Dinh/Dinh 2012/NOTES/2012-2-9

From ZhangLabWiki
Revision as of 16:53, 15 February 2012 by >Dinh
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Seminar by Stephen Briggs[edit]

  • Development of a highly sensitive and reliable protein quantification by MS.
  • System have automated separation of proteins by charge (fewer charge in the machine = higher sensitivity.)
  • Current gene annotation are very accurately. However, ~1K proteins needs to be re-annotated (according to MS results.)
  • Comparison of protein MS with microarray data shows that most proteins are post-translationally regulated
  • Although, a subset of proteins have correlation with microarray data.
  • A small subset of proteins have extremely inverse correlation with mRNA levels. What does this mean?
  • "if you have a hammer you see the whole full of nails"

Working on 24 ChIP data[edit]

  • I'm still analyzing the peak calling results. After normalization and peak calling with MACs, my WIG files did not have the correct number of chromosomes (ie. chr12-19 are often missing). One hypothesis is that the normalization of reads cause MACS to not get any signal from these chromosomes.
  • Currently, I will use MACs to generate WIG files for every single ChIP dataset, then I can perform the Fisher Exact Test (2x2 table.)
  • I generated some summary statistics from the paired-end mapping.
  • Overall, we increased the %mappable reads by PE mapping.
  • The increase is up to ~%50 for R.mC and S.mC! PE mapping were able to rescue reads which mapped to repeats.
File:ChIP 24 mapping.png
  • After removing PCR duplicates:
File:ChIP 24 PCRdup.png
  • I checked the distribution of PE categories
-Proper = properly paired reads - both are in the (->,<-) orientation and within expected length
-Diff_chr = both reads are mapped on different chromosomes
-Indel = both reads are mapped in (->,<-) orientation but much larger or smaller length than expected.
-Inverted = both reads are mapped in (<-,<-) or (->,->) orientation
-Tandem = both reads are mapped in (<-,->) orientation, indicating tandem duplication
  • Could the "improperly" paired reads be in repeats?
    • Can we call peaks on these to see whether some class of repeat is enriched?
File:ChIP 24 PEdist.png