Kun:LabNotes/MONOD/2014-11-22

From ZhangLabWiki
Jump to navigation Jump to search

Nimblegen capture of blood LMS regions[edit]

  • Target selection & probe design: I took the UMR regions Dinh called from the PNAS whole blood WGBS data, did "bedtools subtract" to remove the repeats, then select the 32999 regions that were more than 100bp. The total target size is 29.4Mb.
  • Library preparation & capture: Noi/NOTES/2014-10-29, Noi/NOTES/2014-11-4
  • Read mapping: Dinh/Dinh_2014/NOTES/2014-11-17. From the fraction of targets covered and the enrichment factors, the data from normal plasma appeared to be much better than the data from cancer patients.
  • To take a close look of the read distribution for all these libraries, I extracted the reads for a subset of larger targets (>2k) into smaller bam files for IGV visualization.
   ./get_bam_in_targets.pl /media/Ext12T/DD_Ext12T/MONOD/141112_HiSeqRapidRun/BAMfiles
  • I then load a few bam files with IGV and inspect some of the target regions. Here is a typical case (top two are normal plasma, bottom two are cancer patient plasma). There is some variability across a target region. Normal plasma were more covered with lower variability. The coverage on patient plasma was rather low, and there seems to be many clonal reads, suggesting the starting libraries have low complexity.
 File:Plasma Nimblegen SeqCap EPI bam plot 1.png