How to increase the maximal number of soft-clipped bases

[holger brandl]

Hi,

is it possible to adjust the maximal number of soft-clipped bases? The questions has been asked on https://www.biostars.org/p/93782/ a while ago but without an actual answer.

The reason why I'd like to do so, is that I'm interested in revealing precise transcript-ends. To do so I've filtered my data for poly-A-tail-containing reads. However, to create reportable alignments for those tx-end-overlapping-polya-tail reads properly I would need to relax the soft-clipping limit of STAR.

I know that I cut use cutadapt to trim the polyA tails from the reads, but for sure just tweaking a STAR parameter is more appealing. :-)

Best,

Holger Brandl

[Alexander Dobin]

Hi Holger,

you can control the max allowed number of soft-clipped bases indirectly by controlling the minimum allowed number of matched bases and minimum score:

outFilterScoreMin               0

    int: alignment will be output only if its score is higher than or equal to this value.

outFilterScoreMinOverLread      0.66

    float: same as outFilterScoreMin, but  normalized to read length (sum of mates' lengths for paired-end reads)

outFilterMatchNmin              0

    int: alignment will be output only if the number of matched bases is higher than or equal to this value.

outFilterMatchNminOverLread     0.66

    float: sam as outFilterMatchNmin, but normalized to the read length (sum of mates' lengths for paired-end reads).

All these filters work in the AND fashion, i.e. the most stringent filter wins. For PE read, the filters deal with the total score/number of mapped bases for both mates

For instance, you can set --outFilterScoreMinOverLread 0 --outFilterMatchNminOverLread 0.51, which would allow alignments with number of mapped bases >= 51% of the read length (sum of both mates for PE reads).

I do not recommend going below 0.5 for PE reads as this will allow single end alignments.

Cheers

Alex