MULTITAG SEQUENCING ECOGENOMICS ANALYSIS-US

[0001] This application is a continuation of application Ser. No. 12/515,262 filed May 15, 2009 (now U.S. Pat. No. 8,603,749), which is a National Stage of PCT/US07/84840, filed Nov. 15, 2007, which claims full benefit of priority of U.S. provisional application No. 60/858,948, filed on 15 Nov. 2006, each of which is hereby incorporated by reference in its entirety.

[0002] Work described herein was done partly with Government support under Grant No. 1R43DK074275-01A2 awarded by the U.S. National Institute of Diabetes and Digestive and Kidney Diseases, and the US Government therefore may have certain rights in the invention.

[0003] The invention relates to the determination of polynucleotide sequences. It also relates to determining sequences in multiple samples, in some particulars in multiple environmental samples and in multiple clinical samples.

[0004] Sequence determination technologies for proteins, RNAs and DNAs, have been pivotal in the development of modern molecular biology. During the past fifteen years, DNA sequencing in particular has been the core technology in an on-going revolution in the scope and the depth of understanding of genomic organization and function. The on-going development of sequencing technology is, perhaps, best symbolized by the determination of the complete sequence of a human genome.

[0005] The human genome sequencing project served a number of purposes. It served as a platform for programmatic development of improved sequencing technologies and of genome sequencing efforts. It also served to provide a framework for the production and distribution of sequencing information from increasingly large scale sequencing projects. These projects provided complete genome sequences for a succession of model organisms of increasingly large genetic complements. These accomplishments, culminating in the completion of a human genome sequence, highlight the very considerable power and throughput of contemporary sequencing technology.

[0006] At the same time, however, they highlight the limitations of current technology and the need for considerable improvements in speed, accuracy, and cost before sequencing can be fully exploited in research and medicine. Among the areas that can be seen most readily to require advances in sequencing technology are clinical sequencing applications that require whole genome information, environmental applications involving multiple organisms in mixtures, and applications that require processing of many samples. These are, of course, just a few among a great many areas that either require or will benefit greatly from more capable and less expensive sequencing methods.

[0007] To date, virtually all sequencing has been done by Sanger chain elongation methods. All Sanger methods require separating the elongation products with single base resolution. Currently, while PAGE still is used for this purpose in some commercial sequencers, capillary electrophoresis is the method of choice for high throughput DNA sequencers. Both gel-based and capillary-based separation methods are time consuming, costly, and limit throughput. Chip based methods, such as Affymetrix GeneChips and HySeq's sequencing by hybridization methods, require chips that can be produced only by capital intensive and complex manufacturing processes. These limitations pose obstacles to the utilization of sequencing for many purposes, such as those described above. Partly to overcome the limitations imposed by the necessity for powerful separation techniques in chain termination sequencing methods and the manufacturing requirements of chip-based methods, a number of technologies are currently being developed that do not require the separation of elongation products with integer resolution and do not require chips.

[0008] A lead technology of this type is a bead, emulsion amplification, and pyrosequencing-based method developed by 454 Life Sciences. (See Marguilles, et al. (2005) Nature 437: 376, which is incorporated herein by reference in its entirety, particularly as to the afore-mentioned methods. The method utilizes a series of steps to deposit single, amplified DNA molecules in individual wells of a plate containing several million picoliter wells. The steps ensure that each well of the plate either contains no DNA or the amplified DNA from a single original molecule. Pyrosequencing is carried out in the wells by elongation of a primer template in much the same way as Sanger sequencing. Pyrosequencing does not involve chain termination and does not require separation of elongation products. Instead sequencing proceeds stepwise by single base addition cycles. In each cycle one of the four bases—A, T, G, or C—is included in the elongation reaction. The other three bases are omitted. A base is added to the growing chain if it is complementary to the next position on the template. Light is produced whenever a base is incorporated into the growing complimentary sequence. By interrogating with each of A, C, G, or T in succession, the identity of the base at each position can be determined. Sequencing reactions are carried out in many wells simultaneously. Signals are collected from all the wells at once using an imaging detector. Thus, a multitude of sequences can be determined at the same time.

[0009] In principle, each well containing a DNA will emit a signal for only one of the four bases for each position. In practice, runs of the same base at two or more positions in succession lead to the emission of proportionally stronger signals for the first position in the run. Consequently, reading out the sequence from a given well is a bit more complicated then simply noting, for each position, which of the four bases is added. Nevertheless, because signals are proportional to the number of incorporations, sequences can be accurately reconstructed from the signal strength for most runs.

[0010] The technology has been shown to read accurately an average of about 250 or so bases per well with acceptable accuracy. A device offered by 454 Life Sciences currently uses a 6.4 cm² picoliter well “plate” containing 1,600,000 picoliter sized wells for sequencing about 400,000 different templates. The throughput for a single run using this plate currently is about 100 million bases in four hours. Even though this is a first generation device, its throughput is nearly 100 times better than standard Sanger sequencing devices.

[0011] Numerous other methods are being developed for ultra high throughput sequencing by other institutions and companies. Sequencing by synthesis methods that rely on target amplification are being developed and/or commercialized by George Church at Harvard University, by Solexa, and by others. Ligation sequencing methods have been developed and/or are being commercialized by Applied Biosystems and Solexa, among others. Array and hybridization sequencing methods are commercially available and/or are being developed by Affymetrix, Hyseq, Biotrove, Nimblegen, Illumina, and others. Methods of sequencing single molecules are being pursued by Helicos based on sequencing by synthesis and U.S. Genomics (among others) based on poration.

[0012] These methods represent a considerable improvement in throughput over past methods, in some regards. And they promise considerable improvement in economy as well. However, currently they are expensive to implement and use, they are limited to relatively short reads and, although massively parallel, they have limitations that must be overcome to realize their full potential.

[0013] One particular disadvantage of these methods, for example, is that samples must be processed serially, reducing throughput and increasing cost. This is a particularly great disadvantage when large numbers of samples are being processed, such as may be the case in clinical studies and environmental sampling, to name just two applications.

[0014] The incorporation of indexing sequences by ligation to random shotgun libraries has been disclosed in U.S. Pat. Nos. 7,264,929, 7,244,559, and 7211390, but the direct ligation methods therein disclosed distort the distribution of the components within the samples (as illustrated in FIG. 4 herein) and therefore are inappropriate for enumerating components within each sample.

[0015] Accordingly, there is a need to improve sample throughput, to lower the costs of sequencing polynucleotides from many samples at a time, and to accurately enumerate the components of samples analyzed by high throughput, parallelized and multiplex techniques.

[0016] It is therefore an object of the present invention to provide sequencing methods with improved sample throughput. The following paragraphs describe a few illustrative embodiments of the invention that exemplify some of its aspects and features. They are not exhaustive in illustrating its many aspects and embodiments, and thus are not in any way limitative of the invention. Many other aspects, features, and embodiments of the invention are described herein. Many other aspects and embodiments will be readily apparent to those skilled in the art upon reading the application and giving it due consideration in the full light of the prior art and knowledge in the field.

[0017] Embodiments provide multiplex methods for the quantitative determination of polynucleotides in two or more samples, comprising:

[0018] hybridizing a first primer to polynucleotides in a first sample, said first primer comprising a first tag sequence and a first probe sequence specific for a first target sequence, wherein said first target sequence is 3′ to a variable genetic region;

[0019] elongating primer templates formed thereby to form a first population of tagged polynucleotides comprising: said first primer including said first tag sequence; and sequences of said variable genetic region;

[0020] hybridizing a second primer to polynucleotides in a second sample, said second primer comprising a second tag sequence and a second probe sequence specific for a second target sequence, wherein said second target sequence is 3′ to the same variable genetic region as said first target sequence, wherein further said second probe sequence may be the same as or different from said first probe sequence;

[0021] elongating primer templates formed thereby to form a second population of tagged polynucleotides comprising: said second primer including said second tag sequence; and sequences of said variable genetic region;

[0022] mixing said first and second populations together;

[0023] determining sequences of polynucleotides comprising tag sequences and the sequences of the variable genetic element in said mixture;

[0024] from the tag sequences comprised in the polynucleotide sequences thus determined identifying the sample in which polynucleotide sequences occurred;

[0025] from the sequences of the variable genetic region comprised in the polynucleotide sequences thus determined identifying particular variants of said variable genetic element;

[0026] from this information determining the number of time one or more given variants occur in each sample, and

[0027] from the number for each variant in the polynucleotides thus determined, quantifying said polynucleotides in said samples;

[0028] wherein said sequences are determined without Southern blot transfer and/or without size-separating primer extension products and/or without electrophoresis.

[0029] Embodiments provide multiplex methods for the quantitative determination of polynucleotides in two or more samples, comprising:

[0030] hybridizing a first primer pair to polynucleotides in a first sample, the first primer of said first primer pair comprising a first tag sequence and a first probe sequences specific for a first target sequence and the second primer of said first primer pair comprising a second tag sequence and a second probe sequence specific for a second target sequence, wherein the first and the second probe sequences flank and hybridize to opposite strands of a variable genetic region;

[0031] elongating primer templates formed thereby to from a first population of tagged polynucleotides, each of said polynucleotides comprising: (a) the sequence of said first primer of said first primer pair, a sequence of said variable genetic region, and a sequence complementary to the sequence of said second primer of said first primer pair or (b) a sequence complementary to the sequence of said first primer of said first primer pair, a sequence of said variable genetic region and the sequence of said second primer of said first primer pair;

[0032] hybridizing a second primer pair to polynucleotides in a second sample, the first primer of said second primer pair comprising a third tag sequence and said first probe sequences specific for said first target sequence and the second primer of said second primer pair comprising a fourth tag sequence and said second probe sequence specific for said second target sequence;

[0033] elongating primer templates formed thereby to from a second population of tagged polynucleotides, each of said polynucleotides comprising: (a) the sequence of said first primer of said second primer pair, a sequence of said variable genetic region, and a sequence complementary to the sequence of said second primer of said second primer pair or (b) a sequence complementary to the sequence of said first primer of said second primer pair, a sequence of said variable genetic region and the sequence of said second primer of said second primer pair;

[0034] mixing said first and second populations together;

[0035] determining sequences of polynucleotides in said mixture, comprising the tag sequences and the variable genetic element;

[0036] from the tag sequences comprised in the polynucleotide sequences thus determined identifying the sample in which polynucleotide sequences occurred;

[0037] from the sequences of the variable genetic region comprised in the polynucleotide sequences thus determined identifying particular variants of said variable genetic element;

[0038] from this information determining the number of times given variants occur in each sample, and

[0039] from the number for each variant in the polynucleotides thus determined, quantifying said polynucleotides in said samples.

[0040] wherein said sequences are determined without Southern blot transfer and/or without size-separating primer extension products and/or without electrophoresis.

[0041] Embodiments provide methods in accordance with any of the foregoing or the following wherein given polynucleotide sequences in a sample is quantified by a method comprising normalizing the number occurrences determined for the given sequence. In embodiments the number of occurrences is normalized by dividing the number of occurrences determined for the given polynucleotide sequence by the total number of occurrences of polynucleotide sequences in the sample. In embodiments the given polynucleotide sequences is that of a given variant of a variable genetic region and, in embodiments, the quantity of the given variant in the sample is normalized by dividing the number of occurrences of that variant by the total number of occurrences of all variants of the variable genetic region in the sample.

[0042] Embodiments provide a multiplex method for determining polynucleotide sequences in two or more samples, comprising: attaching a first tag sequence to one or more polynucleotides of a first sample; attaching a second tag sequence different from said first tag sequence to one or more polynucleotides of a second sample; mixing the tagged polynucleotides of said first and second samples together; determining sequences of said polynucleotides comprising said first and said second tags; and identifying said first and second tags in said sequences; thereby identifying sequences of said polynucleotides of said first sample and second samples, wherein said sequences are determined without Southern blot transfer and/or without size-separating primer extension products and/or without electrophoresis.

[0043] Embodiments provide a multiplex method for determining polynucleotide sequences in two or more samples comprising:

[0044] attaching a first tag sequence, t₁, to P_1-1 through P_1-n1 polynucleotides in a first sample, thereby to provide a first plurality of polynucleotides tagged with said first tag, tiPii through t₁P_1-n1;

[0045] attaching a second tag sequence, t₂, to P_2-1 through P_2-n2 polynucleotides in a second sample, thereby to provide a second plurality of polynucleotides tagged with said second tag, t₂P_2-1 through t₂P_2-n2;

[0046] mixing together said polynucleotides tagged with said first and said second tags;

[0047] determining sequences of polynucleotides comprising said tags in said mixture;

[0048] identifying said first and second tags in said sequences and;

[0049] by said first tag identifying polynucleotide sequences of said first sample and by said second tag identifying polynucleotide sequences of said second sample;

[0050] wherein said sequences are determined without Southern blot transfer and/or without size-separating primer extension products and/or without electrophoresis.

[0051] Embodiments provide a method according to any of the foregoing or the following, wherein the number of said polynucleotides in said first sample, n1, is any of 2, 5, 10, 25, 50, 100, 150, 200, 250, 500, 1,000, 1,500, 2,000, 2,500, 5,000, 7,500, 10,000, 12,500, 15,000, 17,500, 20,000, 25,000, 30,000, 35,000, 40,000, 50,000, 75,000, 100,000, 150,000, 200,000, 250,000, 500,000, 1,000,000 or more, and the number of said polynucleotides in said second sample, n₂, is any of 2, 5, 10, 25, 50, 100, 150, 200, 250, 500, 1,000, 1,500, 2,000, 2,500, 5,000, 7,500, 10,000, 12,500, 15,000, 17,500, 20,000, 25,000, 30,000, 35,000, 40,000, 50,000, 75,000, 100,000, 150,000, 200,000, 250,000, 500,000, 1,000,000 or more.

[0052] Embodiments provide a method according to any of the foregoing or the following, wherein the number of said samples and of said different tags therefor is 5, 10, 15, 20, 25, 50, 75, 100, 150, 200, 250, 500, 1,000, 2,500, 5,000, 10,000 or more.

[0053] Embodiments provide a method according to any of the foregoing or the following, wherein the tags are nucleotide sequences that are 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36 nucleotides long or any combination thereof.

[0054] Embodiments provide a method according to any of the foregoing or the following, wherein the tags are incorporated into said polynucleotides by a step of ligation, provided that the step of ligation does not result in biasing.

[0055] Embodiments provide a method according to any of the foregoing or the following, wherein the tags are incorporated into said polynucleotides by a step of ligation and/or by a step of amplification.

[0056] Embodiments provide a method according to any of the foregoing or the following, wherein said tags are comprised in primers for amplification and are incorporated into said polynucleotides by amplification using said primers.

[0057] Embodiments provide a method according to any of the foregoing or the following, wherein said tags are incorporated into said polynucleotides by a process comprising a step of cloning into a vector.

[0058] Embodiments provide a method according to any of the foregoing or the following, wherein the tags are comprised in adapters for amplification and said adapters are ligated to polynucleotides in said samples. Embodiments provide a method in this regard, wherein further, said polynucleotides ligated thereby to said tags are amplified via said adapters. Embodiments provide a method in this regard, wherein further, said adapters comprise a moiety for immobilization. In embodiments said moiety is a ligand; in embodiments it is biotin. Embodiments provide a method in this regard, wherein further, said tags are comprised on adapters for bead emulsion amplification. In embodiments the adapters are suitable for use in a sequencing system of 454 Life Sciences or other sequencing system in which bead emulsion amplification is carried out.

[0059] Embodiments provide a method according to any of the foregoing or the following, wherein the primer for amplification comprises a sequence for PCR amplification, linear amplification, transcriptional amplification, rolling circle replication, or QB replication.

[0060] Embodiments provide a method according to any of the foregoing or the following, wherein the primer for amplification comprises a sequence for PCR amplification.

[0061] Embodiments provide a method according to any of the foregoing or the following, wherein each of said polynucleotides is disposed individually on a bead isolated from other polynucleotides.

[0062] Embodiments provide a method according to any of the foregoing or the following, wherein each of said polynucleotides is disposed individually on a bead isolated from other said polynucleotides, is amplified while disposed therein, and the amplification products thereof also are disposed on said bead.

[0063] Embodiments provide a method according to any of the foregoing or the following, wherein each of said polynucleotides is disposed individually on a bead isolated from other said polynucleotides, is amplified while disposed therein, the amplification products thereof also are disposed on said bead, and each said bead is disposed individually in a well isolated from other said beads.

[0064] Embodiments provide a method according to any of the foregoing or the following, wherein the sequences are determined by pyrosequencing.

[0065] Embodiments provide a method according to any of the foregoing or the following, wherein said samples are biological samples, each comprising one or more species.

[0066] Embodiments provide a method according to any of the foregoing or the following, wherein at least one sequence of said polynucleotides is specific to a particular organism.

[0067] Embodiments provide a method according to any of the foregoing or the following, wherein said sequences comprise a variable 16S rRNA sequence.

[0068] Embodiments provide a method according to any of the foregoing or the following, wherein said sequences comprise a variable 18S rRNA sequence, a variable rRNA ITS sequence, a mitochondrial sequence, a microsatellite sequence, a metabolic enzyme sequence, and/or a genetic disease sequence.

[0069] Embodiments provide a method according to any of the foregoing or the following, wherein the samples are microbial community samples.

[0070] Embodiments provide a method according to any of the foregoing or the following, wherein the samples are microbial community samples for clinical analysis of a patient.

[0071] Embodiments provide a method according to any of the foregoing or the following, wherein the samples are microbial community environmental samples.

[0072] Embodiments provide a method according to any of the foregoing or the following, wherein the samples are microbial community soil samples.

[0073] Embodiments provide a method according to any of the foregoing or the following, wherein the samples are microbial community water samples.

[0074] Embodiments provide a method according to any of the foregoing or the following, wherein the samples are samples for SNP analysis.

[0075] Embodiments provide a method according to any of the foregoing or the following, wherein the samples are samples for genotyping.

[0076] Embodiments provide a multiplex method according to any of the foregoing or the following for determining polynucleotide sequences of two or more samples, comprising,

[0077] amplifying polynucleotides of a first sample to produce first amplified polynucleotides comprising a first tag sequence;

[0078] separately amplifying polynucleotides of a second sample to produce second amplified polynucleotides comprising a second tag sequence different from said first tag sequence;

[0079] wherein the amplification products arising from different individual polynucleotides are spatially separated from one another;

[0080] mixing together amplicons of said first and second samples;

[0081] distributing the amplicons in the mixture into spatially distinct locations; sequencing the amplicons thus distributed using one or more primers that hybridize 5′ to said tag sequences;

[0082] identifying said tag sequences in the sequences of polynucleotides thus determined; and

[0083] identifying by said tags polynucleotides of said first sample and polynucleotides of said second sample.

[0084] Embodiments provide a method according to any of the foregoing or the following, comprising,

[0085] (a) for each sample separately: isolating polynucleotides to be sequenced, ligating said polynucleotides to a common adaptor comprising a tag sequence, and capturing individual ligated polynucleotides onto individual beads under conditions that provide predominately for the immobilization of 0 or 1 molecule per bead;

[0086] (b) thereafter mixing together said beads comprising said polynucleotides.

[0087] Embodiments provide a method according to any of the foregoing or the following, further comprising, amplifying bead-immobilized polynucleotides in droplets of an emulsion thereby to clonally amplify said individual polynucleotides on said beads, wherein amplification comprises amplification of said tag sequence.

[0088] Embodiments provide a method according to any of the foregoing or the following, further comprising, distributing individual droplets containing said amplified polynucleotides into wells under conditions that provide predominantly for 0 or 1 droplet per well, determining in individual wells the sequences of polynucleotides comprising said tag sequences, and by said tag sequences identifying polynucleotides of said first and said second samples.

[0089] In embodiments the invention provides methods in accordance with any of the foregoing or the following, for any one or more of detecting, monitoring, profiling, prognosticating, and/or diagnosing a disorder, disease, or the like.

[0090] In embodiments the invention provides methods in accordance with any of the foregoing or the following, for analyzing the composition, diversity, stability, dynamics, and/or changes in agricultural, food, biosecurity, veterinary, clinical, ecological, zoological, oceanological, and/or any other sample comprising one or more polynucleotides.

[0091] Embodiments provide kits comprising a plurality of two or more primers, each primer in said plurality comprising a tag sequence and a probe sequence specific to a target sequence, wherein:

[0092] (A) in each of said primers the probe sequence is 3′ to the tag sequence, but not necessarily adjacent thereto;

[0093] (B) in each of said primers: the tag sequence is different from the tag sequence of the other in the plurality; the tag sequence is not the complementary sequence to any other tag sequence in the plurality; the tag sequence does not contain any homodinucleotide sequences; the junction sequences between the tag sequence and the adjacent parts of the primer, if any, is not a homodinucleotide sequence;

[0094] (C) in each of said primers the probe sequence is complementary to the target sequence and the target sequence is located 3′ to a variable genetic region, and

[0095] (D) each of said primers is disposed separately from the others in containers in said kit.

[0096] Embodiments provide kits in accordance with any of the foregoing or the following, wherein each of said primers further comprises a priming sequence 5′ to the tag sequence but not necessarily adjacent thereto, and the priming sequence is the same in all of said primers, said kit further comprising a primer complimentary to and effective for polymerization from said priming sequence.

[0097] Embodiments provides kits comprising a plurality of two or more primers pairs, each primer in said plurality comprising a tag sequence and a probe sequence specific to a target sequence, wherein:

[0098] (A) in each of said primer the probe sequence is 3′ to the tag sequence, but not necessarily adjacent thereto;

[0099] (B) in each of said primers: the tag sequence is different from the tag sequence of the other in the plurality; the tag sequence is not the complementary sequence to any other tag sequence in the plurality; the tag sequence does not contain any homodinucleotide sequences; the junction sequences between the tag sequence and the adjacent parts of the primer, if any, is not a homodinucleotide sequence;

[0100] (C) in each of said primers the probe sequence is complementary to the target sequence,

[0101] (D) in each primer pair the probe sequences are specific to target sequences that flank a variable genetic region;

[0102] (E) each of said primers is disposed separately from the others in said kit.

[0103] Embodiments provides kits in accordance with any of the foregoing or the following, wherein, the primers further comprise a priming sequence 5′ to the tag sequence but not necessarily adjacent thereto, the priming sequence either is the same in all the primers, or one member of each pair has the same first priming sequence and the second member of each pair has the same second priming sequence, said kit further comprising disposed separately from one another in one or more containers one or more primers complementary to and effective for elongation from said priming.

[0104] Embodiments provide a kit useful in methods according to any of the foregoing or the following, comprising a set of primers and/or adapters, wherein each primer and/or adapter in said set comprises a tag sequence and a primer sequence. In embodiments the primers and/or adapters further comprise a moiety for immobilization. In embodiments the primers and/or adapters comprise biotin. In embodiments the primers and/or adapters in the set comprise all tag sequences defined by 2, 3, 4, 5, 6, 7, or 8 base polynucleotide sequences, wherein each of said primers and/or adapters are disposed in containers separate from one another. In embodiments there are 1-5, 3-10, 5-15, 10-25, 20-50, 25-75, 50-100, 50-150, 100-200, 150-500, 250-750, 100-1000, or more different tag sequences disposed separately from one another, so as to be useful for uniquely tagging said number of different samples. In embodiments the primers and/or adapters are suitable for use as 454 Life Sciences amplification adapters and/or primers. In embodiments the primers and/or adapters further comprise any one or more of a primer sequence for any one or more of a 16S rRNA sequence, an 18S rRNA sequence, an ITS sequence, a mitochondrial sequence, a microsatellite sequence, a metabolic enzyme sequence, a genetic disease sequence, and/or any other sequence for amplification or analysis.

[0105] In embodiments the invention provides a kit, in accordance with any of the foregoing or the following, comprising a set of primers and/or adapters for use in a method according to any of the foregoing or the following, wherein each primer and/or adapter in said set comprises a tag sequence, the tag sequence of each of said primers and/or adapters is different from that of the other primers and/or adapters in said set, the primers and/or adapters further comprise a priming sequence that is the same in all of the primers and/or adapters in said set, the tag sequences are located 5′ to the priming sequence and the different primers and/or adapters comprising each different tag sequence are disposed separately from one another. In embodiments the tags are any number of bases long. In embodiments the tags are 2, 3, 4, 5, 6, 8, 10, 12 bases long. In embodiments the tags are 4 bases long. In embodiments the priming sequence is specific to any target polynucleotide of interest. In embodiments the priming sequence is specific to a sequence in 16S rRNA. In embodiments the tags differ from each other by at least 2 bases. In embodiments the tags do not contain polynucleotide tracts within the tag. In embodiments the tags do not contain homo-polynucleotide tracts within or at the junction of the tag and PCR primer. In embodiments the tags do not contain polynucleotide tracts within or at the junction of the tag and emulsion PCR adapter. In embodiments, the tags are not reverse compliments of each other.

[0117] The meanings ascribed to various terms and phrases as used herein are illustratively explained below.

[0118] “A” or “an” means one or more; at least one.

[0119] “About” as used herein means roughly, approximately. Should a precise numerical definition be required, “about” means +/−25%.

[0120] “Adapter” means a polynucleotide sequence used to either attach single polynucleotide fragments to beads and/or to prime the emulsion PCR reaction and/or as a template to prime pyrosequencing reactions.

[0121] “ALH” is used herein to mean amplicon length heterogeneity.

[0122] “Amplicon” is used herein to refer to the products of an amplification reaction.

[0123] “Clonally amplified” is used herein generally to mean amplification of a single starting molecule. Typically it also refers to the clustering together of the amplification products, isolated from other amplification templates or products.

[0124] “dsDNA” means double stranded DNA.

[0125] Dysbiosis means a shift in a the species and abundance of species in a microbial community.

[0126] “Flanking” generally is used to mean on each side, such as on the 5′ and the 3′ side of a region of a polynucleotide—with reference to the 5′ and the 3′ ends of one or the other stand of a double stranded polynucleotide. Forward and reverse primers for amplifying a region of a polynucleotide by PCR, for instance, flank the region to be amplified.

[0127] “Microbial community sample” is used herein to refer to a sample, generally of a biological nature, containing two or more different microbes. Microbial community samples include, for instance, environmental samples, as well as biological samples, such as samples for clinical analysis. The term applies as well to preparations, such as DNA preparations, derived from such samples.

[0128] “Multiplex sequencing” herein refers to sequencing two or more types or samples of polynucleotides in a single reaction or in a single reaction vessel.

[0129] “PCO” means principal coordinates analysis.

[0130] “PCA” means principal component analysis.

[0131] “Picotiter plate” means a plate having a large number of wells that hold a relatively small volume, typically more wells than a 96-well microtiter plate, and smaller volumes than those of a typical 96-well microtiter plate well.

[0132] “Primer” means a polynucleotide sequence that is used to amplify PCR productsand/or to prime sequencing reactions.

[0133] “ssDNA” means single stranded DNA.

[0134] “Tag,” “Tag sequence,” etc. means typically a heterologous sequence, such as a polynucleotide sequence that identifies another sequence with which it is associated as being of a given type or belonging to a given group.

[0135] “Variable genetic region” as used herein means a genetic region that varies, such as between individuals of a species and between species. The phrase does not denote a specific length, but, rather is used to denote a region comprising a variation the exact length of which may vary and may differ in different contexts. As to a double stranded polynucleotide, the term includes one or the other and both stands of the region, and may be used to refer to one, the other, or to both strands, and it will generally be clear from the context which is meant. A specific example of a genetic region that varies between individuals, provided for illustration only, is a genetic region that contains an SNP (single nucleotide polymorphism) site. By variable genetic region in this regard is meant a region containing the SNP site. Different sequences of the SNP in this regard constitute the variants of the variable genetic region. A specific example of a variable genetic region that differs between species is the genes for 16S RNA which vary characteristically between microbes and can be used to identify microbes in mixed community samples as described in greater detail in some of the examples herein.

[0136] In certain aspects and embodiments the invention relates to multiplex sequencing analysis using tags. In various aspects and embodiments of the invention in this regard the invention provides methods for sequencing two or more samples simultaneously in a mixture with one another, wherein each sample is first linked to a sample-specific sequence tag, the tagged samples are mixed and sequenced, and the sequences from each sample then are identified by their respective sample-specific sequence tags.

[0137] FIG. 1 provides a general depiction of various aspects and embodiments of the invention in this regard, and the figure is discussed by way of illustration below with reference to sequencing DNA from different samples. A plurality of samples (S₁, S₂, through S_j) is shown topmost in the Figure. Each sample is comprised of a plurality of polynucleotides (P_1-1 to P_1-n1 in S₁; P_2-1 to P_2-n2 in S₂; through P_j-1 to P_j-nj). The polynucleotides in each sample are labeled separately with a tag polynucleotide sequence, all the polynucleotides in a given sample being tagged (in this illustration) with a single tag sequence, designated in the figure as T₁ for S₁, T₂ for S₂, through T_j for S_j. The individual tagged polynucleotides are denoted accordingly. The tagged polynucleotides in each sample are designated collectively, for each sample, T₁S₁, T₂S₂ through T_jS_j. The tagged polynucleotides from the samples are mixed together to form a mixture, designated M. The mixture is sequenced typically by a parallel sequencing method. The tag sequences are identified in the data thus obtained. The sequences are grouped by tag. The sequences from the individual samples are thereby identified.

[0138] In embodiments tags are 3 to 30, 4 to 25, 4 to 20 base long sequences. In embodiments the tags are 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36 nucleotides long or any combination thereof.

[0139] In embodiments there are 1-6, 6-12, 10-15, 10-20, 15-25, 20-40, 25-50, 25-75, 50-100, 50-150, 100-200, 100-250, 50-250, 100-500, 500-1,000, 100-1,000, 500-5,000, 100-10,000, 1,000-25,000, 500-50,000, 100-100,000, 1-1,000,000 or more samples, tagged, respectively, with 1-6, 6-12, 10-15, 10-20, 15-25, 20-40, 25-50, 25-75, 50-100, 50-150, 100-200, 100-250, 50-250, 100-500, 500-1,000, 100-1,000, 500-5,000, 100-10,000, 1,000-25,000, 500-50,000, 100-100,000, 1 1,000,000 or more different tags.

[0140] In embodiments the sequences are determined without the use of gel electrophoresis. In embodiments the sequences are determined without the use of transfer of sequences from a gel onto a membrane or a filter for hybridization. In embodiments, sequences are determined by a parallel sequencing method. In embodiments the sequences are determined by pyrosequencing, sequencing by synthesis, hybridization sequencing, subtractive sequencing, pore sequencing or direct read sequencing.

[0141] In embodiments the tags are incorporated into polynucleotides in samples for sequencing by a step of ligation and/or by a step of amplification.

[0142] In embodiments the tags are comprised in primers for amplification.

[0143] In embodiments the tags are comprised in primers for PCR amplification, transcription amplification, rolling circle amplification, or amplification by Q13 replicase.

[0144] In embodiments the tags are comprised in emulsion PCR adapters and primers for amplification.

[0145] In embodiments the tags are incorporated by a step of cloning into a vector.

[0146] In embodiments the samples are microbial community samples. In embodiments the samples are clinical samples. In embodiments the samples are environmental samples. In embodiments the samples are samples for SNP analysis. In embodiments the samples are samples for genotyping. In embodiments the sequences are determined in one or more picotiter plates.

[0147] In embodiments the samples are fragmented genomic DNAs. In embodiments the samples are fragmented Bacterial genomic DNA, Archae genomic DNA, Fungal genomic DNA, Eukaryotic genomic DNA, chloroplast DNA, and/or mitochondrial DNA. In embodiments the samples are cDNAs. In embodiments the samples are Eukaryotic cDNA, Bacterial cDNA, Archae cDNA, and/or Fungal cDNA. In embodiments the tags are incorporated by a step of ligation and/or by a step of amplification.

[0148] In embodiments the samples are for any one or more of detecting, monitoring, profiling, prognosticating, and/or diagnosing a disorder, disease, or the like.

[0149] In embodiments the samples are for analyzing the composition, diversity, stability, dynamics, and/or changes in agricultural, food, biosecurity, veterinary, clinical, ecological, zoological, oceanological, and/or any other sample comprising one or more polynucleotides.

[0150] In embodiments the sequences are determined in wells of a titer plate. In embodiments the sequences are determined in one or more picotiter plates having a mask. In embodiments the sequences are determined in one more picotiter plates having a mask, wherein the mask defines 2, 4, 8, 16, 32, 64 or more compartments.

[0151] By way of illustration to a 454 picotiter plate, in embodiments there are about 120,000 templates/plate and the read length averages about 250 bases per template. In embodiments relating thereto there are 10 tags of 4 bases per 1/16 plate, 160 tags total, an average of about 750 templates per tag (and per sample), and about 187,500 bases sequenced per tag (and per sample).

[0152] In embodiments there are about 260,000 templates/plate and the read length averages about 250 bases per template. In embodiments relating thereto, there are 12 tags of 4 bases per ⅛ plate, 96 samples total, an average of about 2,708 templates per tag (and per sample) and about 677,083 bases of sequence per tag (and per sample).

[0153] In embodiments there are about 400,000 templates/plate and the read length averages about 250 bases per template. In embodiments relating thereto, there are 96 tags of 6 bases for 96 samples per plate, about 4,166 templates per tag (and per sample) and about 1,041,666 bases of sequence per tag (and per sample).

[0154] In embodiments the tags are 10 base long sequences, there are 192 different tags, and the samples are analyzed in microtiter plate format.

[0155] In embodiments the invention provides algorithms for deconvolving, from a mixture of sequences from two or more samples, the sequences of the samples in the mixture by identifying sample-specific tags in the sequences, grouping the sequences by the tags thus identified, thereby grouping together the sequence from each of said samples, apart from one another.

[0156] In embodiments the invention provides algorithms for deconvolving, from a mixture of sequences from two or more samples, the sequences of the samples in the mixture by identifying sample-specific tags in sequences, as follows:

[0157] 1. Read all sequence reads into an array;

[0158] 2. Search the beginning of each sequence read and identify the tag;

[0159] 3. Build an associative array linking tag with sequence read;

[0160] 4. Sort the keys for the associate array;

[0161] 5. Associate each key with the corresponding sample;

[0162] 6. Pool all sequence reads for each sample;

[0163] 7. Analyze each sample separately.

[0164] 8. Normalize the abundance of each component within each samples with respect to the total reads within that sample.

[0165] In embodiments the algorithm can be implemented in any programming language. In embodiments the algorithm is implemented in C, C++, JAVA, Fortran, or Basic. In embodiments the algorithm is implemented as a PERL script.

[0166] In embodiments the invention provides kits for multiplex sequencing as described herein, comprising a set of primers and/or adapters, wherein each primer and/or adapter in said set comprises a tag sequence, a primer sequence and/or an emulsion PCR adapter. In embodiments the primers and/or adapters further comprise a moiety for immobilization. In embodiments the primers and/or adapters comprise biotin. In embodiments the primers and/or adapters in the set comprise all tag sequences defined by 2, 3, 4, 5, 6, 7, or 8 base polynucleotide sequences, wherein said primers and/or adapters comprising different tag sequences are disposed in containers separate from one another. In embodiments there are 1-5, 3-10, 5-15, 10-25, 20-50, 25-75, 50-100, 50-150, 100-200, 150-500, 250-750, 100-1000, or more different tag sequences disposed separately from one another, so as to be useful for uniquely tagging said number of different samples. In embodiments the primers and/or adapters are suitable for use as 454 Life Sciences amplification adapters and/or primers. In embodiments the primers and/or adapters further comprise any one or more of a primer sequence for any one or more of a 16S rRNA sequence, an 18S rRNA sequence, an ITS sequence, a mitochondrial sequence, a microsatellite sequence, a metabolic enzyme sequence, a genetic disease sequence, and/or any other sequence for amplification or analysis.

[0167] The present invention is additionally described by way of the following illustrative, non-limiting examples.

[0168] 454 Life Sciences, a subsidiary of Roche Diagnostics, provides a device for pyrosequencing approximately 100,000,000 bases of about 400,000 different templates in a single run on a single picotiter plate. The company also provides masks that allows for the processing 2, 4, 8, or 16 different samples on one plate. At maximum capacity using the masked plate, the system provides about 1 million bases of sequence data on about 4,000 templates for each of 16 samples.

[0169] The general process of sequencing using the 454 system is generally as follows: isolate DNA; optionally fragment the DNA; optionally render the DNA double stranded; ligate the DNA to adaptors; separate the strands of the dsDNA, bind the ssDNA to beads under conditions that result in a preponderance of beads that have either no DNA molecule bound to them or a single molecule of DNA bound to them; capture the beads in individual droplets of an emulsion of a PCR reaction mix in oil; carry out a PCR reaction on the emulsion-encapsulated bead-DNAs (whereby amplification products are captured on the beads); distribute the amplification products into picoliter wells so that there is either no bead in a well or one bead; and carry out pyrosequencing on all the beads in all the wells in parallel.

[0170] 454 Life Sciences, a subsidiary of Roche Diagnostics, provides a device for pyrosequencing approximately 100,000,000 bases of sequence for about 400,000 different templates in a single run on a single picotiter plate. At maximum capacity using the plate, the system provides about 10 million bases of sequence data for each of about 4,000 templates for each of 96 multitagged samples. In this example the 96 tags are 6 bases in length and are used along with 6 bases of the forward or reverse primer to identify the reads that belong with each of the 96 individual samples (see FIG. 2).

[0171] Various aspects and embodiments of the invention herein described are illustrated by way of the following general example relating to “ecogenomic” analysis of microbial diversity in biological samples.

[0172] The ability to quantify the number and kinds of microorganisms within a community is fundamental to the understanding of the structure and function of an ecosystem, as discussed in, for instance, Pace 1997 and Theron and Cloete 2000. Traditionally, the analysis of microbial communities has been conducted using microbiological techniques, but these techniques are limited. For instance they are not useful for the many organisms that cannot be cultivated (Ritchie, Schutter et al. 2000; Spring, Schulze et al. 2000). Even for those organisms that can be cultured, these techniques provide little information with which to identify individual microbes or characterize their physiological traits. (Morris, Bardin et al. 2002).

[0173] Recent advances in molecular techniques have overcome some of these disadvantages, and have enabled the identification of many more taxa in microbial communities than traditional microbial techniques. These advances have provided considerable insight into the expression of key functions in species in microbial communities. (Pace 1997; Suzuki 1998; Amann 2000; Frischer, Danforth et al. 2000; Ritchie, Schutter et al. 2000; Spring, Schulze et al. 2000). Among these molecular techniques are Denaturing Gradient Gel Electrophoresis (DGGE), Temperature Gradient Gel Electrophoresis (TGGE), Temporal Temperature Gradient Gel Electrophoresis (TTGE), Terminal-Restriction Fragment Length Polymorphism (T-RFLP), Single Strand Conformation Polymorphism (SSCP), and Length Heterogeneity PCR (LH-PCR) (Frischer, Danforth et al. 2000; Theron and Cloete 2000; Mills, Fitzgerald et al. 2003; Seviour, Mino et al. 2003; Klaper and Thomas 2004).

[0174] Among these, LH-PCR is probably the best technique for fingerprinting. It is inexpensive, fast, and can be used routinely to screen several hundred samples a day. It is useful as a routine survey tool that can be used to monitor the dynamics of natural soil microbial communities, and to quickly identify samples of interest by PCO analysis. LH-PCR has been used to extensively assess natural variation in bacterial communities by profiling the amplified variable regions of 16S rRNA genes in mixed microbial population samples, using PAGE. (See Mills 2000; Litchfield and Gillevet 2002; Lydell, Dowell et al. 2004). The LH-PCR products of the individual species in the population give rise to distinct bands in the gels. The “peak area” of each band is proportional to the abundance of the species in the community. LH-PCR of 16S rRNA variable regions has been used quite successfully to estimate species diversity in bacterioplankton communities, in particular. (See Suzuki, Rappe et al. 1998; Ritchie, Schutter et al. 2000).

[0175] Community functionality cannot be determined directly from 16S rRNA clone data, however, it must be inferred from the data by phylogenetic analysis. Furthermore, LH-PCR and other fingerprinting technologies, while powerful tools for monitoring population dynamics, cannot identify individual species in a community. For this, fingerprinting investigations must be followed up by library construction, cloning, sequencing, and phylogenetic analysis. (Fitzgerald 1999; McCraig 1999; Spring, Schulze et al. 2000; Theron and Cloete 2000; Litchfield and Gillevet 2002; Bowman and McCuaig 2003; Kang and Mills 2004; Eckburg, Bik et al. 2005). Identifying species of a fingerprinting study, thus, is a considerable undertaking that is inconvenient, time-consuming, expensive and subject to technical limitations.

[0176] Grouping samples can, to some extent, reduce the cost, time, and expense of such analyses. For instance, PCO analysis of LH-PCR data can be used to group samples with similar profiles for batch cloning and sequencing. Combining the samples this way reduces the time, expense, and work involved in analyzing the samples. Sequencing of at least 300 random clones is required to identify the bacterial components of the pooled sample down to 1% of the total bacterial populations in typical samples. This level of resolution is similar to that of ALH fingerprinting. Originally a novel approach, pooling similar samples prior to cloning and sequencing has proven to be robust and effective.

[0177] In classic community studies in the literature (Eckburg, Bik et al. 2005), environmental samples are assayed independently. Then the clone sequence data from specific classes/groups are statistically analyzed usually using some sort of averaging metric. Analyses of this type can be extremely costly, especially if the clone libraries are exhaustively analyzed, something that typically involves sequencing thousands of clones. Moreover, for the “averaging” process to be valid, as required for comparing the mixed populations, the samples must be pooled in equal proportions. While simple in principle, in reality, it is difficult to accomplish and, even if accomplished, impossible to verify. A new technique, based on pyrosequencing, offers advantages that overcome a variety of these drawbacks of the fingerprinting technologies mentioned above. The method is implemented on an instrument sold by 454 Life Sciences, Inc., a subsidiary of Curagen Sciences, Inc., using reagents provided by the same company. In addition, 454 Life Sciences provides a custom service for pyrosequencing.

[0178] In this technology, individual DNA molecules are amplified on beads by PCR in individual droplets in an oil-in-water emulsion. Beads then are deposited individually in wells of a picotiter plate. The sequences of all the DNAs in the wells are determined in parallel by pyrosequencing. (See Venter, Levy et al. 2003; Margulies, Egholm et al. 2005; Poinar, Schwarz et al. 2006). In a typical run, there are about 200,000 templates per plate, an average read length of about 100 bases from each template, and a single-plate run generates about 20 million bases of sequence in a single four hour run.

[0179] Although the technology greatly increases throughput over previous methods, it is expensive. In particular, the cost per plate is too high for it to be economically practical to carry out many analyses. To decrease cost, masks can be used that divide a plate into 16 independent sample zones, so that one plate can be used to process 16 different samples, either at the same time or independently. Each 1/16 zone provides about 1,000,000 bases of sequence data from about 10,000 different templates. While this reduces the cost per sample, the expenses associated with using this technology remain undesirably high.

[0180] Various aspects and embodiments of the present invention can be used to further reduce the cost per sample of this technology (as well as other techniques, as described elsewhere herein). The use of multitagging techniques (referred to as, among other things, “Multitag Process”) to the genomic analysis of bacterial populations in according with certain aspects and embodiments of the invention, notably high coverage sequencing of bacterial communities, is referred to herein as “Multitag Ecogenomics” and also as “Multitag Ecogenomic Analysis.”

[0181] (Several publications use the term “Multiplex Pyrosequencing” (Pourmand, Elahi et al. 2002) to refer to generating a composite signal from multiple targets that is read as a signature for a specific sample. The term is not used to refer to tag-based multiplexing in which sequences from different samples in a mixture are determined and then deconvolved from the mixed sequencing data using sample-specific tags incorporated during amplification reactions.)

[0182] As described below the Multitag Process in a relatively simple series of steps accomplishes everything that otherwise would require not only community fingerprinting analysis, but also all of the cloning and sequencing processes previously required for high coverage Ecogenomic Analysis using conventional techniques.

[0183] By way of illustration, the following example describes the use of Multitag Ecogenomic Analysis of variable regions of common genes using tagged universal primers for high coverage analysis of several microbial community samples all at the same time. The analysis is carried out much as described in general above, and further elaborated on in detail below.

[0184] Briefly, short tags are added to the 5′ ends of the forward and reverse PCR primers normally used for community analysis. These tags can be placed between the Emulsion PCR adapters and the PCT primers (see FIG. 2). A different tag is attached to the primers for each of the samples to be combined. For instance primers that span a variable region of 16S rRNA genes may be used for analysis of bacterial and archael communities. 16S rRNA-specific primers with 4 base tags are set out in the Table 1 below. Likewise primers that span a variable region of an ITS gene may be used for analysis of fungal communities. It will be appreciated that the choice of these specific primers is not exclusive, and that a wide variety of other primers suitable to other target regions for amplification may be employed in much the same manner as descried herein for the 16S and ITS genes. Thus, any gene of interest can be used that provides conserved primer sites across a community, and sufficient variation in the region between the primers for the desired resolution of individual species. Thus, for example, genes specific to functional pathways such as anaerobic methane oxidation, or sulphur reduction can serve as targets for the amplification reaction, as well as 16S rRNA sequences.

TABLE 1
Name	Tag	Forward Shared Sequence
		AGCTAGAG TTT GATCMTGGCTCAG
L27FA	AGCT	AGCTAGAGTTTGATCMTGGCTCAG
L27FB	AGTC	AGTCAGAGTTTGATCMTGGCTCAG
L27FC	GATC	GATCAGAGTTTGATCMTGGCTCAG
L27FD	GACT	GACTAGAGTTTGATCMTGGCTCAG
L27FE	CTGC	CTGCAGAGTTTGATCMTGGCTCAG
L27FF	CTAG	CTAGAGAGTTTGATCMTGGCTCAG
L27FG	ATGC	ATGCAGAGTTTGATCMTGGCTCAG
L27FH	ATAG	ATAGAGAGTTTGATCMTGGCTCAG
L27FM	ATCT	ATCTAGAGTTTGATCMTGGCTCAG
L27FO	ATAT	ATATAGAGTTTGATCMTGGCTCAG
Name	Tag	Reverse Shared Sequence
		AGCTGCTGCCTCCCGTAGGAGT
355RA	AGCT	AGCTGCTGCCTCCCGTAGGAGT
355RB	AGTC	AGTCGCTGCCTCCCGTAGGAGT
355RC	GATC	GATCGCTGCCTCCCGTAGGAGT
355RD	GACT	GACTGCTGCCTCCCGTAGGAGT
355RE	CTGC	CTGCGCTGCCTCCCGTAGGAGT
355RF	CTAT	CTATGCTGCCTCCCGTAGGAGT
355RG	ATGC	ATGCGCTGCCTCCCGTAGGAGT
355RH	ATAT	ATATGCTGCCTCCCGTAGGAGT
355RM	ATCT	ATCTGCTGCCTCCCGTAGGAGT
355RO	ATAC	ATACGCTGCCTCCCGTAGGAGT

[0185] Table 1 shows a 16S rRNA-specific primer with a variety of 4 base tag sequences attached. As described herein such primers are useful for amplifying 16S rRNAs in several samples that can then be sequenced together. The 16S rRNA in each sample is amplified using a different tag, but the same 16S primer sequence. The amplified rRNA sequences from the samples are combined and sequenced together. The rRNA sequences from the different samples then are identified and sorted out by their 4 base tag sequence plus the first 4 bases of each primer. It is to be appreciated that the sequences downstream of the shared 16S primer sequence will differ among the samples, as well as the tag sequence.

[0186] In each case, the samples are individually amplified. The resulting amplicons comprise the primer sequences including the tags. Since unique tags are used for each sample, the tags in the amplicons from each sample will be different. The amplified DNAs are then pooled and sequenced by pyrosequencing as described above. The sequence data from a run is analyzed, in part, by grouping together all the sequences having the same tag. In this way, the sequences from each sample are demultiplexed from the sequencing data obtained from the mixture.

[0187] The working of the invention in this regard is illustrated by the following simulation, carried out using conventionally obtained population data from cold seep samples. The algorithm for sequence analysis uses a PERL script to extract the first 100 bases of sequence. It then analyzes all the 100 bases sequences using a custom RDP PERL script. The script works as follows:

[0188] 1. Read all sequence reads into an associate array (Hash 1);

[0189] 2. Extract 100 base subsequences from the beginning of each sequence read;

[0190] 3. Create an associate array (Hash 2) of the sequences;

[0191] 4. Perform a Blast search of the RDP database with Hash 1;

[0192] 5. Perform a Blast search of the RDP database with Hash 2;

[0193] 6. Compare the identifications for the original sequence (Hash 1) and the subsequence (Hash 2);

[0194] 7. Compile a list of similar identifications for Hash 1 and Hash 2;

[0195] 8. Compile a list of different identifications for Hash 1 and Hash 2;

[0196] 9. Calculate the percentage of similar identifications.

[0197] As shown below, there is virtually no difference at the class level in the microbial diversity generated by the sequencing simulation and that derived directly from the 16S rRNA sequences in the data base.

TABLE 2
RDP Class	First 100mer	16S rRNA
ALPHA_SUBDIVISION	3.6%	3.6%
ANAEROBIC_HALOPHILES	3.6%	3.6%
BACILLUS-LACTOBACILLUS-STREPTOCOCCUS SUBDIVISION	3.6%	3.6%
BACTEROIDES_AND_CYTOPHAGA	7.1%	7.1%
CHLOROFLEXUS_SUBDIVISION	3.6%	3.6%
CY.AURANTIACA_GROUP	7.1%	7.1%
CYANOBACTERIA	7.1%	7.1%
DELTA_SUBDIVISION	14.3%	14.3%
ENVIRONMENTAL_CLONE_WCHB1-41_SUBGROUP	7.1%	7.1%
FLX.LITORALIS_GROUP	3.6%	3.6%
GAMMA_SUBDIVISION	10.7%	10.7%
HIGH_G + C_BACTERIA	7.1%	7.1%
LEPTOSPIRILLUM GROUP	3.6%	3.6%
MYCOPLASMA_AND_RELATIVES	3.6%	3.6%
PIRELLULA_GROUP	3.6%	3.6%
SPHINGOBACTERIUM_GROUP	3.6%	3.6%
SPIROCHAETA-TREPONEMA-BORRELIA_SUBDIVISION	3.6%	3.6%
THERMOANAEROBACTER_AND_RELATIVES	3.6%	3.6%

[0198] Inflammatory Bowel Diseases (IBD or IBDs), namely ulcerative colitis (UC) and Crohn's disease (CD), are chronic, lifelong, relapsing illnesses, affecting close to 1 million Americans and costing approximately $2 billion per year to the US healthcare system. IBDs are of unknown cause, have no cure, and are increasing in incidence. The natural course of these diseases is characterized by periods of quiescence (inactive disease) interspersed with flare-ups (active disease). It is now widely accepted that flare-ups of IBD are due to a dysregulated inflammatory reaction to abnormal intestinal microflora dysbiosis), however.

[0199] Specific changes in the microflora of IBD patients that might cause these diseases remain unknown. Narrow searches for a single pathogen that causes IBD have been unsuccessful. (See Guarner and Malagelada 2003). Studies of small bacterial groups have yielded ambiguous results. (See Schultz and Sartor 2000). Only recently have studies of large sets of bacterial flora been attempted. (See Eckburg, Bik, et al. 2005). Improving our knowledge about GI tract microflora has the potential to revolutionize IBD treatment. Development of real-time methods to study microfloral changes may lead to diagnostic tools to predict flare-ups, and to targeted, safe treatments for IBD.

[0200] The key requirement to understanding dysbiosis in polymicrobial diseases is for a method to interrogate widely the microflora in numerous control and disease samples to identify dynamic trends in species composition associated with health and disease progression. In classic community studies (Eckburg, Bik, et al. 2005) environmental samples are assayed independently and then the clone sequence data from specific classes/groups are statistically analyzed usually using some sort of averaging metric. This can be extremely costly, especially if the clone libraries are exhaustively analyzed (i.e., 10,000 clones per sample).

[0201] To improve throughput and reduce cost, Amplicon Length Heterogeneity PCR (ALH-PCR) has been used to study the gut microflora. It offers a rapid way of screening complex microbial communities, allowing for easy fingerprinting of microfloral changes. The LH-PCR fingerprinting is inexpensive and fast, with the ability to screen several hundred samples a day. It can be used as a routine survey tool to monitor the dynamics of natural soil microbial communities or to quickly identify samples of interest using PCO analysis. PCO analysis has been used to group samples with similar profiles, allowing them to be pooled for cloning and sequencing. This greatly reduces the cost of analyzing multiple samples, particularly when the analysis requires sequencing at least 300 random clones to identify bacterial components of the sample down to 1% representation in the total population (which is the resolution limit for ALH fingerprinting). Pooling similar samples before cloning and sequencing has proved to be quite robust. However, equal amounts of the PCR product from each sample must be pooled or the results will be skewed.

[0202] Multitag Pyrosequencing is a novel pyrosequencing technology that allows many community samples to be sequenced together at high coverage without the necessity for fingerprinting, cloning, or the purification and separation techniques required by conventional methods for analyzing microbial communities, as described herein above. Multitag sequencing is more efficient, faster, and less costly than other methods.

[0203] By way of illustration, Multitag Pyrosequencing can be carried out using a set of specific tags on the end of standard universal small ribosomal sub-unit (“SSU”) rRNA primers (See Table 1). A different set of the tagged primers is used to amplify the SSU rRNA in each different environmental sample (FIG. 2—Step 1). The PCR amplicons from all the samples are pooled. Emulsion PCR is performed and the amplicons arising from each molecule are captured on their respective beads. Following amplification, the beads are distributed into the wells of a picoliter plate (FIG. 2—Step 2). The sequences, including the tagged sequences, of the amplicons on each bead are determined by pyrosequencing (FIG. 2—Step 3). A PERL script or other suitable program is used to sort the sequence information using the tags and primer sequence as a key. Sequences with the same tags are identified thereby with their respective sample. The bacteria species in each sample then are identified by matching the SSU rRNA sequences to entries in the database of the Ribosomal Database Project (either RDP 8.1 or RDP 9.0). The normalized frequency with which a bacteria is thus identified in a given sample is indicative of its relative representation in the microbial community. Histograms based on these frequency determinations can be used for the non-parametric analysis of dysbiotic shifts involved in disease states.

[0204] For example, FIG. 3 depicts the results of such an experiment in which six Control, ten Crohns, and eight Ulcerative colitis mucosal samples were analyzed by Multitag Pyrosequencing. Each of the segments in the stacked histogram bars represents the normalized abundance of that specific taxa in a specific sample. In this experiment, identification of the taxa was performed using BLAST analysis of the RDP 8.1 database. It can be seen that some taxa (i.e. Bacillus fragilis subgroup and Rumanococcus gnavus subgroup) are present in the same abundance in both control and disease states. Other taxa, such as Clostridium leptum are more dominant in Ulcerative colitis, while others (i.e. the Gloeothece gloeocapsa subgroup) are indicators of dysbiosis in the disease state.

[0205] However, the standard 454 Life Science process using a ligation step to link the emulsion PCR adapters to the PCR amplicons and produces numerous artifacts in the quantitation of the abundances of each taxa in the samples. In the results displayed in FIG. 3, we algorithmically removed chimeras, reverse reads and truncated products and filtered the data to remove all taxa that were represented by less than 5% abundance. Only then were we able to see a correlation with disease state and specific microbial taxa.

[0206] In one experiment we used tagged PCR primers to amplify the components in duplicate microbial community samples, ligated the Emulsion PCR adapters to these samples, and then subjected these samples to separate pyrosequencing runs. The amplicons are routinely run on an Agilent Bioanalyzer system before and after ligation to quantitate the mixture before emulsion PCR. FIG. 4 depicts a sample run on the Bioanalyzer before and after direct ligation and clearly shows that the ligation step has drastically altered the distribution of the amplicons.

[0207] Additionally, we compared the normalized abundances of the component taxa identified by the multitag process after direct ligation of the Emulsion PCR adapters. In this experiment, identification of the taxa was performed using a Bayesian analysis of the RDP 9.0 database. We can see in FIG. 5 that abundances of the forward and reverse primers for various taxa are different within a sample and between duplicate samples. In several cases, we are missing entire families in the comparison between duplicates. Table 3 summarizes the differences between the forward primers and the reverse primers of the duplicate samples and it is clearly stochastic with no predictable pattern. We hypothesize that this differential ligation efficiency could be due to a number of factors such as internal structure in the amplicons or biases in the terminal nucleotide of either the adapter or amplicon.

TABLE 3
Duplicate Sample Analysis
	FORWARD PRIMERS	REVERSE PRIMER
RDP 9.0 FAMILY	RATIOS	RATIOS
Acidaminococcaceae	544.6%	195.0%
Actinomycetales	144.0%	116.5%
Bacteroidaceae	119.9%	124.5%
Clostridiaceae	97.5%	99.4%
Comamonadaceae	198.0%
Coriobacteriales	181.5%	141.5%
Enterobacteriaceae	4.2%
Eubacteriaceae	88.0%	87.5%
Flavobacteriaceae	34.9%
Incertae sedis 9	106.4%	143.0%
Lachnospiraceae	176.8%	113.1%
Peptococcaceae		91.0%
Peptostreptococcaceae	94.7%	115.4%
Porphyromonadaceae	99.0%	97.3%
Prevotellaceae	264.0%	88.1%
Rikenellaceae	212.2%	106.1%
Streptococcaceae	74.3%	60.7%

[0208] Each of the following publications is incorporated herein by reference in its entirety, particularly as to the above-referenced subject matter, especially relating to methods that can be employed in carrying out multitag sequencing and/or relating to uses thereof.

• [0209] Amann, R. (2000). “Who is out there? Microbial Aspects of Biodiversity.” System. Appl. Microbiol. 23: 1-8.
• [0210] Bowman, J. P. and R. D. McCuaig (2003). “Biodiversity, Community Structural Shifts, and Biogeography of Prokaryotes within Antarctic Continental Shelf Sediment.” Appl. Environ. Microbiol. 69(5): 2463-2483.
• [0211] Eckburg, P. B., E. M. Bik, et al. (2005). “Diversity of the human intestinal microbial flora.” Science 308: 1635-1638.
• [0212] Fitzgerald, K. M. (1999). Microbial Community Dynamics During the Bench-Scale Bioremediation of Petroleum-Contaminated Soil. Department of Biology. Fairfax, Va., George Mason University: 73.
• [0213] Frischer, A. E., J. M. Danforth, et al. (2000). “Whole-cell versus total RNA extraction for analysis of microbial community structure with 16S rRNA-targeted oligonucleotide probes in salt marsh sediments.” Appl. Environ. Microbiol. 66(7): 3037-3043.
• [0214] Guarner, F., and J. R. Malagelada. (2003). “Gut flora in health and disease.” Lancet 361: 512-9.
• [0215] Kang, S. and A. L. Mills (2004). “Soil Bacterial Community Changes Following Disturbance of the Overlying Plant Community.” Soil Science 169: 55-65.
• [0216] Klaper, R. and M. Thomas (2004). “At the crossroads of genomics and ecology: the promise of a canary on a chip.” BioScience 54: 403-412.
• [0217] Litchfield, C. D. and P. M. Gillevet (2002). “Microbial diversity and complexity in hypersaline environments: A preliminary assessment.” Journal of Industrial Microbiology & Biotechnology 28(1): 48-55.
• [0218] Lydell, C., L. Dowell, et al. (2004). “A population survey of members of the phylum Bacteroidetes isolated from salt marsh sediments along the east coast of the United States.” Microbial ecology 48(2): 263-73.
• [0219] Margulies, M., M. Egholm, et al. (2005). “Genome sequencing in microfabricated high-density picolitre reactors.” Nature, 2005 Sep. 15, 437(7057):376-80. Epub: 2005 Jul. 31.
• [0220] McCraig, A. E., L. Glover, J. I. Prosser (1999). “Molecular analysis of bacterial community structure and diversity in unimproved and improved upland grass pastures.” Appl. Environ. Microbiol. 65: 1721-1730.
• [0221] Mills, D. (2000). Molecular Monitoring of Microbial Populations during Bioremediation of Contaminated Soils. Environmental Sciences and Public Policy/Biology. Fairfax, Va., George Mason University: 217.
• [0222] Mills, D. K., K. Fitzgerald, et al. (2003). “A Comparison of DNA Profiling Techniques for Monitoring Nutrient Impact on Microbial Community Composition during Bioremediation of Petroleum Contaminated Soils.” J. Microbiol. Method 54: 57-74.
• [0223] Morris, C. E., M. Bardin, et al. (2002). “Microbial biodiversity: approaches to experimental design and hypothesis testing in primary scientific literature from 1975 to 1999.” Microbiology and Molecular Biology Reviews 66: 592-616.
• [0224] Pace, N. R. (1997). “A Molecular View of Microbial Diversity and the Biosphere.” Science 276: 734-739.
• [0225] Poinar, H. N., C. Schwarz, et al. (2006). “Metagenomics to paleogenomics: large-scale sequencing of mammoth DNA.” Science, 2006 Jan. 20, 311(5759):392-4. Epub: 2005 Dec. 20.
• [0226] Pourmand, N., E. Elahi, et al. (2002). “Multiplex Pyrosequencing.” Nucleic acids research 30(7): 31.
• [0227] Ritchie, N. J., M. E. Schutter, et al. (2000). “Use of Length Heterogeneity PCR and Fatty Acid Methyl Ester Profiles to Characterize Microbial Communities in Soil.” Applied and Environmental Microbiology 66(4): 1668-1675.
• [0228] Schultz, M., and R. B. Sator. (2000). “Probiotics and inflammatory bowel disease.” Am. J. of Gastroenterology 2000 January 95 (1 Suppl): S19-21.
• [0229] Seviour, R. J., T. Mino, et al. (2003). “The microbiology of biological phosphorus removal in activated sludge systems.” FEMS Microbiology Reviews 27: 99-127.
• [0230] Spring, S., R. Schulze, et al. (2000). “Identification and characterization of ecologically significant prokaryotes in the sediment of freshwater lakes: molecular and cultivation studies.” FEMS Microbiology Reviews 24: 573-590.
• [0231] Suzuki, M., M. S. Rappe, et al. (1998). “Kinetic bias in estimates of coastal picoplankton community structure obtained by measurements of small-subunit rRNA gene PCR amplicon length heterogeneity.” Applied and Environmental Microbiology [Appl. Environ. Microbiol.]. 64(11): 4522-4529.
• [0232] Suzuki, M. T. (1998). The Effect of Protistan Bacterivory on Bacterioplankton Community Structure: Dissertation Abstracts International Part B Science and Engineering [Diss. Abst. Int. Pt. B—Sci. & Eng.]. Vol. 59, no. 2, [np]. August 1998.
• [0233] Theron, J. and T. E. Cloete (2000). “Molecular techniques for determining microbial diversity and community structure in natural environment.” Critical Reviews in Microbiology 26: 37-57.
• [0234] Venter, J. C., S. Levy, et al. (2003). “Massive parallelism, randomness and genomic advances.” Nature genetics, 2003 March, 33 Suppl: 219-27.