Hello Community! , I need to estimate read abundances across metagenome assembled genomes (MAGs) using coverM.

My question is regarding to what is considered a good practice when mapping paired end reads against MAGs that were produced by coassembly of reads of more than 1 sample.

For my understanding, when a contig is made by coassembly (using de Bruin graphs) it is likely that this contig is the result of the assembly of reads that came from more than 1 sample. So having this in mind when mapping paired-end reads against the contigs of a MAG , should I consider an abundance of 1 or a count when both pairs of reads map against a contig? or should I be more permissive and consider a corrected aligned read when only one of the pairs is mapped against a contig?