-max-reads 0

[rotheconrad]

Hello! I hope you are well. simka looks like a great tool!

How are the samples normalized with the -max-reads 0 flag? I did not see a description of this in the paper.

Have you considered normalization options such as suggested here:
https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1003531 ?

Or transformation options suggested here:
https://www.frontiersin.org/articles/10.3389/fmicb.2017.02224/full ?

Since the default is not to normalize, is the intended workflow to subset all samples to the same number of reads prior to running simka?

Have you tested how much the size discrepancies actually affect the various distance metrics?

Thanks for the clarification.

best,
Roth