Improving contig/MAG abundance calculations by storing total number of reads in metagenomes in profile databases

[tdelmont]

The need

Get the relative abundance of MAGs across samples. We have a lot of confusion (rightfully in my opinion) within the anvi'o community about the abundance numbers in the summary output.

The solution

Offer the option to incorporate a text file with sample.name as column A and number.of.metagenomic.reads as column B into PROFILE.db.

Then, during summary the relative abundance of MAGs would be calculated properly. The part would be left blank for those that did not import this table. A warning during anvi-summarize would be nice to remind the user of this option.

Beneficiaries

All of those doing genome-resolved metagenomics or metapangenomics, and interested in the portion of reads that mapped to genes or genomes.

Just a though, based on a recent comment in the anvi'o slack.

Best

Tom

[meren]

When people use the anvi'o Snakemake workflows, the actual number of reads are stored in the misc data tables for single and merged profile databases automatically under the name total_num_reads.

The same mechanism can be used to import a misc data layers information via the program anvi-import-misc-data.

We could implement a new routine in anvi-summarize that could take into consideration those numbers to calculate abundance once and for all. Actually we could use the same for even anvi-profile.

I will leave this here with this information, but I this will be post-v7.

Thank you for submitting this, Tom! :)