search

alexdobin / STAR

RNA-seq aligner
MIT License
1.77k stars 495 forks source link

discordant not flagged #2161

Open hwlim opened 4 days ago

hwlim commented 4 days ago

Hi

I am aligning paired end ChIP-seq data using STAR 2.7.4 version with options:

--alignSJDBoverhangMin 999 --alignIntronMax 1 --alignMatesGapMax 2000 --outFilterMultimapNmax 100 --outFilterMismatchNoverLmax 0.05 --outReadsUnmapped None --alignEndsProtrude 0 DiscordantPair

Hoping to distinguish discordant vs concordant alignments. I believe read aligned to "+" downstream of its mate aligned to "-" is discordant. However, I don't see any discordant pairs. I suppose the example below should be discordant? but the flag says they are properly paired.

SAM records

NB501763:286:HV72YBGXB:1:11102:25654:18459 419 chr10 38489187 1 22S16M = 38489181 30 ATGGAATGGAATGGAATAATCCGGTATGGAATGGAATG AAAAAEEEEE6EEEEEEEEEEEEEEE/EEEEEEEEEAE NH:i:4 HI:i:4 AS:i:49 nM:i:1 NB501763:286:HV72YBGXB:1:11102:25654:18459 339 chr10 38489181 1 36M2S = 38489187 -30 TGGAATGGTATGGAATGGAATGGAATGGTACGGAATAG EEEEAEEEEEEEEEEEEEEEEEEEEEEEEEEEEAAAA6 NH:i:4 HI:i:4 AS:i:49 nM:i:1

BEDPE format

chr10 38489180 38489216 chr10 38489186 38489202 NB501763:286:HV72YBGXB:1:11102:25654:18459 1 - +

Or is this a correct report? THanks.