data in the form of OTU count tables. OTU tables and associated data are nlme: Ime function in R. The subject was included as a random effect for both the intercept and the slope of the estimated fit. In all cases, the analysis of trends with gestational time was repeated with multiple measures of stability Study Population and Sampling Procedures. Pregnant women age 18 y or older and diversity resenting to the obstetrical clinics of the Lucille Packard Childrens Hospital selected. The first consisted of 40 women, 11 of whom de.(by subject) two-class testing for differential relative abundance and to delivery. OTUs were considered sig. until delivery and monthly from the time of delivery for up to 12 mo For was <0.1 and if the estimated fold change was >1.5 or <l/. ted Pvalue vere self-collected by participants weekly from the time of study enrollment nificantly differentially abundant between dasses if their adj further information, see S/ Appendix, S/ Methods and Tables S1 and S2 Clustering into CSTs. First, the Bray-Curtis distance between all samples was NA Extraction, 16S rDNA Amplification, and Amplicon Se ng. After ex calculated. This distance matrix was denoised by extracting the most sig. action of genomic DNA, the V3-V5 region of the 16S rRNA gene was PCR. nificant Principal Coordinates Analysis(PCoA)eigenvectors The partitioning he ied from 3, 767 vaginal, stool, saliva, and tooth/gum specimens from around medoids algorithm (pam in R)was applied to these PCoA distances in the first group and analyzed with pyrosequencing as a The number of clusters(k 5)was determined from the gap statistic (S/ iscovery dataset. The V4 reg rom the en in the second group and analyzed with Illumina-based ition Rates. Before vaginal CST transition rates were estimated, the dataset was restricted to pairs of consecutive samples quence Filtering, OTU Clustering, and Chimera Removal. The two datasets collected 4-10 d apart. This set of 652 paired samples had time-separations ere analyzed similarly but separately. Initial quality processing was per- of 4-10 d(mean 6.96 d)and a firs stile, third- quartile, and median of ormed using QIIME version 1.7(qiime. org), followed by global trimming to 7 d. The 1-wk transition rate was quantified as the maximum-likelihood 350 bases and filtering. Raw Illumina read-pairs were quality filtered, estimate from this set of paired samples merged, and de-multiplexed. OTU clustering at a 97% sequence identity hreshold was performed using the UPARSE algorithm After OTU clustering, moval of chimeric sequences was performed in a stringent two-step pr nd Cele quaintance cess. See SI Appendix, SI Methods for details. obstetrical clinics and the labor and delivery unit of Lucille Packard Children s and Statistical Ar erformedusing'r'languageandenvironment[r2014,httpsj//www.r.shoponMetagenomicsThisresearchwassupportedbythemArchofDi project. org version 3.1.1 Appendix, S/ Methods for details. aluating Trends with Gestational Time. Changes in stability and diversity ematical Sciences Grant 1162538(to S.P. H and B.C), and the Thomas Cand over the course of pregnancy were evaluated by LME modeling using the Joan M. Merigan Endowment at Stanford University( D AR ne Cel/ 159(4): 789-799. 19. Verstraelen H, et al. (2009) Longit suggests that L crispatus promotes the stability of the normal vaginal micro- iners are more conducive to the occurrence of 3. Dethlefsen L, McFall 007) An ecological and evolutionary Nat 4. Blencowe H, et al. (2013)Born too soon: The global epidemiology of 15 million subgingival plaque. J Periodontol 52(10): 599-602. effect of female sex hormones 21. Kornman KS Loesch 1980)The subgingival microbial flora during pregnancy. sher SJ (2014)Preterm labor: One syndrome, many J Periodontal Res 15(2): 111-122. 5(6198)760-765 22. Adriaens LM, Alessandri R, SporriS, Lang NP, Perss al. (2008) Microbial prevalence, diversity and abundance in amniotic fluid ta? J Periodontol 80(1): 72-8 investigation. PLos One 3(8]e305 23. Collado MC lsolauri E, Laitinen K, Salminen S(2008) Distinct composition of gut nancy in overweight and normal-weight wom C et al.(2004) identificati nm d Reprod immune 24. Koren O, et al. (2012)He ture labor. Am J Perinatol 21(6): 319-323. 50()470480. 9. etiologic agents of intra-amniotic inflammation leading to preterm birth yna as 25. Gajer P, et al.(2012) Temporal dynamics of the human vaginal microbiota.Sd W. Shen T Chun Med4(132)132ra5 26. Siqueira FM, et al. (2007) Intrauterine growth restriction, low birth weight, and eterm birth: Adverse pregnancy outcomes and their association with maternal 11.Ravel 1, et al. (2011) Vaginal microbiome of reproductive-age women. Proc Natl Acad 27. Brotman RM, Ravel ), Cone RA, Zenilman JM (2010) Rapid fluctuation of the vaginal nal Infections and Prematurity Study Group(1995)Associ- 28. Chaban B, et al. (2014)Characterization of Iota of healthy Cana- of a low-birth-weight infant. 29. Hickey Ri, et al. (2013) 13. Leitich H, et al. (2003)B vaginosis as a risk factor for preterm delivery A meta- 06695-704; discussion70469 neco/ 189(1) 30. Santiago Gl croflora 14. Aagaard k, et al. (2012)A metagenomic approach to characterization of the vaginal 31. Srinivasan S, et al. (2010) Temporal lity of human vaginal bacteria and 15. Hernandez-Ro C et al. (2011)Vaginal microbiota of healthy pregnant Mexi- lationship with bacterial vaginosis. PLos One 5(4)e10197 ci- 32. Thoma ME, et al. (2011)Lor ect Dis Obstet Gynecol 16. Hyman RW, et al. (2014)Diversity of the vaginal microbiome correlates with preterm 17. Romero R, et al. (2014) The vaginal microbiota of pregnant women who subsequently ne84:e61217. ve spontaneous preterm labor and delivery and those with a normal deliver on of fold change and dis- 8. Romero R, et al. (2014) The composition and stability of the vaginal microbiota of normal 35. Tibshirani R, w data with DESeq2. Genome Bio/ 15(12): 550. Iter G, Hastie T(2001) Estimating the number of clusters in a data set via the gap statistic. J R Stat Soc, B 63: 411-42 PNAS I September 1, 2015 I voL. 112 I no 35 11065data in the form of OTU count tables. OTU tables and associated data are provided in Datasets S1 and S2. Study Population and Sampling Procedures. Pregnant women age 18 y or older presenting to the obstetrical clinics of the Lucille Packard Children’s Hospital at Stanford University for prenatal care were enrolled, and two case-control groups were selected. The first consisted of 40 women, 11 of whom de￾livered preterm; the second consisted of nine women, four of whom de￾livered preterm. Specimens from the vagina, stool, saliva, and tooth/gum were self-collected by participants weekly from the time of study enrollment until delivery and monthly from the time of delivery for up to 12 mo. For further information, see SI Appendix, SI Methods and Tables S1 and S2. DNA Extraction, 16S rDNA Amplification, and Amplicon Sequencing. After ex￾traction of genomic DNA, the V3–V5 region of the 16S rRNA gene was PCR￾amplified from 3,767 vaginal, stool, saliva, and tooth/gum specimens from the 40 women in the first group and analyzed with pyrosequencing as a discovery dataset. The V4 region was amplified from 246 vaginal specimens from the nine women in the second group and analyzed with Illumina-based sequencing as a validation dataset. Sequence Filtering, OTU Clustering, and Chimera Removal. The two datasets were analyzed similarly but separately. Initial quality processing was per￾formed using QIIME version 1.7 (qiime.org), followed by global trimming to 350 bases and filtering. Raw Illumina read-pairs were quality filtered, merged, and de-multiplexed. OTU clustering at a 97% sequence identity threshold was performed using the UPARSE algorithm. After OTU clustering, removal of chimeric sequences was performed in a stringent two-step pro￾cess. See SI Appendix, SI Methods for details. Bioinformatics Approach and Statistical Analysis. Statistical analyses were performed using ‘R’ language and environment [R 2014, https://www.r￾project.org version 3.1.1, including phyloseq (33)] and other packages. See SI Appendix, SI Methods for details. Evaluating Trends with Gestational Time. Changes in stability and diversity over the course of pregnancy were evaluated by LME modeling using the nlme::lme function in R. The subject was included as a random effect for both the intercept and the slope of the estimated fit. In all cases, the analysis of trends with gestational time was repeated with multiple measures of stability and diversity. Evaluating Differential Abundance. DESeq2 (34) was used to perform paired (by subject) two-class testing for differential relative abundance and to perform unpaired testing for associations between differential abundance within CST 4 specimens and preterm delivery. OTUs were considered sig￾nificantly differentially abundant between classes if their adjusted P value was <0.1 and if the estimated fold change was >1.5 or <1/1.5. Clustering into CSTs. First, the Bray–Curtis distance between all samples was calculated. This distance matrix was denoised by extracting the most sig￾nificant Principal Coordinates Analysis (PCoA) eigenvectors. The partitioning around medoids algorithm (pam in R) was applied to these PCoA distances. The number of clusters (k = 5) was determined from the gap statistic (SI Appendix, Fig. S14) (35). Estimating Vaginal CST Transition Rates. Before vaginal CST transition rates were estimated, the dataset was restricted to pairs of consecutive samples collected 4–10 d apart. This set of 652 paired samples had time-separations of 4–10 d (mean 6.96 d) and a first-quartile, third-quartile, and median of 7 d. The 1-wk transition rate was quantified as the maximum-likelihood estimate from this set of paired samples. ACKNOWLEDGMENTS. We thank the study participants and Cele Quaintance, Nick Scalfone, Chris Paiji, Kat Sanders, Katie Cumnock, Ana Laborde, March of Dimes Prematurity Research Center study coordinators, the nursing staff in the obstetrical clinics and the labor and delivery unit of Lucille Packard Children’s Hospital, and Stephen Cornell and participants of the Newton Institute Work￾shop on Metagenomics. This research was supported by the March of Dimes Prematurity Research Center at Stanford University, NIH National Center for Advancing Translational Science Clinical and Translational Science Award UL1 TR001085, the Stanford Child Health Research Institute, NIH Grant R01 GM086884 (to S.P.H. and B.J.C.), National Science Foundation Division of Math￾ematical Sciences Grant 1162538 (to S.P.H. and B.J.C.), and the Thomas C. and Joan M. Merigan Endowment at Stanford University (D.A.R.). 1. Goodrich JK, et al. (2014) Human genetics shape the gut microbiome. Cell 159(4):789–799. 2. Ursell LK, et al. (2012) The interpersonal and intrapersonal diversity of human-asso￾ciated microbiota in key body sites. J Allergy Clin Immunol 129(5):1204–1208. 3. Dethlefsen L, McFall-Ngai M, Relman DA (2007) An ecological and evolutionary per￾spective on human-microbe mutualism and disease. Nature 449(7164):811–818. 4. Blencowe H, et al. (2013) Born too soon: The global epidemiology of 15 million preterm births. Reprod Health 10(Suppl 1):S2. 5. Romero R, Dey SK, Fisher SJ (2014) Preterm labor: One syndrome, many causes. Science 345(6198):760–765. 6. DiGiulio DB, et al. (2008) Microbial prevalence, diversity and abundance in amniotic fluid during preterm labor: A molecular and culture-based investigation. PLoS One 3(8):e3056. 7. DiGiulio DB, et al. (2010) Prevalence and diversity of microbes in the amniotic fluid, the fetal inflammatory response, and pregnancy outcome in women with preterm pre-labor rupture of membranes. Am J Reprod Immunol 64(1):38–57. 8. Gardella C, et al. (2004) Identification and sequencing of bacterial rDNAs in culture-nega￾tive amniotic fluid from women in premature labor. Am J Perinatol 21(6):319–323. 9. Han YW, Shen T, Chung P, Buhimschi IA, Buhimschi CS (2009) Uncultivated bacteria as etiologic agents of intra-amniotic inflammation leading to preterm birth. J Clin Microbiol 47(1):38–47. 10. Fredricks DN, Fiedler TL, Marrazzo JM (2005) Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med 353(18):1899–1911. 11. Ravel J, et al. (2011) Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci USA 108(Suppl 1):4680–4687. 12. Hillier SL, et al.; The Vaginal Infections and Prematurity Study Group (1995) Associ￾ation between bacterial vaginosis and preterm delivery of a low-birth-weight infant. N Engl J Med 333(26):1737–1742. 13. Leitich H, et al. (2003) Bacterial vaginosis as a risk factor for preterm delivery: A meta￾analysis. Am J Obstet Gynecol 189(1):139–147. 14. Aagaard K, et al. (2012) A metagenomic approach to characterization of the vaginal microbiome signature in pregnancy. PLoS One 7(6):e36466. 15. Hernández-Rodríguez C, et al. (2011) Vaginal microbiota of healthy pregnant Mexi￾can women is constituted by four Lactobacillus species and several vaginosis-associ￾ated bacteria. Infect Dis Obstet Gynecol 2011:851485. 16. Hyman RW, et al. (2014) Diversity of the vaginal microbiome correlates with preterm birth. Reprod Sci 21(1):32–40. 17. Romero R, et al. (2014) The vaginal microbiota of pregnant women who subsequently have spontaneous preterm labor and delivery and those with a normal delivery at term. Microbiome 2:18. 18. Romero R, et al. (2014) The composition and stability of the vaginal microbiota of normal pregnant women is different from that of non-pregnant women. Microbiome 2(1):4. 19. Verstraelen H, et al. (2009) Longitudinal analysis of the vaginal microflora in preg￾nancy suggests that L. crispatus promotes the stability of the normal vaginal micro￾flora and that L. gasseri and/or L. iners are more conducive to the occurrence of abnormal vaginal microflora. BMC Microbiol 9:116. 20. Jensen J, Liljemark W, Bloomquist C (1981) The effect of female sex hormones on subgingival plaque. J Periodontol 52(10):599–602. 21. Kornman KS, Loesche WJ (1980) The subgingival microbial flora during pregnancy. J Periodontal Res 15(2):111–122. 22. Adriaens LM, Alessandri R, Spörri S, Lang NP, Persson GR (2009) Does pregnancy have an impact on the subgingival microbiota? J Periodontol 80(1):72–81. 23. Collado MC, Isolauri E, Laitinen K, Salminen S (2008) Distinct composition of gut microbiota during pregnancy in overweight and normal-weight women. Am J Clin Nutr 88(4):894–899. 24. Koren O, et al. (2012) Host remodeling of the gut microbiome and metabolic changes during pregnancy. Cell 150(3):470–480. 25. Gajer P, et al. (2012) Temporal dynamics of the human vaginal microbiota. Sci Transl Med 4(132):132ra52. 26. Siqueira FM, et al. (2007) Intrauterine growth restriction, low birth weight, and preterm birth: Adverse pregnancy outcomes and their association with maternal periodontitis. J Periodontol 78(12):2266–2276. 27. Brotman RM, Ravel J, Cone RA, Zenilman JM (2010) Rapid fluctuation of the vaginal microbiota measured by Gram stain analysis. Sex Transm Infect 86(4):297–302. 28. Chaban B, et al. (2014) Characterization of the vaginal microbiota of healthy Cana￾dian women through the menstrual cycle. Microbiome 2:23. 29. Hickey RJ, et al. (2013) Effects of tampons and menses on the composition and diversity of vaginal microbial communities over time. BJOG 120(6):695–704; discussion 704–696. 30. Santiago GL, et al. (2011) Longitudinal study of the dynamics of vaginal microflora during two consecutive menstrual cycles. PLoS One 6(11):e28180. 31. Srinivasan S, et al. (2010) Temporal variability of human vaginal bacteria and re￾lationship with bacterial vaginosis. PLoS One 5(4):e10197. 32. Thoma ME, et al. (2011) Longitudinal changes in vaginal microbiota composition assessed by gram stain among never sexually active pre- and postmenarcheal ado￾lescents in Rakai, Uganda. J Pediatr Adolesc Gynecol 24(1):42–47. 33. McMurdie PJ, Holmes S (2013) phyloseq: An R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One 8(4):e61217. 34. Love MI, Huber W, Anders S (2014) Moderated estimation of fold change and dis￾persion for RNA-seq data with DESeq2. Genome Biol 15(12):550. 35. Tibshirani R, Walter G, Hastie T (2001) Estimating the number of clusters in a data set via the gap statistic. J R Stat Soc, B 63:411–423. DiGiulio et al. PNAS | September 1, 2015 | vol. 112 | no. 35 | 11065 MICROBIOLOGY