis or a poor prognosis, comprising: (a) classifying said individual into one of a plurality of patient classes, said patient classes being differentiated by one or more phenotypic, genotypic or clinical characteristics of said condition; (b) determining the level of expression of a plurality of genes or their encoded proteins in a cell sample taken from the individual relative to a control, said plurality of genes or their encoded proteins comprising genes or their encoded proteins informative for prognosis of the patient class into which said individual is classified; and (c) classifying said individual as having a good prognosis or a poor prognosis on the basis of said level of expression. In a specific embodiment, said condition is cancer, said good prognosis is the non-occurrence of metastases within five years of initial diagnosis, and said poor prognosis is the occurrence of metastases within five years of initial diagnosis. In a more specific embodiment, said cancer is breast cancer. In another specific embodiment, said control is the average level of expression of each of said plurality of genes or their encoded proteins across a plurality of samples derived from individuals identified as having a poor prognosis. In a more specific embodiment, said classifying step (c) is carried out by a method comprising comparing the level of expression of each of said plurality of genes or their encoded proteins to said average level of expression of each corresponding gene or its encoded protein in said control, and classifying said individual as having a poor prognosis if said level of expression correlates with said average level of expression of each of said genes or their encoded proteins in said control more strongly than would be expected by chance. In another specific embodiment, said control is the average level of expression of each of said plurality of genes or their encoded proteins across a plurality of samples derived from individuals identified as having a good prognosis. In a more specific embodiment, said classifying in step (c) is carried out by a method comprising comparing the level expression of each of said plurality of genes or their encoded proteins to said average level of expression of each corresponding gene or its encoded protein in said control, and classifying said individual as having a good prognosis if said level of expression correlates with said average level of expression of each of said genes or their encoded proteins in said control more strongly than would be expected by chance. In another specific embodiment, said plurality of patient classes comprises ER⁻, BRCAI individuals; ER⁻, sporadic individuals; ER+, ER/AGE high individuals; ER+, ER/AGE low, LN+ individuals; and ER+, ER/AGE low, LN⁻ individuals.

[0022] The invention further provides a method of classifying a breast cancer patient as having a good prognosis or a poor prognosis comprising: (a) classifying said breast cancer patient as ER⁻, BRCAI; ER⁻, sporadic; ER+, ER/AGE high; ER+, ER/AGE low, LN+; or ER+, ER/AGE low, LN⁻; (b) determining the level of expression of a first plurality of genes in a cell sample taken from said breast cancer patient relative to a control, said first plurality of genes comprising two of the genes corresponding to the markers in Table 1 if said breast cancer patient is classified as ER⁻, BRCAI; in Table 2 if said breast cancer patient is classified as ER⁻ sporadic; in Table 3 if said breast cancer patient is classified as ER+, ER/AGE high; in Table 4 if said breast cancer patient is classified as ER+. ER/AGE low, LN+; or in Table 5 if said breast cancer patient is classified as ER+, ER/AGE low, LN⁻; and (c) classifying said breast cancer patient as having a good prognosis or a poor prognosis on the basis of the level of expression of said first plurality of genes, wherein said breast cancer patient is “ER/AGE high” if the ratio of the log₁₀(ratio) of ER gene expression to age exceeds a predetermined value, and “ER/AGE low” if the ratio of the log₁₀(ratio) of ER gene expression to age does not exceed said predetermined value. In a specific embodiment, said control is the average level of expression of each of said plurality of genes in a plurality of samples derived from ER⁻, BRCA1 individuals, if said breast cancer patient is ER⁻, BRCA1; the average level of expression of each of said plurality of genes in a plurality of samples derived from ER⁻, sporadic individuals if said breast cancer patient is ER⁻, sporadic; the average level of expression of each of said plurality of genes in a plurality of samples derived from ER+, ER/AGE high individuals, if said breast cancer patient is ER+, ER/AGE high; the average level of expression of each of said plurality of genes in a plurality of samples derived from ER+, ER/AGE low, LN+ individuals where said breast cancer patient is ER+, ER/AGE low, LN+; or the average level of expression of each of said plurality of genes in a plurality of samples derived from ER+, ER/AGE low, LN⁻ individuals where said breast cancer patient is ER+, ER/AGE low, LN⁻. In a more specific embodiment, each of said individuals has a poor prognosis. In another more specific embodiment, each of said individuals has a good prognosis. In an even more specific embodiment, said classifying step (c) is carried out by a method comprising comparing the level of expression of each of said plurality of genes or their encoded proteins in a sample from said breast cancer patient to said control, and classifying said breast cancer patient as having a poor prognosis if said level of expression correlates with said average level of expression of the corresponding genes or their encoded proteins in said control more strongly than would be expected by chance. In another specific embodiment, said predetermined value of ER is calculated as ER=0.1 (AGE−42.5), wherein AGE is the age of said individual. In another specific embodiment, said individual is ER⁻, BRCA1, and said plurality of genes comprises two of the genes for which markers are listed in Table 1. In another specific embodiment, said individual is ER⁻, BRCA1, and said plurality of genes comprises all of the genes for which markers are listed in Table 1. In another specific embodiment, said individual is ER⁻, sporadic, and said plurality of genes comprises two of the genes for which markers are listed in Table 2. said individual is ER⁻, sporadic, and said plurality of genes comprises all of the genes for which markers are listed in Table 2. In another specific embodiment, said individual is ER+, ER/AGE high, and said plurality of genes comprises two of the genes for which markers are listed in Table 3. said individual is ER+, ER/AGE high, and said plurality of genes comprises all of the genes for which markers are listed in Table 3. In another specific embodiment, said individual is ER+, ER/AGE low, LN+, and said plurality of genes comprises two of the genes for which markers are listed in Table 4. In another specific embodiment, said individual is ER+, ER/AGE low, LN+, and said plurality of genes comprises all of the genes for which markers are listed in Table 4. In another specific embodiment, said individual is ER+, ER/AGE low, LN⁻, and said plurality of genes comprises two of the genes for which markers are listed in Table 4. In another specific embodiment, said individual is ER+, ER/AGE low, LN⁻, and said plurality of genes comprises all of the genes for which markers are listed in Table 4. In another specific embodiment, the method further comprises determining in said cell sample the level of expression, relative to a control, of a second plurality of genes for which markers are not found in Tables 1-5, wherein said second plurality of genes is informative for prognosis.

[0023] In another embodiment, the invention provides a method for assigning an individual to one of a plurality of categories in a clinical trial, comprising: (a) classifying said individual as ER⁻, BRCA1, ER⁻, sporadic; ER+, ER/AGE high; ER+, ER/AGE low, LN+; or ER+, ER/AGE low, LN⁻; (b) determining for said individual the level of expression of at least two genes for which markers are listed in Table 1 if said individual is classified as ER⁻, BRCA1; Table 2 if said individual is classified as ER⁻, sporadic; Table 3 if said individual is classified as ER+, ER/AGE high; Table 4 if said individual is classified as ER+, ER/AGE low, LN+; or Table 5 if said individual is classified as ER+, ER/AGE low, LN⁻; (c) determining whether said individual has a pattern of expression of said at least two genes that correlates with a good prognosis or a poor prognosis; and (d) assigning said individual to one category in a clinical trial if said individual has a good prognosis, and assigning said individual to a second category in said clinical trial if said individual has a poor prognosis. In a specific embodiment, said individual is additionally assigned to a category in said clinical trial on the basis of the classification of said individual as determined in step (a). In another specific embodiment, said individual is additionally assigned to a category in said clinical trial on the basis of any other clinical, phenotypic or genotypic characteristic of breast cancer. In another specific embodiment, said method further comprises determining in said cell sample the level of expression, relative to a control, of a second plurality of genes for which markers are not found in Tables 1-5, wherein said second plurality of genes is informative for prognosis of breast cancer, and determining from the expression of said second plurality of genes, in addition to said first plurality of genes, whether said individual has a good prognosis or a poor prognosis.

[0024] The invention further provides a microarray comprising probes complementary and hybridizable to a plurality of the genes for which markers are listed in any of Tables 1-5. The invention further provides a microarray comprising probes complementary and hybridizable to a plurality of the genes for which markers are listed in Table 1, each of the genes for which markers are listed in Table 1, a plurality of the genes for which markers are listed in Table 2, each of the genes for which markers are listed in Table 2, a plurality of the genes for which markers are listed in Table 3, each of the genes for which markers are listed in Table 3, a plurality of the genes for which markers are listed in Table 4, each of the genes for which markers are listed in Table 4, a plurality of the genes for which markers are listed in Table 5, or each of the genes for which markers are listed in Table 5. The invention further provides any one of the above microarrays, wherein said probes are at least 50% of the probes on said microarray. The invention further provides any one of the above microarrays, wherein said probes are at least 90% of the probes on said microarray. The invention further provides microarray comprising probes complementary and hybridizable to a plurality of the genes for which markers are listed in any of Tables 1-5, wherein said probes are complementary and hybridizable to at least 75% of the genes for which markers are listed in Table 1; are complementary and hybridizable to at least 75% of the genes for which markers are listed in Table 2; are complementary and hybridizable to at least 75% of the genes for which markers are listed in Table 3; are complementary and hybridizable to at least 75% of the genes for which markers are listed in Table 4; and are complementary and hybridizable to at least 75% of the genes for which markers are listed in Table 5, wherein said probes, in total, are at least 50% of the probes on said microarray.

[0025] The invention further comprises a kit comprising any one of the above microarrays in a sealed container.

[0026] The invention further provides a method of identifying a set of genes informative for a condition, said condition having a plurality of phenotypic or genotypic characteristics such that samples may be categorized by at least one of said phenotypic or genotypic characteristics into at least one characteristic class, said method comprising: (a) selecting a plurality of samples from individuals having said condition; (b) identifying a first set of genes informative for said characteristic class using said plurality of samples; (c) predicting the characteristic class of each of said plurality of samples; (d) discarding samples for which said characteristic class is incorrectly predicted; (e) repeating steps (c) and (d) at least once; and (f) identifying a second set of genes informative for said characteristic class using samples in said plurality of samples remaining after step (e).

[0027] The invention further provides a method for assigning an individual to one of a plurality of categories in a clinical trial, comprising: (a) classifying the individual into one of a plurality of condition categories differentiated by at least one genotypic or phenotypic characteristic of the condition; (b) determining the level of expression, in a sample derived from said individual, of a plurality of genes informative for said condition category; (c) determining whether said level of expression of said plurality of genes indicates that the individual has a good prognosis or a poor prognosis; and (d) assigning the individual to a category in a clinical trial on the basis of prognosis.

[0028] The invention also provides a method for identifying one or more sets of informative genes or markers for a condition in an organism, comprising: (a) subdividing a plurality of individuals or samples derived therefrom of the organism subject to the condition into a plurality of classes based on one or more clinical, phenotypic or genotypic characteristics of the organism, wherein each class consists of a plurality of individuals or samples derived therefrom of the organism each of which having one or more clinical, phenotypic or genotypic characteristics specific for the class; and (b) attempting to identify for each of one or more of said plurality of classes a set of genes or markers informative for said condition in individuals in said class, wherein, if a set of genes or markers informative for said condition in individuals in said class is obtained for any of said one or more of said plurality of classes, said set of genes or markers is taken as a set of informative genes or markers for said condition in said organism.

[0029] In one embodiment, the method further comprises, for each of one or more of said classes in which a set of genes or markers informative for said condition in individuals in said class cannot be obtained, repeating said steps (a) and (b) on said plurality of individuals or samples derived therefrom in said class such that said plurality of individuals or samples derived therefrom in said class is subdivided into a plurality of additional classes based on one or more clinical, phenotypic or genotypic characteristics of said organism which are different from those used for defining said class, wherein for each of said plurality of additional classes, if a set of genes or markers informative for said condition in individuals in said class is obtained, said set of genes or markers is taken as a set of informative genes or markers for said condition in said organism.

[0030] The invention also provides a method for identifying one or more sets of informative genes or markers for a condition in an organism, comprising: (a) subdividing a plurality of individuals or samples derived therefrom of said organism subject to said condition into a plurality of classes based on one or more clinical, phenotypic or genotypic characteristics of said organism, wherein each said class consists of a plurality of individuals or samples derived therefrom of said organism each having said one or more clinical, phenotypic or genotypic characteristics specific for said class; (b) attempting to identify for each of one or more of said plurality of classes a set of genes or markers informative for said condition in individuals in said class, wherein if a set of genes or markers informative for said condition in individuals in said class is identified for any of said one or more of said classes, said set of genes or markers is taken as a set of informative genes or markers for a condition in said organism; and (c) for each of one or more of said classes in which a set of genes or markers informative for said condition in individuals in said class cannot be obtained, repeating said steps (a) and (b) on said plurality of individuals or samples derived therefrom in said class such that said plurality of samples or individuals in said class is subdivided into a plurality of additional classes based on one or more clinical, phenotypic or genotypic characteristics of said organism which are different from those used those used for defining said class, wherein for each of one or more of said plurality of additional classes, if a set of genes or markers informative for said condition in individuals in said class is obtained, said set of genes or markers is taken as a set of informative genes or markers for a condition in said organism.

[0031] In the methods of the invention, the condition can be a type of cancer. In such an embodiment, each of said sets of genes or markers can be informative of prognosis of individuals in a corresponding class. In one embodiment, the condition is breast cancer, and the one or more clinical, phenotypic or genotypic characteristics comprise age, ER level, ER/AGE, BRAC1 status, and lymph node status.

[0032] In one embodiment, the methods of the invention further comprise generating a template profile comprising measurements of levels of genes or markers of the set of informative genes or markers for said class representative of levels of the genes or markers in a plurality of patients having a chosen prognosis level.

[0033] The invention also provides a method for predicting a breast cancer patient as having a good prognosis or a poor prognosis, comprising: (a) classifying said breast cancer patient into one of the following classes: (a1) ER⁻, BRCAI; (a2) ER⁻, sporadic; (a3) ER+, ER/AGE high; (a4) ER+, ER/AGE low, LN+; or (a5) ER+, ER/AGE low, LN⁻; (b) determining a profile comprising measurements of a plurality of genes or markers in a cell sample taken from said breast cancer patient, said plurality of genes markers comprising at least two of the genes or markers corresponding to the markers in (b1) Table 1 if said breast cancer patient is classified as ER⁻, BRCAI; (b2) Table 2 if said breast cancer patient is classified as ER⁻ sporadic; (b3) Table 3 if said breast cancer patient is classified as ER+, ER/AGE high; (b4) Table 4 if said breast cancer patient is classified as ER+, ER/AGE low, LN+; or (b5) Table 5 if said breast cancer patient is classified as ER+, ER/AGE low, LN⁻; and (c) classifying said breast cancer patient as having a good prognosis or a poor prognosis based on said profile of said plurality of genes or markers, wherein ER⁺ designates a high ER level and ER⁻ designates a low ER level, wherein said ER/AGE is a metric of said ER level relative to the age of said patient, and wherein LN⁺ designates a greater than 0 lymph nodes status in said patient and LN⁻ designates a 0 lymph nodes status in said patient.

[0034] In one embodiment, step (c) is carried out by a method comprising comparing said profile to a good prognosis template and/or a poor prognosis template, and wherein said patient is classified as having a good prognosis if said profile has a high similarity to a good prognosis template or has a low similarity to a poor prognosis template or as having a poor prognosis if said profile has a low similarity to a good prognosis template or has a high similarity to a poor prognosis template. A good prognosis template comprises measurements of said plurality of genes or markers representative of levels of said genes or markers in a plurality of good outcome patients, while a poor prognosis template comprises measurements of said plurality of genes or markers representative of levels of said genes or markers in a plurality of poor outcome patients. Here a good outcome patient is a breast cancer patient who has non-reoccurrence of metastases within a first period of time after initial diagnosis, while a poor outcome patient is a patient who has reoccurrence of metastases within a second period of time after initial diagnosis.

[0035] In another embodiment, the methods for predicting the prognosis of a breast cancer patient further comprise determining said profile, said ER level, said LN status, and/or, said ER/AGE. In one embodiment, said profile is an expression profile comprising measurements of a plurality of transcripts in a sample derived from said patient, wherein said good prognosis template comprises measurements of said plurality of transcripts representative of expression levels of said transcripts in said plurality of good outcome patients, and wherein said poor prognosis template comprises measurements of said plurality of transcripts representative of expression levels of said transcripts in said plurality of poor outcome patients.

[0036] In one embodiment, said expression profile is a differential expression profile comprising differential measurements of said plurality of transcripts in said sample derived from said patient versus measurements of said plurality of transcripts in a control sample.

[0037] In one embodiment, the measurement of each said transcript in said good prognosis template is an average of expression levels of said transcript in said plurality of good outcome patients.

[0038] In one embodiment, the similarity of said expression profile to said good or poor prognosis template is represented by a correlation coefficient between said expression profile and said good or poor prognosis template, respectively, and a correlation coefficient greater than a correlation threshold, e.g., 0.5, indicates a high similarity and said correlation coefficient equal to or less than said correlation threshold indicates a low similarity.

[0039] In another embodiment, the similarity of said expression profile to said good or poor prognosis template is represented by a distance between said cellular constituent profile and said good or poor prognosis template, respectively, and a distance less than a given value indicates a high similarity and said distance equal to or greater than said given value indicates a low similarity.

[0040] In another embodiment, said profile comprises measurements of a plurality of protein species in a sample derived from said patient, wherein said good prognosis template comprises measurements of said plurality of protein species representative of levels of said protein species in said plurality of good outcome patients, and wherein said poor prognosis template comprises measurements of said plurality of protein species representative of levels of said protein species in said plurality of poor outcome patients.

[0041] In one embodiment, said ER level is determined by measuring an expression level of a gene encoding said estrogen receptor, e.g., the estrogen receptor α gene, in said patient relative to expression level of said gene in said control sample, and said ER level is classified as ER⁺ if log 10(ratio) of said expression level is greater than −0.65, and said ER level is classified as ER⁻ if log 10(ratio) of said expression level is equal to or less than −0.65.

[0042] In one embodiment, said ER/AGE is classified as high if said ER level is greater than c·(AGE−d), and said ER/AGE is classified as low if said ER level is equal to or less than c·(AGE−d), wherein c is a coefficient, AGE is the age of said patient, and d is an age threshold.

[0043] In a specific embodiment, said estrogen receptor level is measured by a polynucleotide probe that detects a transcript corresponding to the gene having accession number NM_—000125, said control sample is a pool of breast cancer cells of different patients, and c=0.1 and d=42.5.

[0044] In one embodiment, said control sample is generated by pooling together cDNAs of said plurality of transcripts from a plurality of breast cancer patients. In another embodiment, said control sample is generated by pooling together synthesized cDNAs of said plurality of transcripts and said transcript of said gene encoding said estrogen receptor.

[0045] In one embodiment, said individual is ER⁻, BRCA1, and said plurality of genes comprises at least two of the genes for which markers are listed in Table 1. In one embodiment, said individual is ER⁻, BRCA1, and said plurality of genes comprises all of the genes for which markers are listed in Table 1.

[0046] In another embodiment, the individual is ER⁻, sporadic, and said plurality of genes comprises at least two of the genes for which markers are listed in Table 2. In one embodiment, said individual is ER⁻, sporadic, and said plurality of genes comprises all of the genes for which markers are listed in Table 2.

[0047] In still another embodiment, said individual is ER+, ER/AGE high, and said plurality of genes comprises at least two of the genes for which markers are listed in Table 3. In one embodiment, said individual is ER+, ER/AGE high, and said plurality of genes comprises all of the genes for which markers are listed in Table 3.

[0048] In still another embodiment, said individual is ER+, ER/AGE low, LN+, and said plurality of genes comprises at least two of the genes for which markers are listed in Table 4.

[0049] In one embodiment, said individual is ER+, ER/AGE low, LN+, and said plurality of genes comprises all of the genes for which markers are listed in Table 4.

[0050] In still another embodiment, said individual is ER+, ER/AGE low, LN⁻, and said plurality of genes comprises at least two of the genes for which markers are listed in Table 4. In one embodiment, the individual is ER+, ER/AGE low, LN⁻, and said plurality of genes comprises all of the genes for which markers are listed in Table 4.

[0051] In one embodiment, said profile further comprises one or more genes for which markers are not found in Tables 1-5, which are informative for prognosis.

[0052] The invention also provides a method for assigning an individual to one of a plurality of categories in a clinical trial, comprising assigning said individual to one category in a clinical trial if said individual has a good prognosis as determined by any one of the methods described above, and assigning said individual to a second category in said clinical trial if said individual has a poor prognosis as determined by any one of the methods described above.

[0053] In one embodiment, said individual is additionally assigned to a category in said clinical trial on the basis of the classification of said individual based on said profile, said ER level, said LN status, and/or, said ER/AGE.

[0054] In one embodiment, said individual is additionally assigned to a category in said clinical trial on the basis of one or more other clinical, phenotypic or genotypic characteristic of breast cancer.

[0055] In one embodiment, the method further comprises determining in said cell sample the levels of expression of said one or more genes for which markers are not found in Tables 1-5, and determining from said expression levels of said one or more genes, whether said individual has a good prognosis or a poor prognosis.

[0064] 5.1 Introduction

[0065] The present invention provides methods for classifying individuals having a condition, such as a disease, into one or more subsets of individuals, where individuals in each subset are characterized by one or more phenotypic or genotypic characteristics of the condition. The individuals may be eukaryotes or prokaryotes, may be animals such as mammals, including but not limited to humans, primates, rodents, felines, canines, etc., birds, reptiles, fish, etc. “Individuals” as used herein also encompasses single-celled organisms, or colonies thereof, such as bacteria and yeast. The condition may be a disease, such as cancer, and may be a specific cancer, such as breast cancer. The condition may also be an environmental condition, such as exposure to a toxin, pollutant, drug, proximity to urban or industrial areas, etc.

[0066] The present invention provides methods of determining the prognosis of individuals having a condition, such as cancer, for example, breast cancer, or who are suspected of having the condition, by the use of a combination of clinical, biological or biochemical parameters of the condition and gene expression pattern data. For prognosis, the parameters selected preferably relate to or affect the progression and/or outcome of the condition. The pattern of gene expression within a subset of individuals having the particular condition leads to the identification of sets of genes within a subset that is informative for that subset, for example, for prognosis within that subset. In general, the successful identification of sets of genes informative for prognosis within a particular subset justifies the selection of the plurality of clinical, biological or biochemical parameters of the condition on which division of individuals into condition subsets is based.

[0067] In the example of breast cancer, patient groups are first classified according to at least one of age, lymph node (LN) status, estrogen receptor (ER) level, and BRCA1 mutation status into discrete patient subsets. These clinical factors have been implicated in tumor etiology as well as differences in disease outcome. These characteristics are not limiting; other genotypic or phenotypic characteristics of breast cancer, for example, tumor grade, tumor size, tumor cell type, etc., may also be used, alone or in combination with those listed herein, in order to classify individuals. The differences in gene expression or in tumor fate related to these parameters likely represent differences in tumor origin and tumor genesis, and are therefore good candidates for tumor stratification. Genesets informative for prognosis within each subset are then identified. New breast cancer patients are then classified using the same criteria, and a prognosis is made based on the geneset specific for the patient subset into which the patient falls. In the process of constructing a prognosis classifier within each patient subset, particular attention is paid to the homogeneous patterns related to the tumor outcome. Emergence of such homogeneous prognosis patterns may indicate the most common mechanism to metastasis within a subset. At the same time, successful identification of such patterns also justifies the parameters being used for the tumor stratification. To differentiate this approach from an mRNA-alone approach, the current approach of integrating clinical data with the gene expression data is referred to herein as a “comprehensive prognosis”.

[0068] 5.2 Definitions

[0069] As used herein, “BRCA1 tumor” or “BRCA1 type” means a tumor having cells containing a mutation of the BRCA1 locus.

[0070] The “absolute amplitude” of correlation means the absolute value of the correlation; e.g., both correlation coefficients −0.35 and 0.35 have an absolute amplitude of 0.35.

[0071] “Marker” means a cellular constituent, or a modification of a cellular constituent (e.g., an entire gene, EST derived from that gene, a protein encoded by that gene, post-translational modification of the protein, etc.) the expression or level of which changes between certain conditions. Where a change in a characteristic of the constituent correlates with a certain condition, the constituent is a marker for that condition.

[0072] “Marker-derived polynucleotides” means the RNA transcribed from a marker gene, any cDNA or cRNA produced therefrom, and any nucleic acid derived therefrom, such as synthetic nucleic acid having a sequence derived from the gene corresponding to the marker gene.

[0073] A “similarity value” is a number that represents the degree of similarity between two things being compared. For example, a similarity value may be a number that indicates the overall similarity between a patient's expression profile of specific phenotype-related markers and a template specific to that phenotype (for instance, the similarity to a “good prognosis” template, where the phenotype is a good prognosis). The similarity value may be expressed as a similarity metric, such as a correlation coefficient, or may simply be expressed as the expression level difference, or the aggregate of the expression level differences, between a patient sample and a template.

[0074] A “patient subset” is a group of individuals, all of whom have a particular condition, or are subject to a particular condition, which is distinguished from other individuals having that condition by one or more phenotypic, genotypic or clinical characteristics of the condition, or of a response to the condition. For example, where the condition is breast cancer, individuals may belong to an “ER⁺” or an “ER⁻” patient subset, or may belong to a particular age group patient subset.

[0075] A gene and/or marker is “informative” for a condition, phenotype, genotype or clinical characteristic if the expression of the gene or marker is correlated or anticorrelated with the condition, phenotype, genotype or clinical characteristic to a greater degree than would be expected by chance.

[0076] An individual of a given age can be classified as “ER/AGE high” if the individual's ER level is higher than a threshold value for the given age. The threshold can be age-dependent, i.e., a different threshold for each different age. In one embodiment, the age-dependent threshold value is calculated as c·(AGE−d), where c is a coefficient, AGE is the age of the patient, and d is an age threshold. The parameters c and d depend on the ER level and AGE used. They can be determined by fitting patients' ER level-age distribution to a bimodal distribution of two subgroups each having a different ER level-age dependence. In a specific embodiment, c=0.1 and d=42.5 is used for ER levels represented by a log(ratio) of ER expression level. Thus, for example, the threshold for a 45-year old individual in this embodiment is 0.1 (45-42.5), or 0.25, and if the log(ratio) of ER expression level of the individual is equal to or greater than 0.25, the individual is classified as “ER/AGE high”; otherwise, the individual is classified as “ER/AGE low.”

[0077] 5.3 Identification of Diagnostic and Prognostic Marker Sets

[0078] 5.3.1 Identification of Condition Subsets

[0079] The present invention provides methods of identifying sets of genes and/or markers useful in the diagnosis and prognosis of breast cancer. More generally, the invention also provides methods of identifying sets of genes and/or markers useful in the diagnosis or prognosis of other cancers, and even more generally, of identifying sets of genes and/or markers useful in the differentiation between subgroups of individuals having a particular condition, such as a disease or exposure to a particular environmental condition.

[0080] The method may be applied to any condition for which a plurality of phenotypic or genotypic subsets may be identified. The condition may be a disease; for example, the condition may be cancer, an autoimmune disease, an inflammatory disease, an infectious disease, a neurological disease, a degenerative disease, etc. The condition may be environmental; for example, the condition may be a particular diet, geographic location, etc.; the condition may be exposure to a compound, including, for example, a drug, a toxin, a carcinogen, a foodstuff, a poison, an inhaled compound, an ingested compound, etc.; the condition may be a particular genetic background or predisposition to a medical condition; etc.

[0081] Where the condition is cancer, the condition may be any cancer, for example, without limitation: leukemias, including acute leukemia, acute lymphocytic leukemia, acute myelocytic leukemia, myeloblastic leukemia, promyelocytic leukemia, myelomonocytic leukemia, monocytic leukemia, and erythroleukemia; chronic leukemia, such as chronic myelocytic (granulocytic) leukemia or chronic lymphocytic leukemia; polycythemia vera; lymphomas, such as Hodgkin's disease and non-Hodgkin's disease; multiple myeloma; Waldenström's macroglobulinemia; heavy chain disease; solid tumors, such as sarcomas and carcinomas, fibrosarcoma, myxosarcoma, liposarcoma, chondrosarcoma, osteogenic sarcoma, chordoma, angiosarcoma, endotheliosarcoma, lymphangiosarcoma, lymphangioendotheliosarcoma, synovioma, mesothelioma, Ewing's tumor, leiomyosarcoma, rhabdomyosarcoma, colon carcinoma, pancreatic cancer, breast cancer, ovarian cancer, prostate cancer, squamous cell carcinoma, basal cell carcinoma, adenocarcinoma, sweat gland carcinoma, sebaceous gland carcinoma, papillary carcinoma, papillary adenocarcinomas, cystadenocarcinoma, medullary carcinoma, bronchogenic carcinoma, renal cell carcinoma, hepatoma, bile duct carcinoma, choriocarcinoma, seminoma, embryonal carcinoma, Wilms' tumor, cervical cancer, testicular tumor, lung carcinoma, small cell lung carcinoma, bladder carcinoma, epithelial carcinoma, glioma, astrocytoma, medulloblastoma, craniopharyngioma, ependymoma, pinealoma, hemangioblastoma, acoustic neuroma, oligodendroglioma, menangioma, melanoma, neuroblastoma, or retinoblastoma; etc.

[0082] Rather than stratifying individuals, such as patients or tumor samples derived from patients, by gene expression patterns in the first instance, the method of identifying sets of genes informative for a condition begins by identifying phenotypic, genotypic or clinical subsets of individuals within the larger class of individuals having or affected by the condition.

[0083] In one embodiment, the condition is cancer, and the subsets are distinguished by phenotypic, genotypic, and/or clinical characteristics of the cancer. In this embodiment, groups of individuals are classified according to one or more phenotypic, genotypic, or clinical characteristics relevant to the cancer into patient subsets. At any step in the process of subdividing a patient population into patient subsets, the expression level of one or more genes may be determined in order to identify whether a prognosis-informative set of genes may be identified for the particular patient subset. If an informative gene set is identified, but is not as informative as desired, the patient subset may be further divided and a new geneset identified. These subsets may be further subdivided. For example, a group of individuals affected by a particular cancer may be classified first on the basis of a phenotypic, genotypic or clinical characteristic A into subsets S1 and S2. The levels of expression of a plurality of genes are then determined in tumor samples taken from individuals that fall within subsets S1 or S2 in order to identify sets of genes informative for prognosis within these subsets. Subsets S1 and S2 may then each be subdivided into two or more subsets based on other phenotypic, genotypic or clinical characteristics. The basis for subdivision, if performed, need not be the same for S1 and S2. For example, in various embodiments, S1 is not subdivided, while S2 is subdivided on the basis of characteristic B; or S1 is subdivided based on characteristic B while S2 is not subdivided; or S1 and S2 are both subdivided on the basis of characteristic B; or S1 is subdivided based on characteristic B, while S2 is subdivided according to characteristic C; and so on. For a particular decision matrix leading to a plurality of patient subsets, the preferred outcome is a prognosis-informative set of genes for each patient subset. Different decision matrices may lead to different patient subsets, which, in turn, may result in different sets of prognosis-informative genes.

[0084] In the specific example of breast cancer, a plurality of phenotypic, genotypic or clinical indications are used to classify a patient as being a member of one of a plurality of patient subsets, wherein the indications are medically, biochemically or genetically relevant to breast cancer. For example, a group of patients may be classified into patient subsets based on criteria including, but not limited to, estrogen receptor (ER) status, type of tumor (i.e., BRCA1-type or sporadic), lymph node status, grade of cancer, invasiveness of the tumor, or age. “BRCA1-type” indicates that the BRCA1 mutation is present. In each classification step, a group of cancer patients may be classified into only two classes, for example, ER+ or ER⁻, or into three or more subsets (for example, by tumor grade), depending upon the characteristic used to determine the subsets. As used herein, “ER+” indicates that the estrogen receptor is expressed at some elevated level; for example, it may indicate that the estrogen receptor is detectably expressed, or may indicate that more than 10% of cells are histologically stained for the receptor, etc. Conversely, “ER−” indicates that the estrogen receptor is expressed at a reduced level or not at all; for example, it may indicate that the receptor is not detectably expressed, or that 10% or less of cells are histologically stained for the receptor, etc. Marker gene sets optimized for each phenotypic class are preferably determined after the subsets are established. Where informative markers for a particular patient subset, distinguished from another subset by a particular characteristic of the condition of interest, cannot be determined, the subset may be further divided by another characteristic of the condition to create a plurality of second patient subsets, whereupon genes informative for these second patient subsets may be identified.

[0085] FIG. 1 depicts the process, described in the Examples, of subdivision of a collection of breast cancer patients according to phenotypic and genotypic characteristics relevant to breast cancer, in preparation for identification of genes informative for prognosis. A collection of breast cancer tumor samples was first subdivided by estrogen receptor status. ER status was chosen because the presence or absence of the estrogen receptor greatly influences the expression of other genes. In the ER+patient subset, it was noted that patients appeared to be bimodally distributed by ER level vs. age; that is, ER level dependence upon age tended to fall within two classes, as separated by the solid line in FIG. 2A. This bimodality was used to further subdivide ER+individuals into “ER+, ER/AGE high” individuals and “ER+, ER/AGE low” individuals. A set of informative genes was identified for the ER+, ER/AGE high patient subset. An informative set was not identified for the ER+, ER/AGE low subset, however, so the subset of patients was further divided into LN+ and LN− individuals. Thus, in one embodiment, the present invention provides a method of identifying a set of informative genes or markers for a condition comprising a plurality of phenotypic or genotypic characteristics, comprising (a) classifying each of a plurality of samples or individuals on the basis of one phenotypic or genotypic characteristic into a plurality of first classes; and (b) identifying within each of said first classes a set of informative genes or markers, wherein said set of informative genes or markers within each said first classes is unique to said class.

[0086] 5.3.2 Identification of Marker Sets Informative for Patient Subsets

[0087] Once a patient subset is identified, markers, such as genes, informative for a particular outcome, such as prognosis, may be identified. In one embodiment, the method for identifying marker sets is as follows. This example describes the use of genes and gene-derived nucleic acids as markers; however, proteins or other cellular constituents may be used as markers of the condition.

[0088] After extraction and labeling of target polynucleotides, the expression of a plurality of markers, such as genes, in a sample X is compared to the expression of the plurality markers in a standard or control. In one embodiment, the standard or control comprises target markers, such as polynucleotide molecules, derived from one or more samples from a plurality of normal individuals, or a plurality of individuals not exposed to a particular condition. For example, the control, or normal, individuals may be persons without the particular disease or condition of interest (e.g., individuals not afflicted with breast cancer, where breast cancer is the disease of interest), or may be an individual not exposed to a particular environmental condition. The standard or control may also comprise target polynucleotide molecules, derived from one or more samples derived from individuals having a different form or stage of the same disease; a different disease or different condition, or individuals exposed or subjected to a different condition, than the individual from which sample X was obtained. The control may be a sample, or set of samples, taken from the individual at an earlier time, for example, to assess the progression of a condition, or the response to a course of therapy.

[0089] In a preferred embodiment, the standard or control is a pool of target polynucleotide molecules. However, where protein levels, or the levels of any other relevant biomolecule, are to be compared, the pool may be a pool of proteins or the relevant biomolecule. In a preferred embodiment in the context of breast cancer, the pool comprises samples taken from a number of individuals having sporadic-type tumors.

[0090] In another preferred embodiment, the pool comprises an artificially-generated population of nucleic acids designed to approximate the level of nucleic acid derived from each marker found in a pool of marker-derived nucleic acids derived from tumor samples. In another embodiment, the pool, also called a “mathematical sample pool,” is represented by a set of expression values, rather than a set of physical polynucleotides; the level of expression of relevant markers in a sample from an individual with a condition, such as a disease, is compared to values representing control levels of expression for the same markers in the mathematical sample pool. Such a control may be a set of values stored on a computer. Such artificial or mathematical controls may be constructed for any condition of interest.

[0091] In another embodiment specific to breast cancer, the pool is derived from normal or breast cancer cell lines or cell line samples. In a preferred embodiment, the pool comprises samples taken from individuals within a specific patient subset, e.g., “ER+, ER/AGE high” individuals, wherein each of said individuals has a good prognosis, or each of said individuals has a poor prognosis. Of course, where, for example, expressed proteins are used as markers, the proteins are obtained from the individual's sample, and the standard or control could be a pool of proteins from a number of normal individuals, or from a number of individuals having a particular state of a condition, such as a pool of samples from individuals having a particular prognosis of breast cancer.

[0092] The comparison may be accomplished by any means known in the art. For example, expression levels of various markers may be assessed by separation of target polynucleotide molecules (e.g., RNA or cDNA) derived from the markers in agarose or polyacrylamide gels, followed by hybridization with marker-specific oligonucleotide probes. Alternatively, the comparison may be accomplished by the labeling of target polynucleotide molecules followed by separation on a sequencing gel. Polynucleotide samples are placed on the gel such that patient and control or standard polynucleotides are in adjacent lanes. Comparison of expression levels is accomplished visually or by means of densitometer. In a preferred embodiment, the expression of all markers is assessed simultaneously by hybridization to a microarray. In each approach, markers meeting certain criteria are identified as informative for the prognosis of breast cancer.

[0093] Marker genes are selected based upon significant difference of expression in a condition, such as a disease, as compared to a standard or control condition. Marker genes may be screened, for example, by determining whether they show significant variation within a set of samples of interest. Genes that do not show a significant amount of variation within the set of samples are presumed not to be informative for the disease or condition, and are not selected as markers for the disease or condition. Genes showing significant variation within the sample set are candidate informative genes for the disease or condition. The degree of variation may be estimated by calculating the difference of the expression of the gene, or ratio of expression between sample and control, within the set of samples. The expression, or ratio of expressions, may be transformed by any means, e.g., linear or log transformation. Selection may be made based upon either significant up- or down regulation of the marker in the patient sample. Selection may also be made by calculation of the statistical significance (i.e., the p-value) of the correlation between the expression of the marker and the disease and condition. Preferably, both selection criteria are used. Thus, in one embodiment of the present invention, markers associated with prognosis of breast cancer within a patient subset are selected where the markers show both more than two-fold change (increase or decrease) in expression as compared to a standard, and the p-value for the correlation between the existence of breast cancer and the change in marker expression is no more than 0.01 (i.e., is statistically significant).

[0094] In the context of the present invention, “good prognosis” indicates a desired outcome for a particular condition, especially a particular disease, and “poor prognosis” indicates an undesired outcome of the condition. For example, where the condition is cancer, a “good prognosis” may mean partial or complete remission, and “poor prognosis” may mean reappearance of the cancer after treatment. What constitutes “good prognosis” and “poor prognosis” is specific to the condition of interest, for example, specific to the particular cancer an individual suffers. For example, “good prognosis” for pancreatic cancer may be survival for one or two years after initial diagnosis, while “good prognosis” for Hodgkin's disease may be survival for five years or more. In the specific example of breast cancer, “good prognosis” means the likelihood of non-reoccurrence of metastases within a period of 1, 2, 3, 4, 5 or more years after initial diagnosis, and “poor prognosis” means the likelihood of reoccurrence of metastasis within that period. In a more specific example, “good prognosis” means the likelihood of non-reoccurrence of metastases within 5 years after initial diagnosis, and “poor prognosis” means the likelihood of reoccurrence of metastasis within that period.

[0095] In a more specific embodiment for cancer, for example, breast cancer, using a number of breast cancer tumor samples, markers are identified by calculation of correlation coefficients ρ between the clinical category or clinical parameter(s) {right arrow over (c)} and the linear, logarithmic or any transform of the expression ratio {right arrow over (r)} across all samples for each individual gene. Specifically, the correlation coefficient may be calculated as:
ρ=({right arrow over (c)}·{right arrow over (r)})/(∥{right arrow over (c)}∥·∥{right arrow over (r)}∥)  Equation (1)

[0096] Markers for which the coefficient of correlation exceeds a cutoff are identified as prognosis-informative markers specific for a particular clinical type, e.g., good prognosis, within a given patient subset. Such a cutoff or threshold may correspond to a certain significance of discriminating genes obtained by Monte Carlo simulations. The threshold depends upon the number of samples used; the threshold can be calculated as 3×1/√{square root over (n−3)}, where 1/√{square root over (n−3)} is the distribution width and n=the number of samples. In a specific embodiment, markers are chosen if the correlation coefficient is greater than about 0.3 or less than about −0.3.

[0097] Next, the significance of the correlation is calculated. This significance may be calculated by any statistical means by which such significance is calculated. In a specific example, a set of correlation data is generated using a Monte-Carlo technique to randomize the association between the expression difference of a particular marker and the clinical category. The frequency distribution of markers satisfying the criteria in the Monte-Carlo runs is used to determine whether the number of markers selected by correlation with clinical data is significant.

[0098] Once a marker set is identified, the markers may be rank-ordered in order of significance of discrimination. One means of rank ordering is by the amplitude of correlation between the change in gene expression of the marker and the specific condition being discriminated. Another, preferred, means is to use a statistical metric. In a specific embodiment, the metric is a t-test-like statistic:

[0099] $\begin{array}{cc}t=\frac{\left(\langle x_1\rangle -\langle x_2\rangle \right)}{\sqrt{\left[{\sigma }_1^2\left(n_1-1\right)+{\sigma }_2^2\left(n_2-1\right)\right]\text{/}\left(n_1+n_2-1\right)\text{/}\left(1\text{/}{n}_1+1\text{/}{n}_2\right)}}& \mathrm{Equation}\left(2\right)\end{array}$

[0100] In this equation, x₁ is the error-weighted average of the log ratio of transcript expression measurements within a first clinical group (e.g., good prognosis), x₂ is the error-weighted average of log ratio within a second, related clinical group (e.g., poor prognosis), σ₁ is the variance of the log ratio within the first clinical group (e.g., good prognosis), n₁ is the number of samples for which valid measurements of log ratios are available, σ₂ is the variance of log ratio within the second clinical group (e.g., poor prognosis), and n₂ is the number of samples for which valid measurements of log ratios are available. The t-value represents the variance-compensated difference between two means.

[0101] The rank-ordered marker set may be used to optimize the number of markers in the set used for discrimination. This is accomplished generally in a “leave one out” method as follows. In a first run, a subset, for example five, of the markers from the top of the ranked list is used to generate a template, where out of X samples, X−1 are used to generate the template, and the status of the remaining sample is predicted. This process is repeated for every sample until every one of the X samples is predicted once. In a second run, additional markers, for example five additional markers, are added, so that a template is now generated from 10 markers, and the outcome of the remaining sample is predicted. This process is repeated until the entire set of markers is used to generate the template. For each of the runs, type 1 error (false negative) and type 2 errors (false positive) are counted; the optimal number of markers is that number where the type 1 error rate, or type 2 error rate, or preferably the total of type 1 and type 2 error rate is lowest.

[0102] For prognostic markers, validation of the marker set may be accomplished by an additional statistic, a survival model. This statistic generates the probability of tumor distant metastases as a function of time since initial diagnosis. A number of models may be used, including Weibull, normal, log-normal, log logistic, log-exponential, or log-Rayleigh (Chapter 12 “Life Testing”, S-PLUS 2000 GUIDE TO STATISTICS, Vol. 2, p. 368 (2000)). For the “normal” model, the probability of distant metastases P at time t is calculated as
P=α×exp(−t²/τ²)  Equation (3)

[0103] where α is fixed and equal to 1, and τ is a parameter to be fitted and measures the “expected lifetime”.

[0104] It is preferable that the above marker identification process be iterated one or more times by excluding one or more samples from the marker selection or ranking (i.e., from the calculation of correlation). Those samples being excluded are the ones that can not be predicted correctly from the previous iteration. Preferably, those samples excluded from marker selection in this iteration process are included in the classifier performance evaluation, to avoid overstating the performance.

[0105] It will be apparent to those skilled in the art that the above methods, in particular the statistical methods described above, are not limited to the identification of markers associated with the prognosis of breast cancer within a particular patient subset, but may be used to identify set of marker genes associated with any phenotype or condition, or with any subset of a phenotype or condition defined by one or more characteristics of the phenotype or condition. The phenotype or condition can be the presence or absence of a disease such as cancer, or the presence or absence of any identifying clinical condition associated with that cancer. In the disease context, the phenotype may be a prognosis such as a survival time, probability of distant metastases of a disease condition, or likelihood of a particular response to a therapeutic or prophylactic regimen. The phenotype need not be cancer, or a disease; the phenotype may be a nominal characteristic associated with a healthy individual.

[0106] Thus, the invention provides a method of identifying a set of informative genes or markers for a condition comprising a plurality of phenotypic of genotypic characteristics, comprising: (a) classifying each of a plurality of samples or individuals on the basis of one or more phenotypic or genotypic characteristics of said condition into a plurality of first classes; (b) identifying within each of said first classes a first set of genes or markers informative for said condition, wherein said first set of genes or markers within each of said first classes is unique to said class relative to other classes. In a specific embodiment, samples or individuals in at least one of said first classes are additionally classified on the basis of a phenotypic or genotypic characteristic different from that used to distinguish said first classes into a plurality of second classes, and identifying within at least one of said second classes a second set of informative genes or markers, wherein said second set of informative genes or markers within each of said second classes is unique to said second class relative to other classes. In another embodiment, the invention provides a method of identifying a set of informative genes or markers for a condition comprising a plurality of phenotypic or genotypic characteristics, comprising: (a) classifying each of a plurality of samples or individuals on the basis of one or more phenotypic or genotypic characteristics into a plurality of first classes; (b) classifying at least one of said first classes into a plurality of second classes on the basis of phenotypic or genotypic characteristic different than that used to distinguish said plurality of first classes; (c) identifying within at least one of said first classes or said second classes a set of genes or markers informative for said condition, wherein said set of genes or markers is unique to said class relative to other classes. The invention further provides a method of identifying a set of informative genes or markers for a condition comprising a plurality of phenotypic or genotypic characteristics, comprising: (a) selecting a first characteristic from said plurality of phenotypic or genotypic characteristics; (b) identifying at least two first condition classes differentiable by said first characteristic; (c) selecting a plurality of individuals classifiable into at least one of said first condition classes; and (d) identifying in samples derived from each of said plurality of individuals a set of genes or markers informative for said condition within said at least one of said first condition classes.

[0107] 5.3.3 Classifier Genesets for Five Patient Subsets

[0108] The present invention provides sets of markers useful for the prognosis of breast cancer. The markers were identified according to the above methods in specific subsets of individuals with breast cancer. Generally, the marker sets were identified within a population of breast cancer patients that had been first stratified into five phenotypic categories based on criteria relevant to breast cancer prognosis, including estrogen receptor (ER) status, lymph node status, type of mutation(s) (i.e., BRCA1-type or sporadic) and age at diagnosis. More specifically, patients, and tumors from which samples were taken, were classified as ER⁻, sporadic (i.e., being both estrogen receptor negative and having a non-BRCA1-type tumor); ER⁻, BRCA1 (i.e., being both estrogen receptor negative and having a BRCA1-type tumor); ER+, ER/AGE high (i.e., estrogen receptor positive with a high ratio of the log (ratio) of estrogen receptor gene expression to age); ER+, ER/AGE low, LN+ (i.e., estrogen receptor positive with a low ratio of the log (ratio) of estrogen receptor gene expression to age, lymph node positive); and ER⁺, ER/AGE low, LN⁻ (i.e., estrogen receptor positive with a low ratio of the log (ratio) of estrogen receptor gene expression to age, lymph node negative). The rationale for subdivision of the original patient set into these five subsets is detailed in the Examples (Section 6). The marker sets useful for each of the subsets above are provided in Tables 1-5, respectively.

[0109]

TABLE 1
Geneset of 20 markers used to classify ER−, sporadic individuals.
		Avg	Avg
Accession/		good	poor
Contig No.	Gene	xdev	xdev	Correlation	Description	Sp_xref_keyword_list	SEQ ID
AF055033	IGFBP5	−2.12	0.88	0.54	insulin-like	Growth factor binding,	11
					growth factor	Glycoprotein, Signal, 3D-
					binding protein 5	structure
NM_000599	IGFBP5	−3.41	0.43	0.53	insulin-like	Growth factor binding,	51
					growth factor	Glycoprotein, Signal, 3D-
					binding protein 5	structure
L27560	IGFBP5	−4.55	0	0.52	EST	Hypothetical protein	29
AF052162	FLJ12443	−0.27	1.6	0.52	EST	Hypothetical protein	9
NM_001456	FLNA	−0.61	2.47	0.52	filamin A, alpha	Hypothetical protein,	73
					(actin binding	Actin-binding,
					protein 280)	Phosphorylation, Repeat,
						Polymorphism, Disease
						mutation
NM_002205	ITGA5	−0.37	2.08	0.49	integrin, alpha 5	Integrin, Cell adhesion,	93
					(fibronectin	Receptor, Glycoprotein,
					receptor, alpha	Transmembrane, Signal,
					polypeptide)	Calcium, Repeat
NM_013261	PPARGC1	0.09	1.54	0.47	peroxisome proliferative activated receptor,	231
					gamma, coactivator 1
NM_001605	AARS	0.39	2.36	0.51	alanyl-tRNA	Aminoacyl-tRNA	77
					synthetase	synthetase, Protein
						biosynthesis, Ligase,
						ATP-binding
X87949	HSPA5	−0.03	2.03	0.49	heat shock	ATP-binding, Hypothetical	273
					70 kDa protein 5	protein, Endoplasmic
					(glucose-regulated	reticulum, Signal
					protein, 78 kDa)
Contig50950_RC	NGEF	−1.17	3.2	0.52	neuronal guanine nucleotide exchange	337
					factor
NM_005689	ABCB6	−0.51	2.26	0.48	ATP-binding	ATP-binding, Transport,	187
					cassette, sub-	Transmembrane,
					family B	Mitochondrion, Inner
					(MDR/TAP),	membrane, Transit
					member 6	peptide, Hypothetical
						protein
NM_004577	PSPH	−0.56	3.05	0.51	phosphoserine	Hydrolase, Serine	151
					phosphatase	biosynthesis, Magnesium,
						Phosphorylation
NM_003832	PSPHL	−2.08	2.18	0.5	phosphoserine		131
					phosphatase-like
NM_002422	MMP3	−0.96	2.54	0.5	matrix	Hydrolase,	101
					metalloproteinase	Metalloprotease,
					3 (stromelysin	Glycoprotein, Zinc,
					1, progelatinase)	Zymogen, Calcium,
						Collagen degradation,
						Extracellular matrix,
						Signal, Polymorphism,
						3D-structure
Contig37562_RC		−3.42	−6.02	−0.59	ESTs		293
NM_018465	MDS030	−0.82	−3.28	−0.58	uncharacterized	Hypothetical protein	267
					hematopoietic
					stem/progenitor
					cells protein
					MDS030
Contig54661_RC		−0.79	−2.08	−0.54	ESTs		349
AB032969	KIAA1143	−0.6	−2.85	−0.53	KIAA1143	Hypothetical protein	1
					protein
Contig55353_RC	KIAA1915	−0.27	−1.82	−0.47	KIAA1915	Hypothetical protein	353
					protein
NM_005213	CSTA	2.11	−3.4	−0.49	cystatin A (stefin	Thiol protease inhibitor,	175
					A)	3D-structure

[0110]

TABLE 2
Geneset of 10 markers used to classify ER−, BRCA1 individuals.
		Avg	Avg
Accession/		good	poor		Sequence
Contig No.	Gene	xdev	xdev	Correlation	name	Description	Sp_xref_keyword_list	SEQ ID
AF005487		6.08	0.5	−0.79	HLA-DRB6	Homo sapiens	MHC	3
						MHC class II
						antigen (DRB6)
						mRNA, HLA-
						DRB6*0201 allele,
						sequence.
Contig50728_RC		4.02	0.25	−0.77		ESTs, Weakly similar to S26650	333
						DNA-binding protein 5 - human
						[H. sapiens]
Contig53598_RC		8.41	3.26	−0.77	FLJ11413	hypothetical protein	Hypothetical	343
						FLJ11413	protein
NM_002888	RARRES1	6.9	0.05	−0.87	RARRES1	retinoic acid	Receptor,	109
						receptor responder	Transmembrane,
						(tazarotene	Signal-anchor
						induced) 1
NM_005218	DEFB1	5.14	−3.02	−0.81	DEFB1	defensin, beta 1	Antibiotic,	177
							Signal, 3D-structure
U17077	BENE	2.72	−1.72	−0.77	BENE	BENE protein	Transmembrane	271
Contig14683_RC		1.29	−2.31	−0.74		ESTs		279
Contig53641_RC		−3.29	4.23	0.75	MAGE-E1	MAGE-E1 protein	Hypothetical protein	345
Contig56678_RC		−6.7	−9.73	−0.82		ESTs, Highly similar to	357
						THYA_HUMAN Prothymosin alpha
						[H. sapiens]
NM_005461	KRML	0.88	−3.38	−0.75	MAFB	v-maf	Transcription	181
						musculoaponeurotic	regulation, Repressor,
						fibrosarcoma	DNA-binding,
						oncogene homolog	Nuclear protein,
						B (avian)	Hypothetical protein

[0111]

TABLE 3
Geneset of 50 markers used to classify ER+, ER/AGE high individuals.
		Avg	Avg
Accession/		good	poor
Contig No.	Gene	xdev	xdev	Correlation	Description	Sp_xref_keyword_list	SEQ ID
NM_003600	STK15	−2.93	2.08	0.8	serine/threonine	ATP-binding,	125
					kinase 6	Kinase,
						Serine/threonine-
						protein kinase,
						Transferase
NM_003158	STK6	−1.57	1.42	0.78	serine/threonine	ATP-binding,	113
					kinase 6	Kinase,
						Serine/threonine-
						protein kinase,
						Transferase
NM_007019	UBCH10	−2.98	2.62	0.81	ubiquitin-conjugating	Hypothetical protein,	217
					enzyme E2C	UbI conjugation
						pathway, Ligase,
						Multigene family,
						Mitosis, Cell cycle,
						Cell division
NM_013277	ID-GAP	−2.43	2.43	0.77	Rac GTPase	Hypothetical	233
					activating protein 1	protein
NM_004336	BUB1	−2.04	1.39	0.77	BUB1 budding	Transferase,	147
					uninhibited by	Serine/threonine-
					benzimidazoles 1	protein kinase,
					homolog (yeast)	ATP-binding, Cell
						cycle, Nuclear
						protein, Mitosis,
						Phosphorylation,
						Polymorphism
NM_006607	PTTG2	−1.71	1.49	0.72	pituitary tumor-		211
					transforming 2
AK001166	FLJ11252	−1.33	0.99	0.71	hypothetical protein	Hypothetical	13
					FLJ11252	protein
NM_004701	CCNB2	−4.62	2.01	0.81	cyclin B2	Cyclin, Cell cycle,	153
						Cell division, Mitosis
Contig57584_RC		−3.68	2.04	0.78	likely ortholog of mouse gene rich	359
					cluster, C8 gene
NM_006845	KNSL6	−4.13	1.05	0.73	kinesin-like 6 (mitotic	Hypothetical	215
					centromere-	protein, Motor protein,
					associated kinesin)	Microtubules, ATP-
						binding, Coiled coil,
						Nuclear protein
Contig38901_RC		−3.08	1.15	0.75	hypothetical protein	Hypothetical	299
					MGC45866	protein
NM_018410	DKFZp762E1312	−4.38	1.49	0.75	hypothetical protein	Hypothetical	263
					DKFZp762E1312	protein
NM_003981	PRC1	−3.52	2.17	0.78	protein regulator of		133
					cytokinesis 1
NM_001809	CENPA	−5.04	0.98	0.75	centromere protein	Hypothetical protein,	81
					A, 17 kDa	Chromosomal
						protein, Nuclear
						protein, DNA-binding,
						Centromere, Antigen
NM_003504	CDC45L	−2.67	1.22	0.73	CDC45 cell division	DNA replication,	123
					cycle 45-like (S. cerevisiae)	Cell cycle, Nuclear
						protein, Cell division
Contig41413_RC		−5.43	2.15	0.74	ribonucleotide	Oxidoreductase,	305
					reductase M2	DNA replication,
					polypeptide	Iron
NM_004217	STK12	−2.17	0.73	0.72	serine/threonine	Hypothetical	143
					kinase 12	protein, ATP-
						binding, Kinase,
						Serine/threonine-
						protein kinase,
						Transferase
NM_002358	MAD2L1	−2.65	2.27	0.83	MAD2 mitotic arrest	Cell cycle, Mitosis,	99
					deficient-like 1	Nuclear protein,
					(yeast)	3D-structure
NM_014321	ORC6L	−2.73	1.8	0.75	origin recognition	Hypothetical	241
					complex, subunit 6	protein, DNA
					homolog-like (yeast)	replication, Nuclear
						protein, DNA-binding
NM_012291	KIAA0165	−1.52	1.55	0.71	extra spindle poles	Hypothetical	229
					like 1 (S. cerevisiae)	protein
NM_004203	PKMYT1	−3.64	2.2	0.7	retinoblastoma-like 2	ATP-binding,	137
					(p130)	Kinase,
						Serine/threonine-
						protein kinase,
						Transferase,
						Transcription
						regulation, DNA-
						binding, Nuclear
						protein, Cell cycle,
						Phosphorylation,
						Anti-oncogene
M96577	E2F1	−2.14	1.42	0.75	E2F transcription	Transcription	33
					factor 1	regulation,
						Activator, DNA-
						binding, Nuclear
						protein,
						Phosphorylation,
						Cell cycle, Apoptosis,
						Polymorphism
NM_002266	KPNA2	−3.77	1.78	0.71	karyopherin alpha 2	Transport, Protein	95
					(RAG cohort 1,	transport, Repeat,
					importin alpha 1)	Nuclear protein,
						Polymorphism
Contig31288_RC		−2.63	0.7	0.68	ESTs, Weakly similar to hypothetical	289
					protein FLJ20489 [Homo sapiens]
					[H. sapiens]
NM_014501	E2-EPF	−1.55	1.93	0.7	ubiquitin carrier	UbI conjugation	247
					protein	pathway, Ligase,
						Multigene family
NM_001168	BIRC5	−5.76	2.01	0.78	baculoviral IAP	Apoptosis, Thiol	63
					repeat-containing 5	protease inhibitor,
					(survivin)	Alternative splicing,
						3D-structure,
						Hypothetical
						protein, Protease,
						Receptor
NM_003258	TK1	−4.57	1.38	0.71	thymidine kinase 1,	Transferase,	115
					soluble	Kinase, DNA synthesis,
						ATP-binding
NM_001254	CDC6	−2.46	0.28	0.72	CDC6 cell division	ATP-binding, Cell	67
					cycle 6 homolog	division
					(S. cerevisiae)
NM_004900	DJ742C19.2	−2.96	0.13	0.69	apolipoprotein B	Hydrolase	161
					mRNA editing
					enzyme, catalytic
					polypeptide-like 3B
NM_004702	CCNE2	−3.12	2.13	0.81	cyclin E2	Cell cycle, Cell	155
						division, Cyclin,
						Hypothetical protein,
						Phosphorylation,
						Alternative splicing,
						Nuclear protein
AL160131		−3.07	2.42	0.7	hypothetical protein	Hypothetical	21
					MGC861	protein
NM_016359	LOC51203	−3.22	2.61	0.76	nucleolar protein	Hypothetical protein,	253
					ANKT	Nuclear protein
NM_004856	KNSL5	−1.52	1.1	0.71	kinesin-like 5 (mitotic	Motor protein, Cell	159
					kinesin-like protein	division,
					1)	Microtubules, ATP-
						binding, Coiled coil,
						Mitosis, Cell cycle,
						Nuclear protein
NM_000057	BLM	−1.54	0.76	0.71	Bloom syndrome	Hydrolase,	35
						Helicase, ATP-
						binding, DNA-
						binding, Nuclear
						protein, DNA
						replication,
						Disease mutation
NM_018455	BM039	−2.44	1.18	0.7	uncharacterized bone		265
					marrow protein BM039
NM_002106	H2AFZ	−2.49	1.53	0.72	H2A histone family,	Chromosomal	91
					member Z	protein,
						Nucleosome core,
						Nuclear protein,
						DNA-binding,
						Multigene family
Contig64688		−2.68	3.1	0.73	hypothetical protein	Hypothetical	365
					FLJ23468	protein
Contig44289_RC		−1.65	1.6	0.67	ESTs		315
Contig28552_RC		−1.37	1.53	0.68	diaphanous homolog	Hypothetical	281
					3 (Drosophila)	protein, Coiled coil,
						Repeat, Alternative
						splicing
Contig46218_RC		−1.31	1.56	0.68	ESTs, Weakly similar to T19201	321
					hypothetical protein C11G6.3 -
					Caenorhabditis elegans [C. elegans]
Contig28947_RC		−1.3	0.98	0.67	cell division cycle	Hypothetical protein,	283
					25A	Cell division,
						Mitosis, Hydrolase,
						Alternative splicing,
						Multigene family,
						3D-structure
NM_016095	LOC51659	−1.4	2.13	0.67	HSPC037 protein	Hypothetical protein	249
NM_003090	SNRPA1	−3.26	0.95	0.7	small nuclear	Hypothetical	111
					ribonucleoprotein	protein, Nuclear
					polypeptide A′	protein, RNA-binding,
						Ribonucleoprotein,
						Leucine-rich repeat,
						Repeat, 3D-structure
NM_002811	PSMD7	−2.48	1.89	0.7	proteasome	Proteasome	107
					(prosome, macropain) 26S
					subunit, non-ATPase,
					7 (Mov34 homolog)
Contig38288_RC		−2.34	0.97	0.67	hypothetical protein	Hypothetical	297
					DKFZp762A2013	protein
NM_003406	YWHAZ	−1.5	2.79	0.68	tyrosine 3-	Brain, Neurone,	121
					monooxygenase/tryptophan	Phosphorylation,
					5-monooxygenase	Acetylation,
					activation protein,	Multigene family,
					zeta polypeptide	3D-structure
AL137540	NTN4	2.13	−4.61	−0.69	netrin 4	Hypothetical	19
						protein, Laminin
						EGF-like domain,
						Signal
AL049367		1.9	−3.2	−0.68	EST	Transducer,	15
						Prenylation,
						Lipoprotein,
						Multigene family,
						Acetylation
NM_013409	FST	1.04	−5.78	−0.69	follistatin	Glycoprotein,	235
						Repeat, Signal,
						Alternative splicing
NM_000060	BTD	3.1	−1.45	−0.67	biotinidase	Hydrolase,	37
						Glycoprotein,
						Signal, Disease
						mutation

[0112]

TABLE 4
Geneset of 50 markers used to classify ER+, ER/AGE low, LN+ individuals.
		Avg	Avg
Accession/		good	poor
Contig No.	Gene	xdev	xdev	Correlation	Description	Sp_xref_keyword_list	SEQ ID
NM_006417	MTAP44	−1.5	3	0.69	Fc fragment of IgG,	Hydrolase,	205
					low affinity IIb,	Hypothetical protein,
					receptor for (CD32)	Immunoglobulin
						domain, IgG-
						binding protein,
						Receptor,
						Transmembrane,
						Glycoprotein,
						Signal, Repeat,
						Multigene family,
						Polymorphism,
						NAD, One-carbon
						metabolism, Serine
						protease, Zymogen,
						Protease,
						Alternative splicing,
						Chromosomal
						translocation,
						Proto-oncogene,
						Galaptin, Lectin,
						Antigen
NM_006820	GS3686	−4.3	4.06	0.69	chromosome 1 open	Hypothetical	213
					reading frame 29	protein
NM_001548	IFIT1	−3.4	4.27	0.71	Interferon-induced	Repeat, TPR	75
					protein with	repeat, Interferon
					tetratricopeptide	induction
					repeats 1
Contig41538_RC		−2.5	3.16	0.68	ESTs, Moderately similar to hypothetical	307
					protein FLJ20489 [Homo sapiens]
NM_016816	OAS1	−1.7	3.29	0.75	2′,5′-oligoadenylate	RNA-binding,	255
					synthetase 1,	Transferase,
					40/46 kDa	Nucleotidyltransferase,
						Interferon induction,
						Alternative splicing
Contig51660_RC		−2.1	2.65	0.66	28 kD interferon	Transmembrane	339
					responsive protein
Contig43645_RC		−4.8	1.44	0.63	Homo sapiens,	Hypothetical	313
					clone	protein
					IMAGE: 4428577,
					mRNA, partial cds
AF026941		−4.6	2.71	0.63	EST, Weakly similar	Hypothetical	5
					to 2004399A	protein
					chromosomal
					protein [Homo
					sapiens]
NM_007315	STAT1	−3.5	1.8	0.59	signal transducer	Transcription	225
					and activator of	regulation, DNA-
					transcription 1,	binding, Nuclear
					91 kDa	protein,
						Phosphorylation,
						SH2 domain,
						Alternative splicing,
						3D-structure
NM_002038	G1P3	−4.1	5.64	0.79	interferon, alpha-	Interferon	85
					inducible protein	induction,
					(clone IFI-6-16)	Transmembrane,
						Signal, Alternative
						splicing
NM_005101	ISG15	−5.6	5.34	0.77	interferon-stimulated	Interferon	169
					protein, 15 kDa	induction, Repeat
NM_002462	MX1	−6.1	0.83	0.56	myxovirus (influenza	Hypothetical	103
					virus) resistance 1,	protein, Interferon
					interferon-inducible	induction, GTP-
					protein p78 (mouse)	binding, Multigene
						family, Antiviral
NM_005532	IFI27	−5.8	2.81	0.59	interferon, alpha-	Interferon	183
					inducible protein 27	induction,
						Transmembrane
NM_002346	LY6E	−2.1	3.58	0.75	lymphocyte antigen	Signal, Antigen,	97
					6 complex, locus E	Multigene family,
						Membrane, GPI-anchor
NM_016817	OAS2	−3.6	1.89	0.59	2′-5′-oligoadenylate	RNA-binding,	257
					synthetase 2,	Transferase,
					69/71 kDa	Nucleotidyltransferase,
						Repeat, Interferon
						induction,
						Alternative splicing,
						Myristate
Contig44909_RC		−2.3	1.13	0.55	hypothetical protein	Hypothetical	317
					BC012330	protein
NM_017414	USP18	−4.1	3.37	0.72	ubiquitin specific	Ubl conjugation	259
					protease 18	pathway, Hydrolase,
						Thiol protease,
						Multigene family
NM_004029	IRF7	−2.4	3.67	0.66	interferon regulatory	Collagen,	135
					factor 7	Transcription
						regulation, DNA-
						binding, Nuclear
						protein, Activator,
						Alternative splicing
NM_004335	BST2	−3.2	3.22	0.57	bone marrow	Transmembrane,	145
					stromal cell antigen 2	Glycoprotein,
						Signal-anchor,
						Polymorphism
NM_002759	PRKR	−2.4	1.8	0.58	protein kinase,	Transferase,	105
					interferon-inducible	Serine/threonine-
					double stranded	protein kinase,
					RNA dependent	ATP-binding, Repeat,
						Phosphorylation,
						Interferon
						induction, RNA-
						binding, 3D-structure
NM_006332	IFI30	−3.8	2.65	0.64	interferon, gamma-	Oxidoreductase,	203
					inducible protein 30	Interferon induction,
						Glycoprotein,
						Lysosome, Signal,
						Hypothetical protein
NM_009587	LGALS9	−3.2	2.08	0.6	lectin, galactoside-	Galaptin, Lectin,	227
					binding, soluble, 9	Repeat, Alternative
					(galectin 9)	splicing
NM_003641	IFITM1	−2.4	5.54	0.63	interferon induced	Interferon	127
					transmembrane	induction,
					protein 1 (9-27)	Transmembrane
NM_017523	HSXIAPAF1	−1	2.84	0.7	XIAP associated	Hypothetical	261
					factor-1	protein
NM_014314	RIG-I	−1.3	3.55	0.62	RNA helicase	ATP-binding,	239
						Helicase, Hydrolase,
						Hypothetical protein
Contig47563_RC		−2.2	3.11	0.56	ESTs		325
AI497657_RC		−4.4	5.61	0.74	guanine nucleotide	Transducer,	335
					binding protein 4	Prenylation,
						Lipoprotein,
						Multigene family
NM_000735	CGA	−4.3	2.5	0.58	glycoprotein	Hormone,	53
					hormones, alpha	Glycoprotein,
					polypeptide	Signal, 3D-structure
NM_004988	MAGEA1	−1.4	6.31	0.64	melanoma antigen,	Antigen, Multigene	163
					family A, 1 (directs	family,
					expression of	Polymorphism,
					antigen MZ2-E)	Tumor antigen
Contig54242_RC		−1.2	4.1	0.65	chromosome 17	Hypothetical	347
					open reading frame	protein
					26
NM_004710	SYNGR2	−1.4	3.01	0.54	synaptogyrin 2	Transmembrane	157
NM_001168	BIRC5	−3.7	3.39	0.64	baculoviral IAP	Hypothetical	63
					repeat-containing 5	protein, Protease,
					(survivin)	Receptor,
						Apoptosis, Thiol
						protease inhibitor,
						Alternative splicing,
						3D-structure
Contig41413_RC		−4.4	2.61	0.57	ribonucleotide	Oxidoreductase,	305
					reductase M2	DNA replication,
					polypeptide	Iron
NM_004203	PKMYT1	−3.4	3.79	0.6	retinoblastoma-like 2	ATP-binding,	137
					(p130)	Kinase,
						Serine/threonine-
						protein kinase,
						Transferase,
						Transcription
						regulation, DNA-
						binding, Nuclear
						protein, Cell cycle,
						Phosphorylation,
						Anti-oncogene
Contig48913_RC		−3.1	1.72	0.55	Homo sapiens, Similar to hypothetical	327
					protein PRO1722, clone MGC: 15692
					IMAGE: 3351479, mRNA, complete cds
NM_005804	DDXL	−2.5	1.42	0.58	DEAD/H (Asp-Glu-	ATP-binding,	191
					Ala-Asp/His) box	Helicase, Hydrolase,
					polypeptide 39	Hypothetical protein
NM_016359	LOC51203	−1.7	3.6	0.57	nucleolar protein	Hypothetical	253
					ANKT	protein, Nuclear
						protein
NM_001645	APOC1	−2.9	3.43	0.58	apolipoprotein C-I	Plasma, Lipid	79
						transport, VLDL,
						Signal, 3D-structure,
						Polymorphism
Contig37895_RC		−2	2.05	0.55	ESTs		295
NM_005749	TOB1	−1.3	4.96	0.59	transducer of	Phosphorylation	189
					ERBB2, 1
NM_000269	NME1	−1.3	2.98	0.55	non-metastatic cells	Transferase,	39
					1, protein (NM23A)	Kinase, ATP-
					expressed in	binding, Nuclear
						protein, Anti-
						oncogene, Disease
						mutation
NM_014462	LSM1	−1	4.5	0.57	Lsm1 protein	Nuclear protein,	245
						Ribonucleoprotein,
						mRNA splicing,
						mRNA processing,
						RNA-binding
Contig31221_RC		−1.4	3.83	0.56	HTPAP protein		287
NM_005326	HAGH	−1.9	4.29	0.57	hydroxyacyl	Hydrolase, Zinc,	179
					glutathione	3D-structure
					hydrolase
Contig42342_RC		0.78	−3.2	−0.6	Homo sapiens cDNA	Hypothetical	311
					FLJ39417 fis, clone	protein
					PLACE6016942
AL137540	NTN4	2.24	−3.9	−0.6	netrin 4	Laminin EGF-like	19
						domain, Signal,
						Hypothetical protein
Contig40434_RC		1.64	−5.6	−0.6	wingless-type MMTV	Developmental	301
					integration site	protein,
					family, member 5A	Glycoprotein, Signal
Contig1632_RC		1.03	−3.9	−0.6	hypothetical protein	Hypothetical	275
					MGC17921	protein
NM_014246	CELSR1	0.95	−4.6	−0.6	cadherin, EGF LAG	G-protein coupled	237
					seven-pass G-type	receptor,
					receptor 1 (flamingo	Transmembrane,
					homolog,	Glycoprotein, EGF-
					Drosophila)	like domain,
						Calcium-binding,
						Laminin EGF-like
						domain, Repeat,
						Developmental
						protein,
						Hydroxylation,
						Signal, Alternative
						splicing,
						Hypothetical protein
NM_005139	ANXA3	1.26	−6.2	−0.6	annexin A3	Annexin,	171
						Calcium/phospholipid-
						binding, Repeat,
						Phospholipase A2
						inhibitor, 3D-
						structure,
						Polymorphism

[0113]

TABLE 5
Geneset of 65 markers used to classify ER+, ER/AGE low, LN− individuals.
		Avg	Avg
Accession/		good	poor		Sequence
Contig No.	Gene	xdev	xdev	Correlation	name	Description	Sp_xref_keyword_list	SEQ ID
M55914	MPB1	−2.82	1.25	0.5	ENO1	enolase 1, (alpha)	DNA-binding,	31
							Transcription
							regulation,
							Repressor,
							Nuclear protein,
							Lyase, Glycolysis,
							Magnesium,
							Multigene family,
							Hypothetical
							protein
NM_005945	MPB1	−3.06	1.19	0.49	ENO1	Homo sapiens	Glycolysis,	193
						enolase 1, (alpha)	Hypothetical
						(ENO1), mRNA.	protein, Lyase,
							Magnesium,
							DNA-binding,
							Transcription
							regulation,
							Repressor,
							Nuclear protein,
							Multigene family
NM_001428	ENO1	−2.53	1.18	0.46	ENO1	enolase 1, (alpha)	DNA-binding,	71
							Transcription
							regulation,
							Repressor,
							Nuclear protein,
							Lyase, Glycolysis,
							Magnesium,
							Multigene family,
							Hypothetical
							protein
NM_001216	CA9	−4.72	1.49	0.6	CA9	carbonic	Lyase, Zinc,	65
						anhydrase IX	Transmembrane,
							Glycoprotein,
							Antigen, Signal,
							Nuclear protein,
							Polymorphism
NM_001124	ADM	−5.68	2.99	0.56	ADM	Adrenomedullin	Hormone,	61
							Amidation,
							Cleavage on
							pair of basic
							residues, Signal
NM_000584	IL8	−2.45	2.04	0.54	IL8	interleukin 8	Cytokine,	49
							Chemotaxis,
							Inflammatory
							response, Signal,
							Alternative
							splicing, 3D-structure
D25328	PFKP	−4.19	3.29	0.56	PFKP	Phosphofructo-	Kinase,	25
						kinase, platelet	Transferase,
							Glycolysis,
							Repeat, Allosteric
							enzyme,
							Phosphorylation,
							Magnesium,
							Multigene family
NM_006096	NDRG1	−5.45	5.97	0.77	NDRG1	N-myc downstream	Hypothetical	199
						regulated gene 1	protein, Nuclear
							protein, Repeat
NM_004994	MMP9	−5.53	1.07	0.49	MMP9	matrix	Hydrolase,	165
						metalloproteinase	Metalloprotease,
						9 (gelatinase B,	Glycoprotein,
						92 kDa gelatinase,	Zinc, Zymogen,
						92 kDa type IV	Calcium, Collagen
						collagenase)	degradation,
							Extracellular matrix,
							Repeat, Signal,
							Polymorphism,
							3D-structure
NM_003311	TSSC3	−4.57	5.58	0.68	TSSC3	tumor suppressing subtransferable	117
						candidate 3
NM_006086	TUBB4	−5.19	2.85	0.59	TUBB4	tubulin, beta, 4	G-protein coupled	197
							receptor,
							Transmembrane,
							Glycoprotein,
							Phosphorylation,
							Lipoprotein,
							Palmitate,
							Polymorphism,
							Hypothetical
							protein, GTP-
							binding, Receptor,
							Microtubules,
							Multigene family
NM_006115	PRAME	−4.48	2.77	0.61	PRAME	preferentially	Antigen	201
						expressed antigen
						in melanoma
NM_004345	CAMP	−2.02	1.37	0.49	CAMP	cathelicidin	Antibiotic, Signal	149
						antimicrobial
						peptide
NM_018455	BM039	−2.34	0.76	0.47	BM039	uncharacterized		265
						bone marrow
						protein BM039
Contig49169_RC		−1.17	1.5	0.46	SUV39H2	suppressor of	Hypothetical	329
						variegation 3-9	protein, Nuclear
						(Drosophila)	protein
						homolog 2;
						hypothetical protein
						FLJ23414
Contig45032_RC		−1.37	0.77	0.45	FLJ14813	hypothetical protein	Hypothetical	319
						FLJ14813	protein, ATP-
							binding, Kinase,
							Serine/threonine-
							protein kinase,
							Transferase
NM_000917	P4HA1	−1.54	4.31	0.62	P4HA1	procollagen-	Dioxygenase,	57
						proline, 2-	Collagen,
						oxoglutarate 4-	Oxidoreductase,
						dioxygenase	Iron, Vitamin C,
						(proline 4-	Alternative splicing,
						hydroxylase), alpha	Glycoprotein,
						polypeptide I	Endoplasmic
							reticulum, Signal
NM_002046	GAPD	−2.51	3.42	0.6	GAPD	glyceraldehyde-3-	Glycolysis, NAD,	87
						phosphate	Oxidoreductase,
						dehydrogenase	Hypothetical protein,
							Multigene family
NM_000365	TPI1	−1.81	2.94	0.56	TPI1	triosephosphate	Fatty acid	45
						isomerase 1	biosynthesis,
							Gluconeogenesis,
							Glycolysis,
							Isomerase,
							Pentose shunt,
							Disease mutation,
							Polymorphism,
							3D-structure
NM_014364	GAPDS	−1.08	2.88	0.58	GAPDS	glyceraldehyde-3-	Glycolysis,	243
						phosphate	Oxidoreductase,
						dehydrogenase,	NAD
						testis-specific
NM_005566	LDHA	−2.01	4.01	0.59	LDHA	lactate	Oxidoreductase,	185
						dehydrogenase A	NAD, Glycolysis,
							Multigene family,
							Disease mutation,
							Polymorphism
NM_000291	PGK1	−2.28	1.68	0.51	PGK1	phosphoglycerate	Kinase, Transferase,	41
						kinase 1	Multigene family,
							Glycolysis,
							Acetylation,
							Disease mutation,
							Polymorphism,
							Hereditary
							hemolytic anemia
NM_016185	LOC51155	−2.33	2.82	0.59	HN1	hematological and		251
						neurological
						expressed 1
NM_001168	BIRC5	−4.33	2.78	0.55	BIRC5	baculoviral IAP	Apoptosis, Thiol	63
						repeat-containing 5	protease inhibitor,
						(survivin)	Alternative splicing,
							3D-structure,
							Hypothetical
							protein, Protease,
							Receptor
NM_002266	KPNA2	−3.75	1.34	0.47	KPNA2	karyopherin alpha	Transport,	95
						2 (RAG cohort 1,	Protein transport,
						importin alpha 1)	Repeat, Nuclear
							protein,
							Polymorphism
Contig31288_RC		−2.1	1.27	0.5		ESTs, Weakly similar to hypothetical	289
						protein FLJ20489 [Homo sapiens]
						[H. sapiens]
NM_000269	NME1	−2.15	3.43	0.55	NME1	non-metastatic	Transferase,	39
						cells 1, protein	Kinase, ATP-
						(NM23A)	binding, Nuclear
						expressed in	protein, Anti-
							oncogene,
							Disease mutation
NM_003158	STK6	−1.23	1.73	0.45	STK6	serine/threonine	ATP-binding,	113
						kinase 6	Kinase,
							Serine/threonine-
							protein kinase,
							Transferase
NM_007274	HBACH	−1.83	2.73	0.51	BACH	brain acyl-CoA	Hydrolase, Serine	223
						hydrolase	esterase, Repeat
Contig55188_RC		−2.36	3.28	0.47	FLJ22341	hypothetical protein	Hypothetical	351
						FLJ22341	protein
NM_002061	GCLM	−1.06	1.76	0.48	GCLM	glutamate-cysteine	Ligase,	89
						ligase, modifier	Glutathione
						subunit	biosynthesis
NM_004207	SLC16A3	−3.11	5.07	0.67	SLC16A3	solute carrier family	Transport,	139
						16 (monocarboxylic	Symport,
						acid transporters),	Transmembrane,
						member 3	Multigene family
NM_000582	SPP1	−5.09	5.47	0.53	SPP1	secreted	Hypothetical	47
						phosphoprotein 1	protein,
						(osteopontin, bone	Glycoprotein,
						sialoprotein I, early	Sialic acid,
						T-lymphocyte	Biomineralization,
						activation 1)	Cell adhesion,
							Phosphorylation,
							Signal, Alternative
							splicing
NM_001109	ADAM8	−2.5	3.74	0.45	ADAM8	a disintegrin and	Hydrolase,	59
						metalloproteinase	Metalloprotease,
						domain 8	Zinc, Signal,
							Glycoprotein,
							Transmembrane,
							Antigen
D50402	SLC11A1	−1.05	3.46	0.53	SLC11A1	solute carrier	Transport, Iron	27
						family	transport,
						11 (proton-coupled	Transmembrane,
						divalent metal ion	Glycoprotein,
						transporters),	Macrophage,
						member 1	Polymorphism
AL080235	DKFZP586E1621	−1.23	1.96	0.51	RIS1	Ras-induced	Hypothetical	17
						senescence 1	protein
Contig40552_RC		−1.26	3.96	0.54	FLJ25348	hypothetical protein	Hypothetical	303
						FLJ25348	protein
Contig52490_RC		−0.64	3.33	0.61	LOC116238	hypothetical protein		341
						BC014072
NM_006461	DEEPEST	−2.1	1.85	0.46	SPAG5	sperm associated	Hypothetical	207
						antigen 5	protein
Contig56503_RC		−4.3	3.39	0.55	MGC9753	hypothetical gene	Hypothetical	355
						MGC9753	protein
Contig63525		−1.91	3.34	0.5	FLJ13352	hypothetical protein	Hypothetical	363
						FLJ13352	protein
NM_001909	CTSD	−0.83	4.6	0.51	CTSD	cathepsin D	Hydrolase, Aspartyl	83
						(lysosomal aspartyl	protease,
						protease)	Glycoprotein,
							Lysosome, Signal,
							Zymogen,
							Polymorphism,
							Alzheimer's disease,
							3D-structure
NM_005063	SCD	−2.57	5.15	0.48	SCD	stearoyl-CoA	Hypothetical protein,	167
						desaturase (delta-	Endoplasmic
						9-desaturase)	reticulum, Fatty acid
							biosynthesis, Iron,
							Oxidoreductase,
							Transmembrane
NM_005165	ALDOC	−2.43	5.02	0.48	ALDOC	aldolase C,	Lyase, Schiff base,	173
						fructose-	Glycolysis,
						bisphosphate	Multigene family
NM_000363	TNNI3	−0.54	3.58	0.48	TNNI3	troponin I, cardiac	Hypothetical	43
							protein, Muscle
							protein, Actin-
							binding,
							Acetylation,
							Disease mutation,
							Cardiomyopathy,
							Receptor, Signal
AF035284		−1.63	3.28	0.47	FADS1	EST	Heme,	7
							Hypothetical protein
Contig30875_RC		−0.88	3	0.6		ESTs		285
NM_018487	HCA112	−0.7	3.54	0.58	HCA112	hepatocellular	Hypothetical	269
						carcinoma-	protein
						associated antigen
						112
NM_001323	CST6	−1.63	3.84	0.57	CST6	cystatin E/M	Thiol protease	69
							inhibitor, Signal,
							Glycoprotein
NM_006516	SLC2A1	−1.66	2.22	0.46	SLC2A1	solute carrier	Transmembrane,	209
						family	Sugar transport,
						2 (facilitated	Transport,
						glucose	Glycoprotein,
						transporter),	Multigene family,
						member 1	Disease mutation
NM_007267	LAK-4P	−1.04	3.28	0.61	EVIN1	expressed in	Hypothetical	221
						activated T/LAK	protein
						lymphocytes
NM_004710	SYNGR2	−0.84	4.81	0.56	SYNGR2	synaptogyrin 2	Transmembrane	157
Contig63649_RC		−1.34	6.3	0.75		ESTs, Weakly similar to 2004399A	361
						chromosomal protein [Homo
						sapiens] [H. sapiens]
NM_003376	VEGF	−2.12	2.42	0.46	VEGF	vascular	Hypothetical	119
						endothelial growth	protein, Mitogen,
						factor	Angiogenesis,
							Growth factor,
							Glycoprotein,
							Signal, Heparin-
							binding, Alternative
							splicing, Multigene
							family, 3D-structure
NM_000799	EPO	−0.75	4.01	0.69	EPO	erythropoietin	Erythrocyte	55
							maturation,
							Glycoprotein,
							Hormone, Signal,
							Pharmaceutical,
							3D-structure
NM_006014	DXS9879E	−1.85	3.44	0.54	DXS9879E	DNA segment on chromosome X	195
						(unique) 9879 expressed sequence
NM_007183	PKP3	−0.91	4.14	0.48	PKP3	plakophilin 3	Cell adhesion,	219
							Cytoskeleton,
							Structural
							protein, Nuclear
							protein, Repeat
D13642	SF3B3	−0.65	2.28	0.48	SF3B3	splicing factor 3b,	Hypothetical	23
						subunit 3, 130 kDa	protein,
							Spliceosome,
							mRNA processing,
							mRNA splicing,
							Nuclear protein
NM_003756	EIF3S3	−1.85	2.19	0.46	EIF3S3	eukaryotic	Initiation factor,	129
						translation initiation	Protein
						factor 3, subunit 3	biosynthesis
						gamma, 40 kDa
Contig47096_RC		−0.41	4.52	0.54	PFKFB4	6-phosphofructo-2-	Kinase,	323
						kinase/fructose-	Multifunctional
						2,6-biphosphatase 4	enzyme, Transferase,
							Hydrolase, ATP-
							binding,
							Phosphorylation,
							Multigene family
NM_004209	SYNGR3	−0.31	3.67	0.53	SYNGR3	synaptogyrin 3	Transmembrane	141
Contig3464_RC		0.99	−5.81	−0.52		ESTs		277
Contig31646_RC		1.1	−7.76	−0.5	COL14A1	collagen, type XIV,	Extracellular matrix,	291
						alpha 1 (undulin)	Glycoprotein,
							Hypothetical protein,
							Collagen, Signal
Contig49388_RC		1.73	−1.75	−0.51	FLJ13322	hypothetical protein	Hypothetical	331
						FLJ13322	protein
Contig41887_RC		0.37	−5.74	−0.47	LOC124220	similar to common	Hypothetical	309
						salivary protein 1	protein