The dynamics of a family’s gut microbiota reveal variations on a theme
© Schloss et al.; licensee BioMed Central Ltd. 2014
Received: 6 March 2014
Accepted: 14 June 2014
Published: 21 July 2014
It is clear that the structure and function of the human microbiota has significant impact on maintenance of health and yet the factors that give rise to an adult microbiota are poorly understood. A combination of genetics, diet, environment, and life history are all thought to impact the development of the gut microbiome. Here we study a chronosequence of the gut microbiota found in eight individuals from a family consisting of two parents and six children ranging in age from two months to ten years old.
Using 16S rRNA gene and metagenomic shotgun sequence data, it was possible to distinguish the family from a cohort of normal individuals living in the same geographic region and to differentiate each family member. Interestingly, there was a significant core membership to the family members’ microbiota where the abundance of this core accounted for the differences between individuals. It was clear that the introduction of solids represents a significant transition in the development of a mature microbiota. This transition was associated with increased diversity, decreased stability, and the colonization of significant abundances of Bacteroidetes and Clostridiales. Although the children and mother shared essentially the identical diet and environment, the children’s microbiotas were not significantly more similar to their mother than they were to their father.
This analysis underscores the complex interactions that give rise to a personalized microbiota and suggests the value of studying families as a surrogate for longitudinal studies.
KeywordsMicrobiome Family Feces Dynamics Chronosequence Community Development
Numerous studies have identified associations between deviations in the gut microbiota (that is the community of microorganisms living within the gastrointestinal tract) and diseases as varied as psoriasis, diabetes, colon cancer, and susceptibility to Clostridium difficile infection [1–4]. The mechanisms that give rise to an individual’s microbiota as well as the deviations from their normal microbiota are poorly understood. In light of our growing appreciation for the role of the microbiota in maintaining a healthy state, with isolated exceptions such as fecal microbiota transplant as a treatment for recurrent Clostridium difficile infection , we are largely powerless to manipulate the microbiota to achieve long-term transitions to a healthy state. Fundamental to this problem are the sources of our microbiota and the relative importance of numerous variables that can affect the structure and function of the microbiota.
Genetics, environment, life history characteristics, and diet are expected to have significant long-term impact on the composition of one’s microbiota. Studies of monozygotic and dizygotic twins suggest that monozygotic twins who share an identical set of genes have more similar microbiotas than dizygotic twins who only share half their genes [6, 7]; however, the biological significance of the difference in similarity is likely minimal. Furthermore, shared environments and diet confound the similarity between twins. In addition, the microbiotas of co-habiting individuals tend to be more similar than individuals that are not co-habiting; this argues for the importance of a shared environment and similar diet in shaping the microbiota . Life history characteristics such as whether one was breastfed or bottle-fed or born vaginally or via Cesarean section have been shown to impact the immediate structure of the individual’s microbiota in infancy [9, 10]; however, it is unclear what long-term impacts these characteristics have on the composition of the microbiota. One notable example of such investigations was a 2.3-year time course study of a child’s life starting at birth . Discrete changes in his microbiota were associated with fever and coincident transitions in diet and antibiotic therapies. That study suggests that large perturbations are needed to shift a child’s microbiota from one community structure to another. Several other studies have employed antibiotic perturbations and observed that the structure of the microbiota largely returns to its pre-treatment state after the cessation of the treatment [12–14]. Similar results have been observed among individuals who undergo bowel preparation prior to colonoscopy . In short-term diet perturbation studies, groups of individuals have been given diets that are discordant with their normal diet, and although their microbiota changes, it does not converge to resemble the microbiota of others receiving the same diet. In addition, when the individuals return to their normal diet, their microbiotas also return to their previous community structure [16, 17]. These studies and numerous others indicate that the microbiota is relatively robust to perturbation as numerous studies have shown that the structure of an individual’s microbiota is more similar to itself over time than it is to the microbiota of another individual [6, 18, 19]. The model that emerges from these studies is that the fundamental source for one’s microbiota is the physical and biological environment in which the individual lives. Meanwhile, other factors including genetics, immunological exposures, environment, life history characteristics, diet, and overall philosophy to using antibiotics and other clinical interventions, sculpt the underlying community structure.
Families provide a unique platform for testing the factors that impact the membership and abundance of one’s microbiota because they provide greater opportunities to control for the factors that affect the structure of the microbiota. For example, an analysis of a family where one child becomes a vegetarian would improve our understanding of the effects of diet on the microbiota while controlling for the other factors. In addition, families with various-aged children may represent a chronosequence of the family’s microbiota . Chronosequences could be used to understand how the microbiota develops over time without having to collect samples for numerous years from a single individual. They could also be useful as a tool for determining when an individual’s microbiota deviates from his or her siblings. Over short periods of time, analysis of a family’s microbiota could also inform our understanding of how perturbations to one individual’s microbiota would impact the microbiota of others in the family. These studies have been performed to understand the transmission of pathogens [21–23]. In light of these opportunities, we characterized the gut microbiota of a family with six children over the course of a month relative to a cohort of unrelated adults from the same geographic region.
Sample collection and DNA extraction
This study was approved by the University of Michigan Institutional Review Board. All subjects or their parents granted consent to participate in the study. The members of the family obtained fecal samples by scraping feces from toilet paper at their home and their place of employment using sterile wooden applicators ; the infant’s samples were obtained by scraping feces from his cloth diapers using sterile wooden applicators . The parents obtained the samples for the children and kept a diary of the food the children ate during the course of the study. Because of the size of the family, it was not practical to record the amounts of each food consumed by the family members. Samples from unrelated adults in the broader community were collected from individuals residing in Ann Arbor, MI area (53 males, 102 females; ages 19 to 88 years). Subjects were excluded if they had had any signs of diarrhea in the previous seven days or were pregnant. All fecal samples were immediately stored at -20°C until DNA extraction. Total bacterial DNA was extracted from each fecal sample using the PowerSoil®-htp 96 Well Soil DNA Isolation Kit (MO BIO Laboratories Inc., Carlsbad, CA, USA) on an EpMotion 5075 liquid handling workstation (Eppendorf, Hauppauge, NY, USA).
DNA sequencing and curation
The V3-V5 region of the 16S rRNA gene was amplified and sequenced using the 454 GS FLX pyrosequencing platform at the Baylor College of Medicine as described previously . In parallel to the fecal samples, a mock community was included on each sequencing run for calculating sequencing error rates after curation . All 16S rRNA gene sequences were curated using the mothur software package as previously described [25, 26] and resulted in a final error rate of 0.009%. Sequences were clustered into operational taxonomic units (OTUs) using a 3% distance cutoff with the average neighbor clustering algorithm . Taxonomic assignments were determined using a naïve Bayesian classifier trained using the RDP training set with an 80% bootstrap confidence threshold . All samples were rarefied to 1,827 sequences per sample to avoid the detrimental effects of uneven sampling.
Metagenomic shotgun sequencing and curation
For each of the eight family members, samples were collected at days 1, 15, and 26, which corresponded to the beginning, middle, and end of the study. Random genomic DNA from these samples was sequenced as previously described at the Baylor College of Medicine . SeqPrep was used to remove primer sequences from reads (https://github.com/jstjohn/SeqPrep). All reads were pooled together and normalized using khmer’s digital normalization pipeline . This excluded from assembly any read that had a median k-mer of length 20 that had previously been encountered at least 20 times. The excluded reads were saved for downstream analysis. The remaining reads were filtered by abundance, removing any low abundance and unique k-mers. The filtered reads were assembled by velvet with k-mer lengths of 31 and 35 . The contigs from these assemblies were combined and de-replicated at 99% with CD-HIT . The combined, de-replicated contigs were merged with minimus2 . Merged contigs and singletons from minimus2 were screened by BLASTn for hits to human sequences. Contigs with hits to human sequences and the associated reads were removed from further analysis. Open reading frames (ORFs) were predicted from assembled contigs with MetaGeneAnnotator . Gene counts were obtained by mapping reads to the predicted genes with bowtie . UBLAST was used to assign each translated ORF to Kyoto Encyclopedia of Genes and Genomes (KEGG) orthology categories (KO) [36, 37]. Those ORFs that mapped to genes that were not already assigned to a KO category or lacked a significant match in the KEGG database were pooled into a single category. To assign ORFs to operational protein families (OPFs), we first performed a database-independent all-versus-all BLASTP search of the ORFs. The resulting BLAST scores were used to calculate distances between the ORFs, which were clustered using the average neighbor clustering algorithm with a 25% dissimilarity cutoff . All samples were rarefied to 1,207,904 sequences per sample (approximately 114 Mbp per sample) to avoid the detrimental effects of uneven sampling.
The mothur software package was used to calculate the inverse Simpson alpha diversity index, the θYC measure of community structure, and non-metric dimensional scaling (NMDS) ordinations for both the 16S rRNA gene and metagenomic sequence data . Random Forest analysis of the 16S rRNA gene sequence data was performed using the randomForest R package with 10,000 trees (http://cran.r-project.org/).
The 16S rRNA gene sequence data, metagenomic sequence data, and the associated MIMARKS spreadsheet are available online (http://www.mothur.org/FamilyStudy).
Results and discussion
Descriptive characteristics of family members
Number of OTUs in core microbiota
Infant (0 years old)
2 years old
4 years old
6 years old
8 years old
10 years old
Our analysis of this family’s microbiota demonstrates that they represent a unique island within the possible permutations of microbiota structures. Within the family, each individual had a unique, personalized microbiota that allowed them to be differentiated from other members of their family. These personalized microbiota appear to develop at an early age, likely after weaning. Although it remains to be seen whether that is the microbiota that the children will carry with them through adolescence, it suggests that the differences in genetics and diet, environment, and life history characteristics imprint their effects on the microbiota at an early age. Despite the personalized nature of each microbiota, the overall family is clearly more similar to each other than they are to unrelated individuals from the broader community. Overall, these results confirm the model that individuals who share an environment likely share the same ecological meta-community that can colonize the microbiota and then be selected upon by host genetics, diet, and life history.
Although the microbiota of the family members are personalized to each member and are clearly distinct from those of the Ann Arbor community, there was still a large amount of temporal day-to-day variation. It is interesting that the underlying membership of each microbiota was consistent across the study for each person but the abundances of the individual populations were variable. Furthermore, we were unable to associate these fluctuations with diet, differences in environment, or health. This family experienced many disturbances to their microbiota via fluctuations in the composition of their diet and differences in environment. Yet their gut microbiotas were largely resilient to these disturbances. This suggests that the composition of their individual gut microbiota have been selected for to adapt to these disturbances. This suggests that adaptation by the microbiota to a personalized set of disturbances (for example, food preferences, hygiene, behaviors) helps to select for a personalized microbiota that is resilient to the disturbances.
The family considered in the current study will offer several opportunities to better understand the microbiota. First, our data suggest that the family represents a chronosequence, which can be used to understand the connection between child development and overall microbiota dynamics. For example, as the various children go through different life events such as weaning, puberty, and moving away from home, it will be possible to assess the effects of these events on the microbiota. Here, we saw the profound influence of complete weaning on the microbiota when viewing the 2-year-old who had not yet been weaned as a control for her older siblings. Second, this family offers the ability to better understand the effects of mode of birth on the development of the microbiota when controlling for genetics, environment, and diet, as the infant in this study was the only child to not be born vaginally. Following the development of the microbiota in this child relative to his siblings will help us to better understand the long-term impacts of Cesarean delivery. Finally, in the present analysis, the weaned children had similar microbiotas relative to their parents, suggesting that factors other than age or sex are most important in shaping their microbiota. Tracking these children to identify events that lead to deviations in microbiota structure will allow us to better understand the mechanisms that shape and reinforce the structure of these communities.
Families represent a special cultural entity with shared genetics, environment, diet, and microbiota. Unfortunately, they have been largely ignored as a medium for understanding how genetics, environment, and diet interact to form an individual’s personalized microbiota. All families are different and present different mixtures of genetics, environment, and diet. Although this family may be considered unique because of the large number of children in it, exposure to livestock, and homeschooling, all families have idiosyncrasies that make them unique. As our data suggest, children are born into an environment where they are provided with the family’s microbiota; however, their unique genetics, diet, and life history exert a selection on that microbiota to make their own at a very early age. Therefore, it is critical that we develop a better understanding of how individualized microbiota develop as a function of human social interactions with each other and their environment. How this translates to other communal living arrangements, such as establishing new families, dormitories, hospitals, and assisted living centers, is likely to yield a better understanding of the mechanisms that affect the structure and function of the microbiota.
body mass index
Kyoto Encyclopedia of Genes and Genomes
KEGG orthology categories
non-metric dimensional scaling
operational protein family
open reading frames
operational taxonomic unit.
This work was supported by several grants from the National Institutes for Health (1R01GM099514, R01HG005975, U19AI090871, and P30DK034933). The funding agencies had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Written informed consent was obtained from the patients or their relatives for publication of this manuscript and accompanying images.
- Alekseyenko AV, Perez-Perez GI, Souza AD, Strober B, Gao Z, Bihan M, Li K, Methé BA, Blaser MJ: Community differentiation of the cutaneous microbiota in psoriasis. Microbiota. 2013, 1: 31-View ArticleGoogle Scholar
- Qin J, Li R, Raes J, Arumugam M, Burgdorf KS, Manichanh C, Nielsen T, Pons N, Levenez F, Yamada T, Mende DR, Li J, Xu J, Li S, Li D, Cao J, Wang B, Liang H, Zheng H, Xie Y, Tap J, Lepage P, Bertalan M, Batto JM, Hansen T, Le Paslier D, Linneberg A, Nielsen HB, Pelletier E, Renault P: A human gut microbial gene catalogue established by metagenomic sequencing. Nature. 2010, 464: 59-65.View ArticlePubMedPubMed CentralGoogle Scholar
- Marchesi JR, Dutilh BE, Hall N, Peters WH, Roelofs R, Boleij A, Tjalsma H: Towards the human colorectal cancer microbiota. PLoS One. 2011, 6: e20447-View ArticlePubMedPubMed CentralGoogle Scholar
- Antharam VC, Li EC, Ishmael A, Sharma A, Mai V, Rand KH, Wang GP: Intestinal dysbiosis and depletion of butyrogenic bacteria in Clostridium difficile infection and nosocomial diarrhea. J Clin Microbiol. 2013, 51: 2884-2892.View ArticlePubMedPubMed CentralGoogle Scholar
- Schwan A, Sjolin S, Trottestam U, Aronsson B: Relapsing Costridium difficile enterocolitis cured by rectal infusion of homologous faeces. Lancet. 1983, 2: 845-View ArticlePubMedGoogle Scholar
- Turnbaugh PJ, Hamady M, Yatsunenko T, Cantarel BL, Duncan A, Ley RE, Sogin ML, Jones WJ, Roe BA, Affourtit JP, Egholm M, Henrissat B, Heath AC, Knight R, Gordon JI: A core gut microbiome in obese and lean twins. Nature. 2009, 457: 480-484.View ArticlePubMedPubMed CentralGoogle Scholar
- Yatsunenko T, Rey FE, Manary MJ, Trehan I, Dominguez-Bello MG, Contreras M, Magris M, Hidalgo G, Baldassano RN, Anokhin AP, Heath AC, Warner B, Reeder J, Kuczynski J, Caporaso JG, Lozupone CA, Lauber C, Clemente JC, Knights D, Knight R, Gordon JI: Human gut microbiome viewed across age and geography. Nature. 2012, 486: 222-227.PubMedPubMed CentralGoogle Scholar
- Song SJ, Lauber C, Costello EK, Lozupone CA, Humphrey G, Berg-Lyons D, Caporaso JG, Knights D, Clemente JC, Nakielny S, Gordon JI, Fierer N, Knight R: Cohabiting family members share microbiota with one another and with their dogs. eLife. 2013, 2: e00458-PubMedPubMed CentralGoogle Scholar
- Jakobsson HE, Abrahamsson TR, Jenmalm MC, Harris K, Quince C, Jernberg C, Bjorksten B, Engstrand L, Andersson AF: Decreased gut microbiota diversity, delayed Bacteroidetes colonisation and reduced Th1 responses in infants delivered by Caesarean section. Gut. 2013, 63: 559-566.View ArticlePubMedGoogle Scholar
- Azad MB, Konya T, Maughan H, Guttman DS, Field CJ, Chari RS, Sears MR, Becker AB, Scott JA, Kozyrskyj AL, Investigators CS: Gut microbiota of healthy Canadian infants: profiles by mode of delivery and infant diet at 4 months. CMAJ. 2013, 185: 385-394.View ArticlePubMedPubMed CentralGoogle Scholar
- Koenig JE, Spor A, Scalfone N, Fricker AD, Stombaugh J, Knight R, Angenent LT, Ley RE: Succession of microbial consortia in the developing infant gut microbiome. Proc Natl Acad Sci U S A. 2011, 108 (Suppl 1): 4578-4585.View ArticlePubMedPubMed CentralGoogle Scholar
- Dethlefsen L, Huse S, Sogin ML, Relman DA: The pervasive effects of an antibiotic on the human gut microbiota, as revealed by deep 16S rRNA sequencing. PLoS Biol. 2008, 6: e280-View ArticlePubMedPubMed CentralGoogle Scholar
- Dethlefsen L, Relman DA: Incomplete recovery and individualized responses of the human distal gut microbiota to repeated antibiotic perturbation. Proc Natl Acad Sci U S A. 2010Google Scholar
- Young VB, Schmidt TM: Antibiotic-associated diarrhea accompanied by large-scale alterations in the composition of the fecal microbiota. J Clin Microbiol. 2004, 42: 1203-1206.View ArticlePubMedPubMed CentralGoogle Scholar
- O’Brien CL, Allison GE, Grimpen F, Pavli P: Impact of colonoscopy bowel preparation on intestinal microbiota. PLoS One. 2013, 8: e62815-View ArticlePubMedPubMed CentralGoogle Scholar
- Wu GD, Chen J, Hoffmann C, Bittinger K, Chen YY, Keilbaugh SA, Bewtra M, Knights D, Walters WA, Knight R, Sinha R, Gilroy E, Gupta K, Baldassano R, Nessel L, Li H, Bushman FD, Lewis JD: Linking long-term dietary patterns with gut microbial enterotypes. Science. 2011, 334: 105-108.View ArticlePubMedPubMed CentralGoogle Scholar
- David LA, Maurice CF, Carmody RN, Gootenberg DB, Button JE, Wolfe BE, Ling AV, Devlin AS, Varma Y, Fischbach MA, Biddinger SB, Dutton RJ, Turnbaugh PJ: Diet rapidly and reproducibly alters the human gut microbiome. Nature. 2013, 505: 559-563.View ArticlePubMedPubMed CentralGoogle Scholar
- Caporaso JG, Lauber CL, Costello EK, Berg-Lyons D, Gonzalez A, Stombaugh J, Knights D, Gajer P, Ravel J, Fierer N, Gordon JI, Knight R: Moving pictures of the human microbiome. Genome Biol. 2011, 12: R50-View ArticlePubMedPubMed CentralGoogle Scholar
- Costello EK, Lauber CL, Hamady M, Fierer N, Gordon JI, Knight R: Bacterial community variation in human body habitats across space and time. Science. 2009, 326: 1694-1697.View ArticlePubMedPubMed CentralGoogle Scholar
- Eisenlord SD, Zak DR, Upchurch RA: Dispersal limitation and the assembly of soil Actinobacteria communities in a long-term chronosequence. Ecol Evol. 2012, 2: 538-549.View ArticlePubMedPubMed CentralGoogle Scholar
- Rydberg J, Cederberg A: Intrafamilial spreading of Escherichia coli resistant to trimethoprim. Scand J Infect Dis. 1986, 18: 457-460.View ArticlePubMedGoogle Scholar
- Caugant DA, Levin BR, Selander RK: Distribution of multilocus genotypes of Escherichia coli within and between host families. J Hyg (Lond). 1984, 92: 377-384.View ArticleGoogle Scholar
- Mendez Ede L, Roldan ML, Baroni MR, Mendosa MA, Cristobal SA, Virgolini SM, Faccone D: A case of familial transmission of community-acquired methicillin-resistant Staphylococcus aureus carrying the Inu(A) gene in Santa Fe city. Argent Rev Argent Microbiol. 2012, 44: 303-305.PubMedGoogle Scholar
- Schloss PD, Schubert AM, Zackular JP, Iverson KD, Young VB, Petrosino JF: Stabilization of the murine gut microbiota following weaning. Gut Microbes. 2012, 3: 383-393.View ArticlePubMedPubMed CentralGoogle Scholar
- Schloss PD, Gevers D, Westcott SL: Reducing the effects of PCR amplification and sequencing artifacts on 16S rRNA-based studies. PLoS One. 2011, 6: e27310-View ArticlePubMedPubMed CentralGoogle Scholar
- Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB, Lesniewski RA, Oakley BB, Parks DH, Robinson CJ, Sahl JW, Stres B, Thallinger GG, Van Horn DJ, Weber CF: Introducing mothur: Open-source, platform-independent, community-supported software for describing and comparing microbial communities. Appl Environ Microbiol. 2009, 75: 7537-7541.View ArticlePubMedPubMed CentralGoogle Scholar
- Schloss PD, Westcott SL: Assessing and improving methods used in operational taxonomic unit-based approaches for 16S rRNA gene sequence analysis. Appl Environ Microbiol. 2011, 77: 3219-3226.View ArticlePubMedPubMed CentralGoogle Scholar
- Wang Q, Garrity GM, Tiedje JM, Cole JR: Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol. 2007, 73: 5261-5267.View ArticlePubMedPubMed CentralGoogle Scholar
- The Human Microbiome Consortium: Structure, function and diversity of the healthy human microbiota. Nature. 2012, 486: 207-214.View ArticleGoogle Scholar
- Brown CT, Howe A, Zhang Q, Pyrkosz AB, Brom TH: A reference-free algorithm for computational normalization of shotgun sequencing data. arXiv:12034802v2 [q-bioGN]. 2012Google Scholar
- Zerbino DR, Birney E: Velvet: algorithms for de novo short read assembly using de Bruijn graphs. Genome Res. 2008, 18: 821-829.View ArticlePubMedPubMed CentralGoogle Scholar
- Li W, Godzik A: CD-HIT: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics. 2006, 22: 1658-1659.View ArticlePubMedGoogle Scholar
- Sommer DD, Delcher AL, Salzberg SL, Pop M: Minimus: a fast, lightweight genome assembler. BMC Bioinformatics. 2007, 8: 64-View ArticlePubMedPubMed CentralGoogle Scholar
- Yok NG, Rosen GL: Combining gene prediction methods to improve metagenomic gene annotation. BMC Bioinformatics. 2011, 12: 20-View ArticlePubMedPubMed CentralGoogle Scholar
- Langmead B, Trapnell C, Pop M, Salzberg SL: Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol. 2009, 10: R25-View ArticlePubMedPubMed CentralGoogle Scholar
- Kanehisa M, Goto S, Sato Y, Kawashima M, Furumichi M, Tanabe M: Data, information, knowledge and principle: back to metabolism in KEGG. Nucleic Acids Res. 2014, 42: D199-D205.View ArticlePubMedPubMed CentralGoogle Scholar
- Edgar RC: Search and clustering orders of magnitude faster than BLAST. Bioinformatics. 2010, 26: 2460-2461.View ArticlePubMedGoogle Scholar
- Schloss PD, Handelsman J: A statistical toolbox for metagenomics. BMC Bioinformatics. 2008, 9: 34-View ArticlePubMedPubMed CentralGoogle Scholar
- Koren O, Goodrich JK, Cullender TC, Spor A, Laitinen K, Backhed HK, Gonzalez A, Werner JJ, Angenent LT, Knight R, Backhed F, Isolauri E, Salminen S, Ley RE: Host remodeling of the gut microbiome and metabolic changes during pregnancy. Cell. 2012, 150: 470-480.View ArticlePubMedPubMed CentralGoogle Scholar
- Avershina E, Storro O, Oien T, Johnsen R, Wilson R, Egeland T, Rudi K: Bifidobacterial succession and correlation networks in a large unselected cohort of mothers and their children. Appl Environ Microbiol. 2013, 79: 497-507.View ArticlePubMedPubMed CentralGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.