Question: Allele specific expression
gravatar for monaallouba
12 months ago by
monaallouba0 wrote:

Hi All, I would like to analyze the allele-specific expression in one sample at a specific het site (using rna seq data). I am not sure which of these two equations is correct. Could you please help me

  1. Equation one: alt reads/total reads
  2. Equation two: | 0.5 - alt reads/total reads |

I found equation two in several papers but I don’t know what the 0.5 means

Thank you very much

ADD COMMENTlink modified 12 months ago • written 12 months ago by monaallouba0

Thank you very much for your response @geek_y :)

The problem now is that my BAM file shows the the ref and alt reads but this site is completely absent in the VCF :/..Any ideas how I can resolve that? or can I just determine the number of ref and alt reads from the BAM file?

ADD REPLYlink written 12 months ago by monaallouba0

Please use ADD COMMENT/ADD REPLY when responding to existing posts to keep threads logically organized. This comment show go under @geek_y's answer.

SUBMIT ANSWER is for new answers to original question.

ADD REPLYlink written 12 months ago by genomax85k

How did you get that VCF file ? Genotyping array ?

ADD REPLYlink written 12 months ago by geek_y11k

no targeted sequencing

ADD REPLYlink written 11 months ago by monaallouba0
gravatar for geek_y
12 months ago by
geek_y11k wrote:

Thats the deviation from the expected 0.5.

If you have 100 reads overlapping a het SNP, if 25 coming from ref and 75 coming from alt, Allelic Ratio (AR) w.r.t ref allele is ref/total

25/100 = 0.25 . So the deviation is 0.5-0.25=0.25

If there is no allelic imbalance, the ref reads would be 50 and alternate reads would be 50, AR is 0.5 and deviation would be 0.5-0.5=0, i.e no allelic imbalance.

I would just do ref/total to have AR. No need to do (0.5-ref/total). You need to do a binomial test to see the "significant" deviation.

PS: It doesnt matter if numerator is ref or alt count. When you interpret, it should be w.r.t what you use as numerator. Like in gene expression fold change, depending on if you compare condition1 vs condition2 or condition2 vs condition1

ADD COMMENTlink modified 12 months ago • written 12 months ago by geek_y11k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1116 users visited in the last hour