The focus was on large scale microbial sequencing efforts under the NIH microbiome grant. Presentations by JCVI, NIAID, Scripps.
Karen Nelson (JCVI director, Rockville)
Mining the Human Microbiome
Craig Venter's ocean sampling resulted in 1800 species -> huge backlog for seq -> so instead of seq single organisms, they new seq species level. Aim is to look for microbial diversity in air, water, humanbody, soil.
Human microbiome: extensive microbial diversity in humans (trends at population level) -> host and microbial interactions not known right now -> example in the GI tract -> anybody can seq their microbial diversity -> Eckburg et al. 2005 Science 308 (human colon, identified 395 phylotypes from 11831 bacterial and 1524 archaeal) -> stomach : 128 16S rDNA phylotypes identified from 1833 full-length 16S DNA samples -> first human metagenomic paper (Gill et al. Science 2009) -> human dna contamination in the microbes creates IRB issues -> NIH roadmap human microbiome project : $157million 2007-2013: seq 3000 reference bacterial genomes; viral and eukaryotic genomes, 300 healthy humans diverse body sites, reagent repository, database and resource center, bioinformatic tools development ->Nelson et al. Science 2010 (A catalog of reference genomes from the human microbiome) has 178 microbial genomes -> data is in DACC (HMP) -> Single cell sequencing of uncultured bacteria.
-> Understanding the human virome : in process of seq eukaryotic genomes and viruses -> HMP metagenomics : 554 subjects, sampled body sites in healthy 18-40 year olds, 5 body sites -> body sites have distinct communities -> even two teeth have distinct communities -> several disease-specific metagenomic studies (T1D, colon cancer, obesity models, oral disease, skin, acne, psoriasis, nasopharynx microbiome, bacterial vaginosis etc) -> esophagus cancer relative to other cancers has gone up 6 times -> microbial signatures can be used diagnostically. I asked if the study design captured diet as a variable to correlate with the microbial diversity (particularly when capturing diversity in fecal sample, large contribution of what you usually eat)... but they have not (surprisingly!).
Mihai Pop
Genome assembly is impossible in bacteria (NP hard): you have to try all possible combinations before finding the right solution. Most assemblers take 100bp as input and output 36bp assembly. Solution: go for longer reads so that assembly is easier. HMP settled on SOAPdenovo for assembly. Recipe: http://goo.gl/cMD7C. Overall PGA statistics -> assembly collector's curves (one stool sample)
-> MetAMOS - a new metagenomic assembler: principles: 1) assembly is not an isolated/independent process and 2) Metagenomic assembly is 'special': coverage statistics dont work, multiple haplotypes at different levels of divergence and haplotype structure have a biological meaning -> Bambus2, MetaGeneMark, FPC, Metaphyler -> sequencing based microbial diversity capture has higher power than 16s DNA profiling.
Ian Wilson (Scripps Research Institute)
JCSG -> Structure determination pipeline -> Structural genomics: structural coverage of the protein universe -> Commensal colonies of organisms - Human microbiome -> Sequences to structure : mission: 3D atomic resolution structures of proteins from knowledge of their corresponding DNA sequences : individual proteins to protein families : 5000 structures : domains of unknown function (DUFs): look for similarity to known structures
Maria Giovanni (NIAID, NIH)
Basic research, clinical trials, genomics, proteomics, bioinformatics, systems biology: for microbiology and infectious diseases and to develop new targets and diagnostics. 2 largest funding for NIAID after NCI amongst NIH's 27 centers. Sequenced 2534 genomes for bacteria, fungi, parasite, vectors, plants and worms. Over hundred computational tools developed. Three genome sequencing centers : JCVI, Broad, Univ of Maryland, Baltimore. 3860 bacterial genomes ongoing. Broad -> Ferret. Large scale bioinformatics, data storage, compression. NIAID eukaryotic progam bioinformatics EupathDB. Data has been transferred to Influenza Research Database (partnership with centers).
Karen Nelson (JCVI director, Rockville)
Mining the Human Microbiome
Craig Venter's ocean sampling resulted in 1800 species -> huge backlog for seq -> so instead of seq single organisms, they new seq species level. Aim is to look for microbial diversity in air, water, humanbody, soil.
Human microbiome: extensive microbial diversity in humans (trends at population level) -> host and microbial interactions not known right now -> example in the GI tract -> anybody can seq their microbial diversity -> Eckburg et al. 2005 Science 308 (human colon, identified 395 phylotypes from 11831 bacterial and 1524 archaeal) -> stomach : 128 16S rDNA phylotypes identified from 1833 full-length 16S DNA samples -> first human metagenomic paper (Gill et al. Science 2009) -> human dna contamination in the microbes creates IRB issues -> NIH roadmap human microbiome project : $157million 2007-2013: seq 3000 reference bacterial genomes; viral and eukaryotic genomes, 300 healthy humans diverse body sites, reagent repository, database and resource center, bioinformatic tools development ->Nelson et al. Science 2010 (A catalog of reference genomes from the human microbiome) has 178 microbial genomes -> data is in DACC (HMP) -> Single cell sequencing of uncultured bacteria.
-> Understanding the human virome : in process of seq eukaryotic genomes and viruses -> HMP metagenomics : 554 subjects, sampled body sites in healthy 18-40 year olds, 5 body sites -> body sites have distinct communities -> even two teeth have distinct communities -> several disease-specific metagenomic studies (T1D, colon cancer, obesity models, oral disease, skin, acne, psoriasis, nasopharynx microbiome, bacterial vaginosis etc) -> esophagus cancer relative to other cancers has gone up 6 times -> microbial signatures can be used diagnostically. I asked if the study design captured diet as a variable to correlate with the microbial diversity (particularly when capturing diversity in fecal sample, large contribution of what you usually eat)... but they have not (surprisingly!).
Mihai Pop
Genome assembly is impossible in bacteria (NP hard): you have to try all possible combinations before finding the right solution. Most assemblers take 100bp as input and output 36bp assembly. Solution: go for longer reads so that assembly is easier. HMP settled on SOAPdenovo for assembly. Recipe: http://goo.gl/cMD7C. Overall PGA statistics -> assembly collector's curves (one stool sample)
-> MetAMOS - a new metagenomic assembler: principles: 1) assembly is not an isolated/independent process and 2) Metagenomic assembly is 'special': coverage statistics dont work, multiple haplotypes at different levels of divergence and haplotype structure have a biological meaning -> Bambus2, MetaGeneMark, FPC, Metaphyler -> sequencing based microbial diversity capture has higher power than 16s DNA profiling.
Ian Wilson (Scripps Research Institute)
JCSG -> Structure determination pipeline -> Structural genomics: structural coverage of the protein universe -> Commensal colonies of organisms - Human microbiome -> Sequences to structure : mission: 3D atomic resolution structures of proteins from knowledge of their corresponding DNA sequences : individual proteins to protein families : 5000 structures : domains of unknown function (DUFs): look for similarity to known structures
Maria Giovanni (NIAID, NIH)
Basic research, clinical trials, genomics, proteomics, bioinformatics, systems biology: for microbiology and infectious diseases and to develop new targets and diagnostics. 2 largest funding for NIAID after NCI amongst NIH's 27 centers. Sequenced 2534 genomes for bacteria, fungi, parasite, vectors, plants and worms. Over hundred computational tools developed. Three genome sequencing centers : JCVI, Broad, Univ of Maryland, Baltimore. 3860 bacterial genomes ongoing. Broad -> Ferret. Large scale bioinformatics, data storage, compression. NIAID eukaryotic progam bioinformatics EupathDB. Data has been transferred to Influenza Research Database (partnership with centers).
No comments:
Post a Comment