rob-p
commented
8 years ago Hi @CTLife,
The default value of 31 should work well for a wide range of organisms and read lengths. The only reason you might consider selecting a smaller value is to reduce memory usage (if you have e.g. a particularly huge de novo transcriptome), or if you're dealing with very short reads (e.g. <50 bases).
P.S. You mention genomes above, but please note that RapMap is designed for alignment to a transcriptome (a FASTA of all of the transcripts / RNA-products to which you wish to map). We strongly advise against building the index on the genome itself.
CTLife
rapmap quasiindex --help:
-k (positive integer less than 32), --klen (positive integer less than32), The length of k-mer to index.
But how to select the k for different genome, such as human, mouse and c.elegans ?
Thank you.