WO2016049932A1

WO2016049932A1 - Biomarkers for obesity related diseases

Info

Publication number: WO2016049932A1
Application number: PCT/CN2014/088062
Authority: WO
Inventors: Qiang FENG; Dongya ZHANG; Longqing TANG; Jun Wang
Original assignee: Bgi Shenzhen Co., Limited; Bgi Shenzhen
Priority date: 2014-09-30
Filing date: 2014-09-30
Publication date: 2016-04-07
Also published as: CN107075446A; CN107075446B

Abstract

The present invention provides biomarkers and methods for predicting the risk of a disease related to microbes, in particular obesity or related diseases.

Description

BIOMARKERS FOR OBESITY RELATED DISEASES

CROSS-REFERENCE TO RELATED APPLICATION

None

FIELD

The present invention relates to biomarkers and methods for predicting the risk of a disease related to microbes, in particular obesity or related diseases.

BACKGROUND

Obesity, which is prevalent in developed countries, has increased considerably worldwide (de Carvalho Pereira et al. , 2013) . It is reported that the prevalence of overweight and obesity combined rose by 27.5％ for adults and 47.1％ for children between 1980 and 2013 in the world. The number of overweight individuals increased from 857 million in 1980, to 2.1 billion in 2013, and of these, 671 million are affected by obesity. More than 50％ of which live in ten countries, and USA has the largest number of obese individuals, followed by China (Ng et al. , 2014) .

There is a growing body of evidence suggesting that patients who are diagnosed by their physician that they are overweight are more likely to lose weight relative to those who are not diagnosed. However, low rates of physician diagnosis and advice for behavioral health risk factors related to obesity is concerning (Bleich et al. , 2011) .

In children, the diagnosis of obesity is based on age-and gender-specific body mass index (BMI) cut-points. This is in contrast to adults, in which an obesity diagnosis is based on a BMI regardless of age or gender. Unlike adults, for whom obesity diagnostic criteria are simpler, fewer obese children being accurately diagnosed for the more complicated diagnostic criteria and change in terminology for pediatric obesity (Walsh et al. , 2013) . Moreover, limitations of BMI in terms of identification of the different populations should be considered (Nevill et al. , 2006) . Therefore, waist circumference (WC) can be considered a reliable and useful tool for epidemiological studies to assess abdominal adiposity, but this measurement seems to be harder to perform (Miguel-Etayo et al. , 2014) . What’s more, regional studies of diagnosis of pediatric obesity using International Classification of Diseases, Ninth Revision (ICD-9) , National Ambulatory Medical Care Survey (NAMCS) , and National Hospital Medical Care Survey (NHAMCS) have shown relatively low sensitivity of a clinical diagnosis (Walsh et al. , 2013) .

Recent insight suggests that the human gut microbiota could play an important role in obesity. An early report, based on sequencing of amplified 16S rRNA genes, indicated a much higher ratio of Firmicutes to Bacteroidetes in faecal samples from 12 obese humans than in two lean controls (Ley et al. , 2006) . Recent observational studies using metagenomic sequencing in human obesity have demonstrated reduced bacterial diversity, a relative depletion of Bacteroidetes , and enrichment in genes involved in carbohydrate and lipid metabolism (Allin and Pedersen, 2014) . These correlative findings indicated the altered gut microbiota is a causal factor in the pathogenesis of obesity. This indicating that maybe we can use the characteristics of gut microbiota as criteria to diagnosis of obesity.

In summary, there are considerable missed opportunities and low sensitivity in the diagnosis of obesity. A more valid (less biased) assessment of overweight and/or obesity need to be developed.

SUMMARY

Embodiments of the present disclosure seek to solve at least one of the problems existing in the prior art to at least some extent.

The present invention is based on the following findings by the inventors:

Assessment and characterization of gut microbiota has become a major research area in human disease, including obesity. To carry out analysis on gut microbial content in obesity patients, the inventors carried out a protocol for a Metagenome-Wide Association Study (MGWAS) (Qin, J. et al. A metagenome-wide association study of gut microbiota in type 2 diabetes. Nature 490, 55–60 (2012) , incorporated herein by reference) based on deep shotgun sequencing of the gut microbial DNA from 158 individuals. The inventors identified and validated 54 obesity-associated gut microbes. To exploit the potential ability of obesity classification by gut microbiota, the inventors calculated probability of illness through a random forest model based on the 54 obesity-associated gut microbes. The inventors' data provide insight into the characteristics of the gut metagenome related to obesity risk, a paradigm for future studies of the pathophysiological role of the gut metagenome in other relevant disorders, and the potential usefulness for a gut-microbiota-based approach for assessment of individuals at risk of such disorders.

It is believed that 54 obesity-associated gut microbes are valuable for increasing obesity detection at earlier stages due to the following. First, the markers of the present invention are more specific and sensitive as compared with conventional markers. Second, analysis of stool promises accuracy, safety, affordability, and patient compliance. And samples of stool are transportable. Thus, the present invention relates to an in vitro method, which is comfortable and noninvasive, so people will participate in a given screening program more easily. Third, the markers of the present invention may also serve as tools for therapy monitoring in obesity patients to detect the response to therapy.

In one aspect of present disclosure, there is provided with a biomarker set for predicting a disease related to microbiota in a subject consisting of:

gut biomarker comprising Bacteroides intestinalis DSM 17393, Alistipes shahii WAL 8301, Faecalibacterium prausnitzii L2-6, Bacteroides ovatus 3_8_47FAA, Bacteroides intestinalis DSM 17393, Bacteroides sp. 1_1_30, Coprococcus eutactus ATCC 27759, Klebsiella pneumoniae 342, Veillonella sp. oral taxon 158 str. F0412, Bacteroides sp. 1_1_30, Dialister invisus DSM 15470, Bacteroides intestinalis DSM 17393, Faecalibacterium prausnitzii L2-6, Bacteroides sp. 3_1_33FAA, Faecalibacterium cf. prausnitzii KLE1255, Klebsiella oxytoca KCTC 1686, Bacteroides thetaiotaomicron VPI-5482, Bacteroides ovatus 3_8_47FAA, Haemophilus parainfluenzae T3T1, Haemophilus parainfluenzae T3T1, Faecalibacterium prausnitzii L2-6, Bacteroides intestinalis DSM 17393, Haemophilus parainfluenzae T3T1, Ruminococcus sp. 5_1_39BFAA, Coprococcus comes ATCC 27758, Eubacterium hallii DSM 3353, Dorea formicigenerans ATCC 27755, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Dorea longicatena DSM 13814, Ruminococcus obeum A2-162, Eubacterium hallii DSM 3353, Ruminococcus torques L2-14, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Ruminococcus obeum A2-162, Eubacterium hallii DSM 3353 or

microbes, with genomic DNA sequences comprising SEQ ID NO: 1 to 48497

Alternatively, the biomarker set consists of at least one of the species listed in Table 3, preferably at least 10％ , at least 20％, at least 30％, at least 40％, at least 50％, at least 60％, at least 70％, at least 80％, at least 90％, at least 100％of the species listed in Table 3.

In another aspect of present disclosure, there is provided with a biomarker set for predicting a disease related to microbiota in a subject consisting of:

a gut biomarker comprises at least a partial sequence of SEQ ID NO: 1 to 48497 as stated in Table 4..

According to embodiments of present disclosure, the disease is obesity or related disease.

In another aspect of present disclosure, there is provided with a kit for determining the gene marker set described above, comprising primers used for PCR amplification and designed according to the DNA sequecne as set forth in at least a partial sequence of SEQ ID NO: 1 to 48497.

In another aspect of present disclosure, there is provided with a kit for determining the gene marker set described above, comprising one or more probes designed according to the genes as set forth in SEQ ID NO: 1 to 48497.

In another aspect of present disclosure, there is provided with use of the gene marker set described above for predicting the risk of obesity or related disorder in a subject to be tested, comprising:

(1) collecting a samplefrom the subject to be tested；

(2) determining the relative abundance information of each biomarker of the biomarker set according to any one of claims 1 to 3 in the samples obtained in step (1) ；

(3) obtaining a probability of obesity by comparing the relative abundance information of each biomarker of subject to be tested with a training dataset using a Multivariate statistical model,

wherein the probability of obesity greater than a cutoff indicates that the subject to be tested has or is at the risk of developing the obesity or related disorder.

According to embodiments of present disclosure, the training dataset is constructed based on the relative abundance information of each biomarker of a plurality of subjects having obesity and a plurality of normal subjects using a Multivariate statistical model, alternatively the Multivariate statistical model is a randomForest model.

According to embodiments of present disclosure, the training dataset is a matrix with each row representing each biomarker of the biomarker set according to any one of claims 1 to 3, each column representing samples, each cell representing relative abundance profile of a biomarker in the sample, and sample disease status is a vectot, with 1 for obesity and 0 for control.

According to embodiments of present disclosure, the relative abundance infromation of each of Bacteroides intestinalis DSM 17393, Alistipes shahii WAL 8301, Faecalibacterium prausnitzii L2-6, Bacteroides ovatus 3_8_47FAA, Bacteroides intestinalis DSM 17393, Bacteroides sp. 1_1_30, Coprococcus eutactus ATCC 27759, Klebsiella pneumoniae 342, Veillonella sp. oral taxon 158 str. F0412, Bacteroides sp. 1_1_30, Dialister invisus DSM 15470, Bacteroides intestinalis DSM 17393, Faecalibacterium prausnitzii L2-6, Bacteroides sp. 3_1_33FAA, Faecalibacterium cf. prausnitzii KLE1255, Klebsiella oxytoca KCTC 1686, Bacteroides thetaiotaomicron VPI-5482, Bacteroides ovatus 3_8_47FAA, Haemophilus parainfluenzae T3T1, Haemophilus parainfluenzae T3T1, Faecalibacterium prausnitzii L2-6, Bacteroides intestinalis DSM 17393, Haemophilus parainfluenzae T3T1, Ruminococcus sp. 5_1_39BFAA, Coprococcus comes ATCC 27758, Eubacterium hallii DSM 3353, Dorea formicigenerans ATCC 27755, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Dorea longicatena DSM 13814, Ruminococcus obeum A2-162, Eubacterium hallii DSM 3353, Ruminococcus torques L2-14, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Ruminococcus obeum A2-162 and Eubacterium hallii DSM 3353 is obtained based on the relative abundance information of SEQ ID NO: 1 to 48497.

According to embodiments of present disclosure, the training dataset is at least one of Table 5-1、 5-2、 5-3 and 5-4, and the probability of obesity being at least 0.5 indicates that the subject to be tested has or is at the risk of developing the obesity or related disorder.

In another aspect of present disclosure, there is provided with use of the gene marker set described above for preparation of a kit for predicting the risk of obesity or related disorder in a subject to be tested, comprising:

(1) collecting a sample from the subject to be tested；

According to embodiments of present disclosure, the training dataset is at least one of Table 5-1、 5-2、 5-3 and 5-4, and the probability of obesity being at least 0.5 indicates that the subject to be tested has or is at the risk of developing the obesity or related disorder. In another aspect of present disclosure, there is provided with a method of diagnosing whether a subject has an abnormal condition related to microbiota or is at the risk of developing an abnormal condition related to microbiota, comprising:

determining the relative abundance of the biomarkers described above in a sample from the subject, and

determining whether a subject has an abnormal condition related to microbiota or is at the risk of developing an abnormal condition related to microbiota based on the relative abundance.

According to embodiments of present disclosure, the method comprises:

(1) collecting a sample from the subject to be tested；

BRIEF DISCRIPTION OF DRAWINGS

These and other aspects and advantages of the present disclosure will become apparent and more readily appreciated from the following descriptions taken in conjunction with the drawings, in which:

Fig. 1 The association analysis of Obese p-value distribution identified a disproportionate over-representation of strongly associated markers at lower P-values.

Fig. 2 The ROC were drawn by the probability of the illness in training set, and AUC＝0.9651.

Fig. 3 To validate 54 MLG markers, the inventors used random forest to set a model, then predicted the 42 samples, and the probability of the illness in each samples were calculated. The predict error rate is 8/42＝19.05％.

Fig. 4 The ROC in test set (42 samples) were drawn by the probability of the illness in test set, and AUC＝0.9188.

Fig. 5 The probability of the illness were calculated in each before and after operation samples.

Fig. 6 shows the variation of probability of the illness in each samples.

Fig. 7 shows the probability of the illness among the three groups. The probability after the operation samples were significantly lower than before.

EXAMPLES

Terms used herein have meanings as commonly understood by a person of ordinary skill in the areas relevant to the present invention. Terms such as “a” , “an” and “the” are not intended to refer to only a singular entity, but include the general class of which a specific example may be used for illustration. The terminology herein is used to describe specific embodiments of the invention, but their usage does not delimit the invention, except as outlined in the claims.

The present invention is further exemplified in the following non-limiting Examples. Unless otherwise stated, parts and percentages are by weight and degrees are Celsius. As apparent to one of ordinary skill in the art, these Examples, while indicating preferred embodiments of the invention, are given by way of illustration only, and the agents were all commercially available.

Example 1. Identifying biomarkers for evaluating obesity risk

1.1 Sample collection

Fecal samples from 158 Chinese subjects, including 78 obesity patients and 80 control subjects (training set) , were collected by Rui Jin Hospital Shanghai Jiao Tong Univeristy School of Medicine in 2012. Obesity patients were age from 18 to 30 with BMI over 25. Subjects were asked to collect fresh feces samples at hospital. Collected samples were put in sterile tubes and stored at -80℃ immediately until further analysis.

The complete ethical approval has been obtained, and all the patients gave written informed consent. The study was approved by the Institutional Review Board of Rui Jin Hospital Shanghai Jiao Tong Univeristy School of Medicine.

1.2 DNA extraction

Fecal samples were thawed on ice and DNA extraction was performed using the Qiagen QIAamp DNA Stool Mini Kit (Qiagen) according to manufacturer`s instructions. Extracts were treated with DNase-free RNase to eliminate RNA contamination. DNA quantity was determined using NanoDrop spectrophotometer, Qubit Fluorometer (with the Quant-iTTMdsDNA BR Assay Kit) and gel electrophoresis.

1.3 DNA library construction and sequencing of fecal samples

DNA library construction was performed following the manufacturer`s instruction (Illumina, insert size 350bp, read length 100bp) . The inventors used the same workflow as described previously to perform cluster generation, template hybridization, isothermal amplification, linearization, blocking and denaturation, and hybridiza-tion of the sequencing primers. The inventors constructed one paired-end (PE) library with insert size of 350 bp for each sample, followed by a high-throughput sequencing to obtain around 30 million PE reads of length 2x100bp. High-quality reads were obtained by filtering low-quality reads with ambiguous `N' bases, adapter contamination and human DNA contamination from the Illumina raw reads, and by trimming low-quality terminal bases of reads simultaneously.

The inventors totally output about 5.9 Gb per sample of fecal micbiota sequencing data (high quality clean data) (Table 1) from 158 samples (78 cases and 80 controls) on Illumina HiSeq 2000 platform.

Table 1 Summary of metagenomic data. Fourth column reports results from Wilcoxon rank-sum tests.

1.4 Metagenomic data processing and analysis

1.4.1 Reads mapping

The inventors used the updated human gut gene catalogue established in Li, J. et al. An integrated catalog of reference genes in the human gut microbiome. Nat. Biotechnol. (2014) (incorporated herein by reference) and mapped the high quality reads to it with the alignment criteria identity >＝ 90. The average reads mapping rate was shown on Table 1. This mapping rate was close to the samples in Li, J. et al. 2014, supra, which indicated that this mapping rate was sufficient for the further study. After reads mapping, the inventors derived the gene profile (9.9Mb genes) from the mapping result using the same method as Li, J. et al. 2014, supra.

Taxonomic assignment of genes. Taxonomic assignment of the predicted genes was performed using an in-house pipeline which had described in the published paper (Li, J. et al. 2014, supra) .

1.4.2 Data profile construction

Gene profile. Based on the reads mapping results, the inventors use the same method described in the published T2D paper (Qin et al. 2012, supra) to compute the relative gene abundance.

1.4.3 Analysis of factors influencing gut microbiota gene profile. The inventors used the permutational multivariate analysis of variance (PERMANOVA) to assess the effect of 6 clinical parameters, including age, sex, height, weight, BMI and obese, based on gene profile . The inventors performed the analysis using the method implemented in package ″vegan″ in R, and the permuted p-value was obtained by 10,000 times permutations. The inventors also corrected for multiple testing using ″p. adjust″ in R with Benjamini-Hochberg method to get the q-value for each test. PERMANOA identified three significant factors associated with gut microbe (based on gene profiles) (q <0.05, Table 2) . The analysis indicated weight, BMI and obese status were strong associated markers, supporting the diseases (obese) status was the major determinant influencing the composition of gut microbiota.

Table 2 PERMANOVA based on euclidean distance analysis of gene profile. The analysis was conducted to test whether clinical parameters, and obese status have significant impact on the gut microbiota with q-value <0.05.

phenotype	Df	Sums Of Sqs	Mean Sqs	F. Model	R²	Pr (>F)
phenotype	Df	Sums Of Sqs	Mean Sqs	F. Model	R²	Pr (>F)	Age	1	0.317034738	0.317034738	1.004112579	0.006395454	0.4094
Sex	1	0.377329497	0.377329497	1.196542903	0.007611763	0.1727	Age	1	0.317034738	0.317034738	1.004112579	0.006395454	0.4094
Sex	1	0.377329497	0.377329497	1.196542903	0.007611763	0.1727	Height	1	0.331409667	0.331409667	1.049947284	0.006685435	0.3291
Weight	1	0.969536515	0.969536515	3.111941857	0.019558192	1.00E-04	Height	1	0.331409667	0.331409667	1.049947284	0.006685435	0.3291
Weight	1	0.969536515	0.969536515	3.111941857	0.019558192	1.00E-04	BMI
	1	0.954186893	0.954186893	3.0617069	0.019248548	1.00E-04	BMI
	1	0.954186893	0.954186893	3.0617069	0.019248548	1.00E-04	Obese
	1	0.972185352	0.972185352	3.120613959	0.019611626	2.00E-04	Obese

1.4.4 Identification of obesity associated markers

Identification of obesity associated genes. To identify the association between the metagenomic profile and obesity, a two-tailed Wilcoxon rank-sum test was used in 9,879,897 high occurrence gene (genes that were present in less than 10 samples across all 158 samples were removed) profiles. 396, 100 gene markers were obtained, which were enriched in either case or control with p-value < 0.01, FDR＝ 3.8％ (Fig. 1) .

Estimating the false discovery rate (FDR) . Instead of a sequential p-value rejection method, the inventors applied the “q-value” method proposed in a previous study to estimate the FDR (Storey, J. D. A direct approach to false discovery rates. Journal of the Royal Statistical Society 64, 479-498 (2002) , incorporated herein by reference) .

Receiver Operator Characteristic (ROC) analysis. The inventors applied the ROC analysis to assess the performance of the obesity classification based on metagenomic markers. The inventors then used the “pROC” package in R to draw the ROC curve.

1.4.5 Construction of MLG and identification of obesity associated MLG species markers

237 MLG species based on the 396, 100 obesity associated maker genes profile. The inventors used the 396, 100 gene markers to built the metagenomic linkage group (MLG) using the same method described in the published T2D paper (Qin et al. 2012, supra) . All the 396, 100 genes were annotated by aligning these genes to the 4, 653 reference genomes in IMG v400. An MLG was assigned to a genome if more than 50％ constitutive genes were annotated to that genome, otherwise it was termed as unclassified. Total 237 MLG genomes with gene number > 100 were selected (P-value <0.01) . To estimate the relative abundance of an MLG species, the inventors estimated the average abundance of the genes of the MLG species, after removing the 5％ lowest and 5％ highest abundant genes (Qin et al. 2012, supra) .

1.5 MLG-based classifier

A random forest model (R. 2.14, randomForest4.6-7 package) (Liaw, Andy & Wiener, Matthew. Classification and Regression by randomForest, R News (2002) , Vol. 2/3 p. 18, incorporated herein by reference) was trained using the MLG abundance profile of the training cohort (158 samples) to select the optimal set of MLG markers. The model was tested on one or more testing sets and the prediction error was calculated.

About the randomForest model, using “randomForest4.6-7 package” package in R vision 2.14, input is a training dataset (namely relative abundance profiles of selected MLGs in training samples) , sample disease status (sample disease status of training samples is a vectot, 1 for obesity, 0 for control) , and a test set (just the relative abundance profiles of selected MLGs in test set) . Then the inventors used the randomForest function from randomForest package in R software to build the classification, and predict function was used to predict the test set. Output is the prediction results (probability of illness； cutoff is 0.5 and if the probability of illness ≥ 0.5, the subject is at risk of obesity)

54 MLG species marker identification. To identify 237 MLG species makers, the inventors used “randomForest4.6-7 package” package in R vision 2.14 based on the 237 obesity associated MLG species. Firstly, the inventors sorted all the 237 MLG species by the importance given by the “randomForest” method (Liaw, Andy & Wiener, Matthew. Classification and Regression by randomForest, R News (2002) , Vol. 2/3 p. 18, incorporated herein by reference) . MLG marker sets were constructed by creating incremental subsets of the top ranked MLG species, starting from 1 MLG species and ending at all 237 MLG species. For each MLG makers set, the inventors calculated the false predication ratio in the 158 samples. Finally, the 54 MLG species sets with lowest false prediction ratio were selected out as MLG species makers (Table 3 and Table 4) . Furthermore, the inventors drew the ROC curve using the OOB (out of bag) prediction probability of illness from randomForest model based on the selected MLG species markers (Table 5-0、 5-1、 5-2、 5-3、 5-4) and the area under the ROC curve (AUC) was 0.9651 in the 158 samples (Fig. 2) . At the best cutoff 0.5294, true positive rate (TPR) was 0.8625, and false positive rate (FPR) was 0.07692, indicating that the 54 MLG markers could be used to accurately classify obesity individuals.

Table 3 54 most discriminant MLGs (species markers) associated with obesity

Table 4. SEQ ID of the 54 MLG species

MLG ID	SEQ ID NO:	genes number
MLG ID	SEQ ID NO:	genes number	mlg_id: 11523	1～1762	1762
mlg_id: 127	1763～1867	105	mlg_id: 11523	1～1762	1762
mlg_id: 127	1763～1867	105	mlg_id: 3640	1868～2332	465
mlg_id: 13556	2333～2457	125	mlg_id: 3640	1868～2332	465
mlg_id: 13556	2333～2457	125	mlg_id: 10626	2458～2598	141
mlg_id: 3600	2599～3865	1267	mlg_id: 10626	2458～2598	141
mlg_id: 3600	2599～3865	1267	mlg_id: 139	3866～5780	1915
mlg_id: 3231	5781～5888	108	mlg_id: 139	3866～5780	1915
mlg_id: 3231	5781～5888	108	mlg_id: 15383	5889～6022	134
mlg_id: 6315	6023～6157	135	mlg_id: 15383	5889～6022	134
mlg_id: 6315	6023～6157	135	mlg_id: 2514	6158～8015	1858
mlg_id: 10488	8016～8816	801	mlg_id: 2514	6158～8015	1858
mlg_id: 10488	8016～8816	801	mlg_id: 141	8817～8982	166
mlg_id: 8678	8983～10386	1404	mlg_id: 141	8817～8982	166
mlg_id: 8678	8983～10386	1404	mlg_id: 844	10387～12442	2056
mlg_id: 13519	12443～12728	286	mlg_id: 844	10387～12442	2056
mlg_id: 13519	12443～12728	286	mlg_id: 12243	12729～12944	216
mlg_id: 74	12945～13047	103	mlg_id: 12243	12729～12944	216
mlg_id: 74	12945～13047	103	mlg_id: 15399	13048～13425	378
mlg_id: 252	13426～13576	151	mlg_id: 15399	13048～13425	378
mlg_id: 252	13426～13576	151	mlg_id: 12049	13577～13848	272
mlg_id: 31817	13849～14267	419	mlg_id: 12049	13577～13848	272
mlg_id: 31817	13849～14267	419	mlg_id: 8585	14268～17734	3467
mlg_id: 61	17735～17890	156	mlg_id: 8585	14268～17734	3467
mlg_id: 61	17735～17890	156	mlg_id: 10628	17891～18031	141
mlg_id: 1180	18032～18234	203	mlg_id: 10628	17891～18031	141
mlg_id: 1180	18032～18234	203	mlg_id: 115	18235～19185	951
mlg_id: 11108	19186～20157	972	mlg_id: 115	18235～19185	951
mlg_id: 11108	19186～20157	972	mlg_id: 994	20158～20261	104
mlg_id: 13037	20262～20391	130	mlg_id: 994	20158～20261	104
mlg_id: 13037	20262～20391	130	mlg_id: 20358	20392～20644	253
mlg_id: 10879	20645～20749	105	mlg_id: 20358	20392～20644	253
mlg_id: 10879	20645～20749	105	mlg_id: 2728	20750～21085	336
mlg_id: 605	21086～23716	2631	mlg_id: 2728	20750～21085	336
mlg_id: 605	21086～23716	2631	mlg_id: 10526	23717～24188	472
mlg_id: 2060	24189～28313	4125	mlg_id: 10526	23717～24188	472
mlg_id: 2060	24189～28313	4125	mlg_id: 4035	28314～28498	185
mlg_id: 7361	28499～30352	1854	mlg_id: 4035	28314～28498	185
mlg_id: 7361	28499～30352	1854	mlg_id: 15056	30353～30659	307
mlg_id: 12221	30660～33317	2658	mlg_id: 15056	30353～30659	307
mlg_id: 12221	30660～33317	2658	mlg_id: 8423	33318～33423	106

mlg_id: 4404	33424～35102	1679
mlg_id: 4404	33424～35102	1679	mlg_id: 6546	35103～35487	385
mlg_id: 11335	35488～35965	478	mlg_id: 6546	35103～35487	385
mlg_id: 11335	35488～35965	478	mlg_id: 3647	35966～38270	2305
mlg_id: 94	38271～38506	236	mlg_id: 3647	35966～38270	2305
mlg_id: 94	38271～38506	236	mlg_id: 13978	38507～40389	1883
mlg_id: 12929	40390～40550	161	mlg_id: 13978	38507～40389	1883
mlg_id: 12929	40390～40550	161	mlg_id: 3665	40551～40691	141
mlg_id: 12935	40692～41880	1189	mlg_id: 3665	40551～40691	141
mlg_id: 12935	40692～41880	1189	mlg_id: 31770	41881～42212	332
mlg_id: 5711	42213～44891	2679	mlg_id: 31770	41881～42212	332
mlg_id: 5711	42213～44891	2679	mlg_id: 15213	44892～47552	2661
mlg_id: 2245	47553～48497	945	mlg_id: 15213	44892～47552	2661

Table 5-0 Prediction results of 54 MLGs in 158 samples

Example 2. Validating the 54 MLG biomarkers in 42 samples (test set)

The inventors validated the discriminatory power of the obesity classifier using another new independent study group, including 17 obesity patients and 25 non- obesity controls that were also collected in Rui Jin Hospital Shanghai Jiao Tong Univeristy School of Medicine .

For each sample, DNA was extracted and a DNA library was constructed followed by high throughput sequencing as described in Example 1. The inventors calculated the gene abundance profile for these samples using the same method as described in Qin et al. 2012, supra. Then the gene relative abundance of each of the markers as set forth in SEQ ID NOs: 1-48497 was determined. The inventors estimated the relative abundance of a MLG in all samples by using the relative abundance values of genes from this MLG (Qin et al. 2012, supra) .

To validate 54 MLG markers, the inventors used random forest to set a model. About the randomForest model, using “randomForest4.6-7 package” package in R vision 2.14, input is a training dataset (namely relative abundance profiles of selected MLGs in training samples, Table 5-1、 5-2、 5-3、 5-4) , sample disease status (sample disease status of training samples is a vectot,1 for obesity, 0 for control) , and a test set ( just the relative abundance profiles of selected MLGs in test set) . Then the inventors used the randomForest function from randomForest package in R software to build the classification, and predict function was used to predict the test set. Output is the prediction results (probability of illness； cutoff is 0.5 and if the probability of illness ≥ 0.5, the subject is at risk of obesity) .

Then the inventors predicted the 42 samples, and the probability of the illness in each samples were calculated (Table 6) . The predict error rate is 8/42＝19.05％ (Fig. 3) . And most of obesity patients (16/17) were diagnosed as obesity correctly. Also, the ROC in test set were drawn by the probability of the illness in test set, and AUC＝0.9188 (Fig. 4) , validating that the 54 MLG markers could be used to accurately classify obesity individuals. At the best cutoff 0.592, true positive rate (TPR) was 0.8824, and false positive rate (FPR) was 0.16.

Table 6 Prediction results of 54 MLGs in 42 samples

Example 3. Validating the 54 MLG biomarkers in 22 samples (test set)

The inventors validated the discriminatory power of the obesity classifier using another test set samples (Table 7) , including 9 case samples and 13 control samples (5 samples after operation 1 month and 8 samples after operation 3 month) that were also collected in Rui Jin Hospital Shanghai Jiao Tong Univeristy School of Medicine . Case means before operation samples, control means after operation 1 month and 3 month.

Table 7 22 samples information

Before	1-M	3-M
Before	1-M	3-M	DB62	DB-S1-62	DB-S3-62
DB67		DB-S3-67	DB62	DB-S1-62	DB-S3-62
DB67		DB-S3-67	DB68	DB-S1-68	DB-S3-68
DB78		DB-S3-78	DB68	DB-S1-68	DB-S3-68
DB78		DB-S3-78	DB85	DB-S1-85
DB124	DB-S1-124	DB-S3-124	DB85	DB-S1-85
DB124	DB-S1-124	DB-S3-124	DB125	DB-S1-125	DB-S3-125
DB126		DB-S3-126	DB125	DB-S1-125	DB-S3-125
DB126		DB-S3-126	DB01		DB-S3-01

*Before: before the operation； 1-M: operation after one month； 3-M: operation after three month.

The probability of the illness were calculated in each before and after operation samples. It shows that after the operation, the probability of the illness are becoming low (Fig. 5, Table 8) . The error rate was 9％ (2/22) .

To know the variation of each samples, the probability of the illness were showed in Fig. 6 . And among the three groups, the probability of illness after the operation samples were significantly lower than before (Fig. 7, Table 9) , validating that the 54 MLG markers could be used to accurately classify obesity individuals.

Table 8 Prediction results of 54 MLGs in 22 samples

Samples (DB： obesity)	probability of obesity
Samples (DB： obesity)	probability of obesity	DB85	0.882
DB68	0.826	DB85	0.882
DB68	0.826	DB78	0.822
DB01	0.81	DB78	0.822
DB01	0.81	DB125	0.776
DB126	0.732	DB125	0.776
DB126	0.732	DB62	0.662
DB.S1.68	0.654	DB62	0.662
DB.S1.68	0.654	DB124	0.606
DB67	0.592	DB124	0.606
DB67	0.592	DB.S1.62	0.5
DB.S3.126	0.496	DB.S1.62	0.5
DB.S3.126	0.496	DB.S1.125	0.476
DB.S3.125	0.458	DB.S1.125	0.476
DB.S3.125	0.458	DB.S3.62	0.412
DB.S1.85	0.396	DB.S3.62	0.412
DB.S1.85	0.396	DB.S3.124	0.368
DB.S3.67	0.354	DB.S3.124	0.368
DB.S3.67	0.354	DB.S1.124	0.34
DB.S3.01	0.336	DB.S1.124	0.34
DB.S3.01	0.336	DB.S3.68	0.256
DB.S3.78	0.252	DB.S3.68	0.256

Table 9 P value of probability of illness in three groups

	Before	1-M	3-M
	Before	1-M	3-M	Before	0.0007437	4.114e-05
1-M			0.8584	Before	0.0007437	4.114e-05

Thus the inventors have identified and validated 54 obesity-associated gut microbes by a random forest model based on obesity-associated genes markers. And the inventors have constructed a method to evaluate the risk of obesity disease based on these 54 obesity-associated gut microbes.

Although explanatory embodiments have been shown and described, it would be appreciated by those skilled in the art that the above embodiments can not be construed to limit the present disclosure, and changes, alternatives, and modifications can be made in the embodiments without departing from spirit, principles and scope of the present disclosure.

Claims

A biomarker set for predicting a disease related to microbiota in a subject consisting of:

gut biomarker comprising Bacteroides intestinalis DSM 17393, Alistipes shahii WAL 8301, Faecalibacterium prausnitzii L2-6, Bacteroides ovatus 3_8_47FAA, Bacteroides intestinalis DSM 17393, Bacteroides sp. 1_1_30, Coprococcus eutactus ATCC 27759, Klebsiella pneumoniae 342, Veillonella sp. oral taxon 158 str. F0412, Bacteroides sp. 1_1_30, Dialister invisus DSM 15470, Bacteroides intestinalis DSM 17393, Faecalibacterium prausnitzii L2-6, Bacteroides sp. 3_1_33FAA, Faecalibacterium cf. prausnitzii KLE1255, Klebsiella oxytoca KCTC 1686, Bacteroides thetaiotaomicron VPI-5482, Bacteroides ovatus 3_8_47FAA, Haemophilus parainfluenzae T3T1, Haemophilus parainfluenzae T3T1, Faecalibacterium prausnitzii L2-6, Bacteroides intestinalis DSM 17393, Haemophilus parainfluenzae T3T1, Ruminococcus sp. 5_1_39BFAA, Coprococcus comes ATCC 27758, Eubacterium hallii DSM 3353, Dorea formicigenerans ATCC 27755, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Dorea longicatena DSM 13814, Ruminococcus obeum A2-162, Eubacterium hallii DSM 3353, Ruminococcus torques L2-14, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Ruminococcus obeum A2-162, Eubacterium hallii DSM 3353, or

microbes, with genomic DNA sequences comprising SEQ ID NO: 1 to 48497.
A biomarker set for predicting a disease related to microbiota in a subject according to claim 1 consisting of:

a gut biomarker comprises at least a partial sequence of SEQ ID NO: 1 to 48497.
The biomarker set for predicting a disease related to microbiota in a subject of claim 1 or 2, wherein the disease is obesity or related disease.
A kit for determining the gene marker set of any one of claims 1 to 3, comprising primers used for PCR amplification and designed according to the DNA sequecne as set forth in claim 2.
A kit for determining the gene marker set of any one of claims 1 to 3, comprising one or more probes designed according to the genes as set forth in claim 2.
Use of the gene marker set of any one of claims 1 to 3 for predicting the risk of obesity or related disorder in a subject to be tested, comprising:

(1) collecting a sample from the subject to be tested；

(2) determining the relative abundance information of each biomarker of the biomarker set according to any one of claims 1 to 3 in the samples obtained in step (1) ；

(3) obtaining a probability of obesity by comparing the relative abundance information of each biomarker of subject to be tested with a training dataset using a Multivariate statistical model,

wherein the probability of obesity greater than a cutoff indicates that the subject to be tested has or is at the risk of developing the obesity or related disorder.
The use of claim 6, wherein the training dataset is constructed based on the relative abundance information of each biomarker of a plurality of subjects having obesity and a plurality of normal subjects using a Multivariate statistical model, alternatively the Multivariate statistical model is a randomForest model.
The use of claim 7, wherein the training dataset is a matrix with each row representing each biomarker of the biomarker set according to any one of claims 1 to 3, each column representing samples, each cell representing relative abundance profile of a biomarker in the sample, and sample disease status is a vectot, with 1 for obesity and 0 for control.
The use of claim 7, wherein the relative abundance infromation of each of Bacteroides intestinalis DSM 17393, Alistipes shahii WAL 8301, Faecalibacterium prausnitzii L2-6, Bacteroides ovatus 3_8_47FAA, Bacteroides intestinalis DSM 17393, Bacteroides sp. 1_1_30, Coprococcus eutactus ATCC 27759, Klebsiella pneumoniae 342, Veillonella sp. oral taxon 158 str. F0412, Bacteroides sp. 1_1_30, Dialister invisus DSM 15470, Bacteroides intestinalis DSM 17393, Faecalibacterium prausnitzii L2-6, Bacteroides sp. 3_1_33FAA, Faecalibacterium cf. prausnitzii KLE1255, Klebsiella oxytoca KCTC 1686, Bacteroides thetaiotaomicron VPI-5482, Bacteroides ovatus 3_8_47FAA, Haemophilus parainfluenzae T3T1, Haemophilus parainfluenzae T3T1, Faecalibacterium prausnitzii L2-6, Bacteroides intestinalis DSM 17393, Haemophilus parainfluenzae T3T1, Ruminococcus sp. 5_1_39BFAA, Coprococcus comes ATCC 27758, Eubacterium hallii DSM 3353, Dorea formicigenerans ATCC 27755, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Dorea longicatena DSM 13814, Ruminococcus obeum A2-162, Eubacterium hallii DSM 3353, Ruminococcus torques L2-14, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Ruminococcus obeum A2-162 and Eubacterium hallii DSM 3353 is obtained based on the relative abundance information of SEQ ID NO: 1 to 48497.
The use of claim 7, wherein the training dataset is at least one of Table 5-1、 5-2、 5-3 and 5-4, and the probability of obesity being at least 0.5 indicates that the subject to be tested has or is at the risk of developing the obesity or related disorder.
Use of the gene marker set of any one of claims 1 to 3 for preparation of a kit for predicting the risk of obesity or related disorder in a subject to be tested, comprising:

(1) collecting a sample from the subject to be tested；

(2) determining the relative abundance information of each biomarker of the biomarker set according to any one of claims 1 to 3 in the samples obtained in step (1) ；

(3) obtaining a probability of obesity by comparing the relative abundance information of each biomarker of subject to be tested with a training dataset using a Multivariate statistical model,

wherein the probability of obesity greater than a cutoff indicates that the subject to be tested has or is at the risk of developing the obesity or related disorder.
The use of claim 11, wherein the training dataset is constructed based on the relative abundance information of each biomarker of a plurality of subjects having obesity and a plurality of normal subjects using a Multivariate statistical model, alternatively the Multivariate statistical model is a randomForest model.
The use of claim 12, wherein the training dataset is a matrix with each row representing each biomarker of the biomarker set according to any one of claims 1 to 3, each column representing samples, each cell representing relative abundance profile of a biomarker in the sample, and sample disease status is a vectot, with 1 for obesity and 0 for control.
The use of claim 12, wherein the relative abundance infromation of each of Bacteroides intestinalis DSM 17393, Alistipes shahii WAL 8301, Faecalibacterium prausnitzii L2-6, Bacteroides ovatus 3_8_47FAA, Bacteroides intestinalis DSM 17393, Bacteroides sp. 1_1_30, Coprococcus eutactus ATCC 27759, Klebsiella pneumoniae 342, Veillonella sp. oral taxon 158 str. F0412, Bacteroides sp. 1_1_30, Dialister invisus DSM 15470, Bacteroides intestinalis DSM 17393, Faecalibacterium prausnitzii L2-6, Bacteroides sp. 3_1_33FAA, Faecalibacterium cf. prausnitzii KLE1255, Klebsiella oxytoca KCTC 1686, Bacteroides thetaiotaomicron VPI-5482, Bacteroides ovatus 3_8_47FAA, Haemophilus parainfluenzae T3T1, Haemophilus parainfluenzae T3T1, Faecalibacterium prausnitzii L2-6, Bacteroides intestinalis DSM 17393, Haemophilus parainfluenzae T3T1, Ruminococcus sp. 5_1_39BFAA, Coprococcus comes ATCC 27758, Eubacterium hallii DSM 3353, Dorea formicigenerans ATCC 27755, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Dorea longicatena DSM 13814, Ruminococcus obeum A2-162, Eubacterium hallii DSM 3353, Ruminococcus torques L2-14, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Ruminococcus obeum A2-162 and Eubacterium hallii DSM 3353 is obtained based on the relative abundance information of SEQ ID NO: 1 to 48497.
The use of claim 12, wherein the training dataset is at least one of Table 5-1、 5-2、 5-3 and 5-4, and the probability of obesity being at least 0.5 indicates that the subject to be tested has or is at the risk of developing the obesity or related disorder.
A method of diagnosing whether a subject has an abnormal condition related to microbiota or is at the risk of developing an abnormal condition related to microbiota, comprising:

determining the relative abundance of the biomarkers of any one of claims 1 to 3 in a sample from the subject, and

determining whether a subject has an abnormal condition related to microbiota or is at the risk of developing an abnormal condition related to microbiota based on the relative abundance.
The method according to claim 16, comprising:

(1) collecting a sample from the subject to be tested；

(2) determining the relative abundance information of each biomarker of the biomarker set according to any one of claims 1 to 3 in the samples obtained in step (1) ；

(3) obtaining a probability of obesity by comparing the relative abundance information of each biomarker of subject to be tested with a training dataset using a Multivariate statistical model,

wherein the probability of obesity greater than a cutoff indicates that the subject to be tested has or is at the risk of developing the obesity or related disorder.
The method of claim 17, wherein the training dataset is constructed based on the relative abundance information of each biomarker of a plurality of subjects having obesity and a plurality of normal subjects using a Multivariate statistical model, alternatively the Multivariate statistical model is a randomForest model.
The method of claim 18, wherein the training dataset is a matrix with each row representing each biomarker of the biomarker set according to any one of claims 1 to 3, each column representing samples, each cell representing relative abundance profile of a biomarker in the sample, and sample disease status is a vectot, with 1 for obesity and 0 for control.
The method of claim 18, wherein the relative abundance infromation of each of Bacteroides intestinalis DSM 17393, Alistipes shahii WAL 8301, Faecalibacterium prausnitzii L2-6, Bacteroides ovatus 3_8_47FAA, Bacteroides intestinalis DSM 17393, Bacteroides sp. 1_1_30, Coprococcus eutactus ATCC 27759, Klebsiella pneumoniae 342, Veillonella sp. oral taxon 158 str. F0412, Bacteroides sp. 1_1_30, Dialister invisus DSM 15470, Bacteroides intestinalis DSM 17393, Faecalibacterium prausnitzii L2-6, Bacteroides sp. 3_1_33FAA, Faecalibacterium cf. prausnitzii KLE1255, Klebsiella oxytoca KCTC 1686, Bacteroides thetaiotaomicron VPI-5482, Bacteroides ovatus 3_8_47FAA, Haemophilus parainfluenzae T3T1, Haemophilus parainfluenzae T3T1, Faecalibacterium prausnitzii L2-6, Bacteroides intestinalis DSM 17393, Haemophilus parainfluenzae T3T1, Ruminococcus sp. 5_1_39BFAA, Coprococcus comes ATCC 27758, Eubacterium hallii DSM 3353, Dorea formicigenerans ATCC 27755, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Dorea longicatena DSM 13814, Ruminococcus obeum A2-162, Eubacterium hallii DSM 3353, Ruminococcus torques L2-14, Dorea longicatena DSM 13814, Collinsella aerofaciens ATCC 25986, Ruminococcus obeum A2-162 and Eubacterium hallii DSM 3353 is obtained based on the relative abundance information of SEQ ID NO: 1 to 48497.
The method of claim 18, wherein the training dataset is at least one of Table 5-1、 5-2、 5-3 and 5-4, and the probability of obesity being at least 0.5 indicates that the subject to be tested has or is at the risk of developing the obesity or related disorder.