Coverage analysis for any assembler type

[cerebis]

Currently, bin3C cheats to obtain coverage information by relying on SPAdes contig naming convention.

This is obviously brittle but we only use this information to report it to users per-cluster. Later, it would be interesting to add the step of identifying outliers within clusters and offering them as a per-cluster analysis target. This has been done manually in the literature and outliers are enriched for plasmids (and you would expect other mobile elements or conserved sequence).

With that feature implemented, handling any assembler would be important.

[cerebis]

Added 3 possibilities, "generic", "spades" and "megahit".

• Generic assumes the information is not available and excludes it from reports.
• Megahit also embeds this information in the FASTA headers.
• SPAdes as before

[cerebis]

Additionally, users can now supply coverage information as a separate table as a tab-delimted table.

To produce this table, we reccommend using CoverM. In our work, the metabat calculation format is fine (--methods metabat), while users could experiment with other methods at their choosing.

{contig_id}\t{coverage-value}