Some minor bugs in the catalog

[jeremyButtler]

Thank you for your quick response on my last issue.

This is not an issue, but more a couple errors in the 2023 database I noticed while converting it. I am only mentioning these to make sure that you are aware of them.

The first is that there are entries in the master catalog that have a "See genome indices" note, but that can not be found in the genome indices tab. These are all in the form of "gene_deletion". Here are the four I know of. All of these four entries are resistant to at least one antibiotic (I ignored anything that did not confer antibiotic resistance, so there might be more there).

• fgd1_deletion
• ethA_deletion
• katG_deletion
• pncA_deletion

The other minor issue I found is the the "gene_LoF" variants have no unique identifier in the genome indices tab. I can find them in the genome indices tab, but I have no way of telling them apart from other LoF's for the same gene. I suspect that these LoF variants get duplicated when there are multiple "gene_LoF" tags for the same gene.

Sorry about being the person who is raising issues.

Thanks for all the hard work you all do.

[sachalau]

Dear Jeremy,

Thank you for your report.

Did you have a look at the Instruction of use for these files available as PDF in the same folder?

https://github.com/GTB-tbsequencing/mutation-catalogue-2023/blob/main/Final%20Result%20Files/Instruction%20of%20use%20for%20incorporation%20of%20the%20mutations%20catalogue%20version%202%20results%20into%20bioinformatic%20pipeline.pdf

Quoting:

We do not provide genomic-variants for deletion graded-variants (for instance, for the graded-variant “pncA_deletion”)

So you are right that the deletion graded variant entries do not appear in the genomic_coordinates sheet.

Regarding your second issue, it is correct that gene_LoF have more than one genomic_coordinate entry in the second sheet (as a matter of fact, not only gene_LoF entries are duplicated, there are plenty of missense variant that might appear more than once as well).

Quoting again :

Each graded-variant can be linked to more than one genomic-variant (for instance, different genomic-variants can lead to identical missense)

For LoF graded variants, this is expected because any start_gained, frameshift or start_lost variants are included in those entry. So for instance, pncA_LoF will appear many times in the genomic_coordinate file because all these variants start_gained, frameshift and start_lost that we have observed in our database will be there. So you will know if you identify one in your data that it's a pncA_LoF and you can then match it to the appropriate grading.

I hope that resolves your concerns

[jeremyButtler]

Thanks for your reply

I did read the pdf, but was more focused on figuring out how to parse the database. Also at that time I did not have enough of a grip of the naming system or the catalog to understand that section.

Thanks again for taking the time.