Does salmon use phred scores while pseudoaligning (and/or validating mapping)?

Question

[Anthony]

Dear Biostar Community,

I would like to know if salmon uses phred scores while pseudoaligning (and/or validating mapping)?

• I checked Patro R, Duggal G, Love MI, Irizarry RA, Kingsford C. Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods. 2017;14(4):417–419. doi:10.1038/nmeth.4197 and I did not find any reference to phred score;
• In the code, there is only one reference to phred and it does not seem to be used anywhere else.

In the negative, what is the reason behind this the choice?

Thanks in advance,

Anthony.

Answer 1

[Rob]

Hi Anthony,

No; salmon does not use the PHRED score, either while mapping or aligning. The main reason for not using it during alignment or mapping is that it just tends not to be very useful for this purpose. If a nucleotide inside a reasonably long exact match is bad quality (and therefore likely wrong), it’s almost certainly wrong in all of the mappings where it appears. In terms of finding the best locus (or loci) for the read, the quality score just turns out not to be very useful, while using it incurs extra processing cycles. Clearly, there _are_ cases where the quality scores are useful (e.g. in downstream variant detection, etc.) — however, we found them not to be helpful for our purposes and so we don’t use them.

Best, Rob