I know it is a common standard to use these metrics with metabarcoding data, when we have an OTU abundance matrix between samples and we want to compare the microbial community shape between conditions. But I was wondering if we could do the same with a MAGs (metagenome assemblies-genomes) abundance matrix obtained from shotgun metagenomics data.
In short, I reconstructed the MAGs after binning the contigs in my assembly with various binning tools. Then, I aligned the cleaned raw reads from my samples with the contigs belonging to the different MAGs, which allows me to know the number of reads belonging to the different genomes in my dataset. After that, I normalize the number of reads by the length of the contigs, to get the average coverage per MAG between my samples.
I thus finally have a MAGs-coverage matrix with the samples in column and the MAGs in raw. So the structure is the same as an OTU abundance matrix derived from metabarcoding data, and I want to compare my different samples to potentially show patterns between my biological conditions.
I was thinking of using for example the Bray-Curtis index to calculate distances between my samples, but is this method correct with a MAG-coverage matrix?
If you have any advice for me, I would be very grateful.
