alignment to a special part

Question

[alirezamomeni707]

I have a RNA-seq data and want to get the proportion of reads that map to a window of 100pb between 5'UTR and CDS (50 bp at 5'UTR and 50 bp at the CDS). since I am using Tophat to align and bowtie to build indexes, I guess I have to get the sequence of that part to build indexes. do you guys know how to find the sequence of that part?

[Devon Ryan]

You have BAM files and a GTF or something like that already, yes?

[alirezamomeni707]

yes I have bam file. I can also download GTF FILE.

Answer 1

[Istvan Albert]

I would advise against creating a custom genome index that only contains the intervals of interest. The alignments are most likely to be negatively affected.

Instead, you should align your reads against the full genome then intersect these resulting alignment with a custom interval file that contains only the regions of interests.

Tools like bedtools or bedops will allow you to do both, create upstream regions relative to the intervals of interest as well as to intersect the alignment files.