only CAZy annotation information in distilled files

[gruningerrj]

I appear to be having the same issue as #53 with richly annotated MAGS but only CAZy annotation information in the final distillate. I note the this was solved by adding KOs to the KEGG descriptions however I am not sure what files to find this information in and the easiest way to add this information to the appropriate files. I would be grateful if anyone could help me figure out how to do this.

Thanks

[shafferm]

When you set up DRAM did you give it the location of your KEGG proteins file or are you using KOfam?

[gruningerrj]

yes. I gave it to the location file prokaryotes.pep in the KEGG database. A colleague that use KOfam before we had access to KEGG didn't have any trouble. I am not sure if I used the wrong KEGG protein file?

[shafferm]

This maybe because of a change in format of the KEGG pep file headers. We haven't renewed our KEGG subscription for a bit over a year so that could be causing the issue. Could you run grep '>' prokaryotes.pep | head and share the output? You could potentially fix this issue by providing the --gene_ko_link_loc flag during set up. This is a file that has all gene IDs and KO IDs in a two column file. I can't remember where in the KEGG flat file database it's stored but I think it's called something like genes_ko.list.gz. I could help you rerun the processing of KEGG with that file added so that you don't need to rerun all of it if that would help.

[gruningerrj]

Here is the output from prokaryotes.pep

eco:b0001 thrL; thr operon leader peptide eco:b0002 thrA; fused aspartate kinase/homoserine dehydrogenase 1 eco:b0003 thrB; homoserine kinase eco:b0004 thrC; threonine synthase eco:b0005 yaaX; DUF2502 domain-containing protein YaaX eco:b0006 yaaA; peroxide stress resistance protein YaaA eco:b0007 yaaJ; putative transporter YaaJ eco:b0008 talB; transaldolase B eco:b0009 mog; molybdopterin adenylyltransferase eco:b0010 satP; acetate/succinate:H(+) symporter

The format of the genes_ko.list is below grep 'eco:' genes_ko.list | head eco:b3957 ko:K01438 eco:b3958 ko:K00145 eco:b3959 ko:K00930 eco:b3962 ko:K00322 eco:b3968 ko:K01977 eco:b3970 ko:K01980 eco:b3971 ko:K01985 eco:b3980 ko:K02358 eco:b3981 ko:K03073 eco:b3982 ko:K02601

[rmFlynn]

The use of kegg is considered advanced still, but we have tools that can be used to build compatible pep files. I hope this process will become more streamlined in the future. Until then, I will call this issue closed, as it is not relevant to the current code base.

[mw55309]

It might be good to show somewhere in the documentation what you think the kegg.pep headers should look like, particularly for genes that more than one KO :)