I am using salmon to evaluate how the L.donovani gene expression varies in the two different growth phases (promastigotes and amastigotes). For that I downloaded two RNA-seq data from SRA and I aligned those to the reference L.donovani coding sequence coding sequences (CDS)
In order to quantify gene expression with salmon I have to index the CDS using a specific k-mer value. In the own salmon manual they use a 31-mer for 75bp reads. My question is whether is reasonable to choose this same value i.e. 0,413reads length or if there's any other advisable value. I heard that some people use about 0,33reads length whereas I also found this post where is is metioned between 0,5 and 0,66*reads length (although this if rather for de novo genome assembly, which is not my case as I am mapping to a reference CDS.)
Thanks in advance.