Friday, October 12 Notes

[klgray25]

Friday, October 12

Project Background

Prior Sample/Data Processing

• Samples collected from marine environments around the world
• Water sample collected
• Filtered to isolate viral component and bacteria (+viruses) component
• Selected for dsDNA viruses only
• Fractions underwent shotgun metagenomic sequencing
• Assembled into contigs
• Generated RPKM values
• Compared to metabolomics databases (MetaCyc, COG, KEGG, 3 more) using FAST - combined the rpkm values for all enzymes in pathway with set of rules (specific to each pathway)
• Output was RPKM value for enriched pathways for every sample
• EBI Link: https://www.ebi.ac.uk/metagenomics/studies/ERP001736
• Metadata Link: https://www.ebi.ac.uk/ena/submit/tara-oceans-checklist

Data Contents

• Rpkm - After the contigs are assembled, the reads were aligned back to the contigs - this was used to generate rpkm (essentially a normalized genomic abundance metric - reads per kilobase of transcript per million mapped reads)
• Sample name - c=non-viral (includes bacterial and viral), ERR=viral only
• Type - single includes only one fraction - multi includes the viral and bacterial fraction data in a single analysis
  Date information exists - Simon Rao
• See comment below for more details
• It's okay to make the data public

Notes From Steve Hallam Visit

• Tara Oceans Project - project that underwent the sampling expedition: http://ocean-microbiome.embl.de/companion.html
• International consortium of oceanography/marine biologists - made a standardized sample collection process (data is comparable)
• First expedition - photic samples - didn’t sample very deep
• PathwayTools - prediction engine - need licence - made MetaCyc identifiers from this
• PathoLogic - has harmonized names
• New idea - metabolically functional genes encoded in viruses - more widespread than imagined before
• Talked about this paper: http://www.pnas.org/content/108/39/E757.short
• Cyanobacteria normally have fast turn-over - slow down and halt photosynthesis in response to viruses (sequester them and protect neighbouring cells) - virus carries genes that are part of the photosystem - overcomes the defence mechanism and promotes photosynthesis, cellular division
• Pathway tools - KEGG Atlas - have diagrams for metabolism - recommended using these
• Envisions this turning into a manuscript - Nature Scientific Data publication
• Heatmap with distribution of pathways good starting point (something similar to KEGG atlas ideal though)
• Want to be able to do things like compare samples in Indian Ocian to x Ocean
• Pathways by location heatmap
• Metaviriome - attracted to certain pathways - want to visualize the pathways that are affected

MetaCyc Notes

• Reference database of enzymes and metabolic pathways
• Mostly small molecule pathways (but updated versions add macromolecular metabolic pathways)
• Tool PathoLogic uses to predict metabolic networks of organism with annotated genome files - generates Pathway/Genome databases - BioCyc stores the databases generated by SRI
• Used to generate organism-specific pathway/genome databases
• Curated from experimentally validated results/academic papers

Project Update

Approximate Task Divvy Up

• Analysis - Javier, Heather, Jasmine
• Map - Ogan, Dan, Olga, Arjun, Kristen
• Reading/Writing - Kristen, Arjun
• Integration - Arjun

http://www.gutcyc.org - similar initiative - gut microbiome

Website Goals

• Map
  • Data points are plotted to the map with latitude and longitude values
  • Want to be able to query by location, depth, other metadata (temperature, salinity, etc)
  • Clicking on a sample should pull up data on sample information, pathway information, etc. (Want some figures to make data visual - likely by metabolic category) and link out to MetaCyc information
• Analysis Functionality
  • Want differential comparison of metabolic pathway activation for samples with given set of characteristics
  • Want to have a way to filter out pathways that are generally present everywhere
• Interactive KEGG Atlas-like Visualization (if time permits)

Website Progress

• Elected to use shiny to build website
• Set groundwork and set up the map using leaflet (package for maps)
• Working on query module - if filtering, zooming, changing parameters, query returns subset of data
• Began basic formatting of site
• Current site functionality: The site has three pages - a Home/Welcome, Map, and Analysis page - accessible from the sidebar menu. The map appears on the map page and the points are plotted. When a given sample is selected, a field appears with some basic sample data. A Control panel has been added for the map, but it is not yet filled in or functional

[tsa87]

RPKM: Reads Per Kilobase of transcript per Million mapped reads
SRF: Surface Water Layer
DCM: Deep Chlorophyll Maximum
MES: Mesopelagic
OMZ: Oxygen minimum zone