Moore Notes 11 13 13
From OpenWetWare
Jump to navigationJump to search
Group Call
- Participants: Katie, Tom, Josh, Ladan, Guillaume
- PICRUSTs vs. Shotmap analysis (both with Kegg Orthology groups (KOs))
- Filtered: only using KOs present in PICRUSTs results http://edhar.genomecenter.ucdavis.edu/~gjospin/picrust_test/picrust_test/filtered_results/
- Hamming distance between samples is good (presence-absence)
- Check what families these are (e.g., via Patrick's MetaCyc analysis)
- But abundance vectors are only weakly correlated (Spearman rho)
- Hamming distance between samples is good (presence-absence)
- Top 100: most abundant KOs from PICRUSTs http://edhar.genomecenter.ucdavis.edu/~gjospin/picrust_test/picrust_test/top100/
- Correlations are not better for these KOs
- If shotmap is more sensitive, then using stricter thresholds for read classification should improve correlation with PICRUSTs (and Jack's metagenome annotation)
- Tom and Guillaume will check this and report back next week
- Filtered: only using KOs present in PICRUSTs results http://edhar.genomecenter.ucdavis.edu/~gjospin/picrust_test/picrust_test/filtered_results/
- Shotmap paper
- Wrapping up analyses in next few weeks
- Then finish writing
- Patrick could contribute MetaCyc annotation results
- ISME 2014: South Korea
- Any one going?
- Next call: Guillaume and Josh
- Two weeks from now: Ladan