search

InesdeSantiago / BaalChIP

0 stars 4 forks source link

broad peaks #1

Closed PoisonAlien closed 7 years ago

PoisonAlien commented 7 years ago

Hi,

Can this be used for broad peaks such as H3K27Ac and super enhancers to see if its on single allele or both allele ? Are there any caveats considering these peaks are broader ? I was writing my own script to do so then I came across your package (looks very interesting).

InesdeSantiago commented 7 years ago

Hi! Yes it can be used for any region, as long as you have a list of regions in a bed format. I have tested and published the vignette with TF and FAIRE-seq but I also tested with histone data and it works ok. The software computes the allelic ratio of SNPs within those regions, then it will perform the statistical test to see if the read counts are biased to any of the alleles.

For broad peaks, I wouldn't use the simulation filter (named filterIntbias) which was designed for TF type of data. best, ines

On Thu, Apr 6, 2017 at 3:19 AM, PoisonAlien notifications@github.com wrote:

Hi,

Can this be used for broad peaks such as H3K27Ac and super enhancers to see if its on single allele or both allele ? Are there any caveats considering these peaks are broader ? I was writing my own script to do so then I came across your package (looks very interesting).

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/InesdeSantiago/BaalChIP/issues/1, or mute the thread https://github.com/notifications/unsubscribe-auth/AGp8R5ECeSBUJkjSVCgebgBV13G-otfDks5rtEu5gaJpZM4M1CZy .

PoisonAlien commented 7 years ago

Got it! just wanted to make sure.

I have another question. How do you choose heterozygous SNP's (from your vignette) ? I was planning on calling variants and choosing all high QC heterozygous variants, but in the example it seems all are dbSNP entries. How were they chosen ? Sorry if this has been touched in vignette or in manuscript and I missed it.

InesdeSantiago commented 7 years ago

good! :-) I chose only the ones on dbSNP because I wanted to get their coordinates and the data I had (downloaded from ENCODE) did not included SNP coordinates. But you can chose heterozygous SNPs in any way that makes sense to your study as long as you organise them in the right format (ID CHROM POS REF ALT and RAF(Raf is optional)). Best, ines

On Thu, Apr 6, 2017 at 11:17 AM, PoisonAlien notifications@github.com wrote:

Got it! just wanted to make sure.

I have another question. How do you choose heterozygous SNP's (from your vignette) ? I was planning on calling variants and choosing all high QC heterozygous variants, but in the example it seems all are dbSNP entries. How were they chosen ? Sorry if this has been touched in vignette or in manuscript and I missed it.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/InesdeSantiago/BaalChIP/issues/1#issuecomment-292130906, or mute the thread https://github.com/notifications/unsubscribe-auth/AGp8R0v_VNR-zpY5d8EJdUGg62CK4JFDks5rtLvLgaJpZM4M1CZy .

PoisonAlien commented 7 years ago

Got it :)