Post date: Dec 5, 2015 10:12:05 PM
Updates on the case control analysis:
- Slides with prelim results are available in the Syncplicity folder - please let me know if you would like and do not have access
- Further review of QC noted a few issues: a) MESA still had some inflation, with some very small P-values for some SNPs that appeared to have low MACs. Updating the MAC cutoff to 20 seems to have solved this problem, b) eQTL dataset had some inflation d/t incorrect estimation of betas/Ps possibly from rounding. c) Added GC correction per-study, which is slightly more conservative.
- 12 genome-wide significant results in the combined analysis, of which 3 could be novel (IL27 being reported, is a 4th) if additionally replicated - the rest previously found for COPD or lung fx
- UKBiLeve agrees to do a replication: propose 5x10-6, of which several likely to replicate if considering prior associations in GWAS
- Timeline is short: BiLEVE and Spirometa plan to be done with their large collaborative project by December - so trying to get a draft out by the end of December
- 'Validation' analysis of at least some of the top hits by Kim de Jeong
- eQTL analysis by Yohan
- Need methods sections, co-authors, affiliations
- Secondary analyses to be done.
Future directions:
- Emphysema analysis - likely to be discussed, next call in January
- 23AndMe collaboration - discussed problems with self reported diagnosis, seemed to be a general lack of enthusiasm