Ns126:Calendar/NOTES/2015-8-31

From ZhangLabWiki
Revision as of 21:54, 23 September 2015 by >Shicheng
Jump to navigation Jump to search


compare methylation haplotype region with Hic Data

  • liftover epigenetic annotation from hg18 to hg19
./liftOver Hic.topological.domain.hESC.hg18.bed hg18ToHg19.over.chain Hic.topological.domain.hESC.hg19.bed tmp
./liftOver Hic.topological.domain.IMR90.hg18.bed  hg18ToHg19.over.chain Hic.topological.domain.IMR90.hg19.bed tmp
./liftOver Hic.boundary.IMR90.hg18.bed hg18ToHg19.over.chain Hic.boundary.IMR90.hg19.bed tmp
./liftOver Hic.boundary.hESC.hg18.bed hg18ToHg19.over.chain Hic.boundary.hESC.hg19.bed tmp
./liftOver Hic.common.boundary.hESC.IMR90.hg18.bed hg18ToHg19.over.chain Hic.common.boundary.hESC.IMR90.hg19.bed tmp
/home/sguo/monod/phase2/high.gsi.genome.cor.txt


compare the cluster analysis with raw methylation signal

Achieve raw methylFreq files

  • I found the overlapped CpG sites between methylation haplotype and raw methylFreq (depth>5) was only 49. The reason when we calculate the methylation haplotype we did not get rid of low coverage reads. Therefore, I need collect the raw methylation methylFreq file and merage them again without low coverage reads discarding.
  • After discussing with Dinh, the methylFreq files of 106 samples were collected (10 N37, 36 Salk, 57 MONOD and 3 Heyn2013Age)
  • 651 CpG sites were found overlapped with 180 high GSI methylation haplotype regions. among them, 35 sites were found have more than 30% missing value and then were filtered in the further analysis.
  • the heatmap based on raw methylation signals were as the following(right).

File:GSI.RRBS.BSPP.WGBS.RawSignal.png

Comare MHL and Average methylation level