Weighted standard deviation of AGS

[TlaskalV]

Hi Ben,
I am running microbial_fraction with few metagenomes and it is great. I was thinking that it should be possible to include the weighted standard deviation of the average genome size in the calculation below, right? This would help to get an idea of the variance in genome sizes.

singlem/singlem/read_fraction.py

Line 168 in f9045bb

average_genome_size = average_genome_size_numerator / average_genome_size_denominator

Thank you for considering such feature.

Cheers,
Vojtech

[wwood]

Thanks for the suggestion @TlaskalV

We had considered this, but there's a few negatives to this approach. The main thing is that genome size changes are just one of the uncertainties. Another is the coverage of each lineage in the profile, or uncertainty in the taxonomic assignments.

Having uncertainty measures just from genome sizes could therefore be a bit misleading to the user, if you follow?

Thanks, ben

[TlaskalV]

Thanks for taking the time to review my suggestion and explaining your reasoning. While I still lean toward the variance idea, I get where you're coming from - looking forward to continuing to use the microbial_fraction.