Entering edit mode
3.3 years ago
qliang2016
•
0
Hi community, I have a very simple question that, if I want to get the allele-specific expression of a gene from RNA-seq, what is the simplest way to do it. I know there are tools such as GATKAlleleCounter, but they are counting at each SNP sites, based on the vcf file. Thus if two or more SNPs are in one read, it will be counted multiple times if I simply add the count for each SNP within a gene body. Is there any tool that do the gene-level ASE aggregation? The phasing is known in my case. Thanks!
MBASED can generate gene-level ASE with/without phasing information (paper, R package).
Thanks. I read the paper and actually, it also did not try to avoid multiple counting. Also it seems that the authors did SNV which I guess very rarely have two in one read (say 75-150 bp short reads). For SNP there are more sites and one read covering multiple SNPs are not as rare. For SNP, my feeling is that for most cases, if there is real significant differences between two alleles, multiple counting will kind of enhance the result (ratio) but not change the 'fact' (true differences); and if robustly, there are no differences at all, then the multiple counting is fair for both alleles. This could be the reason people do not care so much about the true number of reads for each allele. All multiple counting would affect is marginal cases.