2011 marks the third IHMC and included over 500 members of the medical, microbial and computational fields who gathered this year in Vancouver, Canada. These research leaders made it a truly international event, travelling from all over Europe, Asia, Australia, Africa and North America to attend.
The Genome Institute’s Associate Director, Dr. George Weinstock, was one of the key organizers of the conference. As a leader in the Human Microbiome Project, Dr. Weinstock is applying next generation sequencing technologies to help characterize the millions of microbes that live in and on the human body.
The Genome Institute had a strong presence at the IHMC, with a number of speakers and posters outlining the incredible advances in metagenomics, a field that involves recovering genetic material from massive groups of microbes present in an environment (such as the human body) without first culturing (or growing) these microbes separately.
Here are some highlights of The Genome Institute’s human microbiome research presented at the 2011 IHMC:
Looking for Biology in the Human Microbiome, George Weinstock
Dr. Weinstock described his view on the current state of the NIH’s Human Microbiome Project (HMP). He outlined the difficulties in creating a ‘reference’ microbiome (a far more complex undertaking than the human reference genome, which was created by the Human Genome Project) and comparing the healthy and diseased or ‘perturbed’ microbiome in individuals. He also summarized various new approaches to metagenomic analyses, what we have learned so far and the challenges that lie ahead.
Temporal Effects of the 7-Valent Pneumococcal Conjugate Vaccine (PCV7) on the Nasopharyngeal Microbiome Among Gambian Infants, Martin Antonio
Dr. Weinstock filled in for collaborator Dr. Martin Antonio from the UK’s Medical Research Council, who could not attend the meeting. Dr. Antonio’s work looks at the effects of the pneumococcus vaccine on the nasopharyngeal microbiome in Gambian Infants. Pneumococcus is one of the leading causes of child pneumonia deaths, particularly in Africa and South Asia. The Genome Institute is helping to sequence and compare samples from infants who have received the vaccine and those who have not.
The Microbiome of Healthy Humans: Commonalities and Variations, Makedonka Mitreva
Dr. Makedonka Mitreva, Assistant Director at The Genome Institute, spoke about advantages and disadvantages of the types of sequence data generated by the Human Microbiome Project (HMP) and how these techniques are applied to the characterization of microbial communities. She outlined the tools The Genome Institute is using to analyze samples from both the HMP and the European project, MetaHIT (Metagenomics of the Human Intestinal Tract). Dr. Mitreva and colleagues compared the microbiomes of healthy patient samples from both projects and found a great deal of correspondence between the two, though more research is still needed to make any final conclusions. Their goal is to determine the compositions of the various microbial communities in the body, how these microbes function and how they compare within an individual and among populations. This information will help further the study of human disease, as it relates to our microbiome.
The Characterization of the Human Virome in Children and Adults, Kristine M. Wylie
The virome, or the complement of viruses that inhabit our bodies, remains a less frequently studied part of the human microbiome. Dr. Kristine Wylie spoke about work she and colleagues are doing to characterize the virome in children with fever of unknown origin and those without fever. Her group found that children who presented with fever had more viruses deep within their nasal cavities and within their plasma than those children without fever. Dr. Wylie also outlined preliminary studies on the viromes of healthy individuals from the HMP sampled over multiple visits. Using an unbiased sequencing approach on samples taken from multiple sites around the body, the group identified many viruses known to cause acute and chronic infections, such as herpes viruses and human papilloma viruses (HPV). They also detected certain commonly found chronic infections in some of the individuals. The results are preliminary but compelling as researchers seek to characterize all of the organisms present within the human microbiome.
Longitudinal Assessment of Organismal and Functional Variability in the Human Microbiome
Abubucker, Sahar; Kota, Karthik; Martin, John; Zhou, Yanjiao; Gao, Hongyu; HMP Data Analysis Working Group; Sodergren, Erica; Weinstock, George; Mitreva, Makedonka
The HMP provides samples from multiple body sites taken from 100 healthy individuals. Some individuals in this group have been sampled more than once, providing data for comparison over time. This poster looked at differences in individual HMP microbial communities from the same body site sampled during multiple visits. The researchers looked at correlations between visits at each body site using 16S and whole genome sequencing, and by characterizing the microbial function in the communities. They found that at a given body site, individuals varied slightly in their microbial makeup but the organisms’ functions remained very stable. Function is also well conserved between different individuals in the same body site. They also examined single base changes (SNPs) in bacterial strains between visits to characterize slight changes in variation.
Taxonomic and Functional Characterization of Human Microbial Community Structure
Gao, Hongyu; Weinstock, George; Sodergren, Erica; Mihindukulasuriya, Kathie; Wylie, Kristine; Zhou,Yanjiao; Abubucker, Sahar; Mitreva, Makedonka
When examining the human microbiome in different individuals, there are many differences - even within the same body site. This poster surveyed multiple individuals’ microbial communities at various body sites at the more general genus level. The group found that variation between individuals is relatively low at sites such as the mouth and stool but relatively high in the skin and vaginal sites in terms of both bacterial make-up and abundance. Using unprecedented large datasets, the researchers identified certain subgroups (called biome subtypes) with similar bacterial genus composition, even though the overall community composition varies at those sites. These results demonstrate that while differences abound, there are some commonalities in individuals’ microbial communities. This can help provide a baseline for future studies that aim to characterize the microbiome in healthy and diseased states.
New Goal for HMP Reference Genome Catalog: the Human Microbiome Strain Collection
Lobos, Elizabeth; Dunne, Michael; Hoppe-Bauer, Joan; Sodergren, Erica; Cejkova, Darina; Strouhal, Michal; Fulton, Robert; Courtney, Laura; Weinstock, George
The Human Microbiome Project (HMP) has recently raised its targeted number of sequenced reference microbial genomes from 900 to 3000. This new goal is possible through continued increases in sequencing performance and decreases in cost. The ‘rate-limiting step’ for reaching this goal has now become the intake and processing of microbial samples. To address this problem, The Genome Institute is collaborating with the Clinical Microbiology Laboratory at Barnes-Jewish Hospital to isolate bacteria from clinical samples, in order to establish the Human Microbiome Strain Collection (HMSC). The goal of the HMSC is to collect 10,000 strains. HMP researchers will sequence the 16S ribosomal RNA gene from the samples to classify them. They will retain strains for the HMSC that add to the diversity of the existing HMP reference genomes, as these are good candidates for whole genome sequencing. They will send the HMSC strains to the Biodefense and Emerging Infections Research Resources Repository (BEI) so other scientists may access them. This new pipeline for samples, described in this poster, could lead to the discovery of novel species. It will also improve the HMP reference genome database and enhance microbial classification and analysis.
Optimizing Read Mapping to Reference Genomes to Determine Composition and Species Prevalence in Microbial Communities
Martin, John; Sykes, Sean; Young, Sarah; Kota, Karthik; Sanka, Ravi; Sheth, Nihar; Orvis, Joshua; HMP Data Processing Working Group; Weinstock, George; Mitreva, Makedonka
The HMP comprises a massive amount of data from microbial communities spanning 18 human body sites. Characterizing this vast amount of metagenomic sequence data represents a challenge that requires updated computational tools and methodologies. This poster describes efforts to evaluate a number of currently available tools and strategies that align and map metagenomic data from large microbial communities. The group evaluated these tools and strategies based on their ability to detect and classify these communities and how well they scale for this volume of data. They performed this evaluation using software and mapping strategies to map a mock community of known composition to a genome reference database of the organisms within that community. They judged the various methods based on sensitivity, accuracy, speed, scalability and ease of use. Once they determined the winning method, they then found the optimal parameters that would most improve classification of the community, while maintaining a low false positive identification rate. Having robust and scalable software and a sound strategy for handling large metagenomic datasets is important to the success of the HMP and will assist in its goal of characterizing the healthy human microbiome.
Variations in the Microbiome of Healthy Humans Defined by Comparison of Cumulative Abundance Distribution
Weinstock, George; Kota, Karthik; Mitreva, Makedonka; Martin, John; Wang, Zhengyuan
It is difficult to use common statistical tests to compare and evaluate large data sets such as those found in metagenomic microbial communities. This poster evaluates the Kolmogorov-Smirnov statistical test (KS-test) when comparing HMP metagenomic samples. Using this test, the group looked at the differences in microbial communities from the mouth and the stool. Traditional statistical tests, which work best with smaller data sets, tend to inflate differences in large data samples. The KS-test worked best for these large data analyses as it detected the true differences present in the samples, without inflating statistical significance. Determining the best statistical test for such metagenomic data comparisons will ultimately be important for diagnosing individuals based on their microbiomes.
Identification of Novel 16S Sequences in Metagenomic Data Sets
Wylie, Kristine; Mihindukulasuriya, Kathie; Zhou, Yanjiao; Sodergren, Erica; Weinstock, George
16S sequencing is a common way to classify both known and new bacteria from HMP samples. Much of the 16S data currently available has not yet been classified. This poster describes the use of whole genome shotgun sequencing to validate this 16S data and to help determine if this data indicates the presence of novel forms of bacteria in a set of HMP samples. Using this technique, the group has been able to identify new 16S sequences and has paved the way for further complete genome sequencing for a number of bacterial candidates. Such enhanced identification of novel organisms will improve the bacterial genome database currently used when analyzing communities within the human microbiome.