QC based on Samtools Flagstat

Question

[indranipoddar]

Hello,

It will be very helpful if somebody can please suggest me if the this a poor alignment or not and I need to exclude QC-failed reads.( I did read other posts on Flagstat but those examples doesn't have these many QC-failed reads).

I did a tophat alignment from paired end reads from Illumina and executed Samtools flagstat on the aligned BAM file,it shows this stats below:

117590453 + 26766923 in total (QC-passed reads + QC-failed reads)
17600285 + 3418020 duplicates
101724522 + 8960862 mapped (86.51%:33.48%)
117590453 + 26766923 paired in sequencing
58968738 + 13380987 read1
58621715 + 13385936 read2
93403124 + 5491156 properly paired (79.43%:20.51%)
93590388 + 5502264 with itself and mate mapped
8134134 + 3458598 singletons (6.92%:12.92%)
0 + 0 with mate mapped to a different chr
0 + 0 with mate mapped to a different chr (mapQ>=5)

Please suggest.
Thanks in advance

Answer 1

[Sukhi Singh]

The Meaning Of Samtools Flagstat Output Of Illumina'S Bam File

What Does Qc-Passed/Failed Reads Mean In Samtools Flagstat?

Nina and Pierre has an answer for you.

[indranipoddar]

Thanks Sukhdeep. I read those threads before. In my flagstat stats, I see so many QC failed and also properly paired is 79%.

Is this acceptable in a RNA -seq alignment. I mean which is the most important stat should I investigate before excluding anything.

Thanks