Check is RNA-seq mapping (tophat) is performed ok (hg38)?

Question

[manekineko]

Hi,

I have used default options in Galaxy (just set the Mean Inner Distance between Mate Pairs to 50 which I did not sure it is ok) to map my RNA-seq library to Human genome hg38. Is there a way to check is the mapping good and ok, for example check a housekeeping gene or a gene what I can see how should look like in IGV?

[Nicolas Rosewick]

just open your bam file in IGV and check several genes. Don't forget to index it before open it in IGV (samtools index alignment_file.bam)

[manekineko]

I see this - http://s22.postimg.org/4s9xmq4u9/Screen_Shot_2015_07_27_at_17_51_47.png

but cannot decide if it is normal or not for RNA-seq (as I work mainly with miRNAs not with RNA-seq so far :)

[Nicolas Rosewick]

Seems OK to me. Are your data stranded?

[manekineko]

R1 read is forwarded and R2 is reverse.

I mainly was worried for setting mean-inner-mate option in tophat as I have problems calculating it. I just know that my reads are 101bp, and the RNA was selected from fragments 300-400.So do not know how much a wrong -r option will harm the mapping. I see from the mapping there is a lot reads overlapping mates

Answer 1

[Dan D]

For a comprehensive check, I recommend using Picard's CollectRnaSeqMetrics tool.

IGV, like NicoBxl suggested, is a great way to check individual genes. Galaxy also has its own visualization solution called Trackster.