CheckM2 missing files and result inconsistencies

[gkatahon]

CheckM2 missing files and result inconsistencies

[chklovski]

Hi,

CheckM2 does not use (single-copy) marker genes but a full set of KEGG protein annotation, as well as total counts of each amino acid, number of coding sequences, and length of genome as inputs.

As a result, small variations in these inputs can produce minor variations in output as is seen here. As CheckM1 uses only single-copy marker genes, it is less likely to vary with small bin changes.

Functionally your bin is 61% complete and 0.5% contaminated as estimated by CheckM2. Variations of <1% at this level of completeness is far below the error variance of CheckM2 (and even more of CheckM1) and is functionally meaningless (well below the resolution capacity of either tool).

You can also check the 'diamond_output' folder which contains all KEGG annotations identified in any given bin/contig. If the KEGG annotations are the same, then the difference is down to minor changes in genomic properties (amino acid distribution, etc). Hope this answers your question.

[chklovski]

Closed due to inactivity.