How to interpret duplicate output rows and negative percent identity (rRNA gene variant model)

[danrlu]

We can't tell whether this is a bug or it is intended and it's us not interpreting the results properly.

Describe the bug/confusion

• One AMR hit has 8 rows in the RGI main output.
• Only the ORF_ID is different and all other columns are identical, but the query contigs in the ORF_ID don't match the Contig column. I'm guessing this AMR hit was found in 8 different contigs, but wouldn't the Contig column show their corresponding contig names?
• Separately in the Percentage Length of Reference Sequence column, the number is -15.74 and < 0 (b/c the orientation of AMR gene and the contig are the opposite direction). This is the only time I see this column < 0 in all RGI outputs so wondering whether this is intended.

Input rgi main -i contigs.fasta -o contig_amr_report -t contig -a BLAST --include_nudge --clean

Input file https://github.com/danrlu/debug_data/tree/main/rgi

Error log No error.

CARD Version RGI version latest version of rgi CARD 3.2.6 (same result when uploading the contig.fasta to CARD website)

Thanks very much for this great tool and database and continuous improvements!!!

Originally noticed by @lvreynoso.

[raphenya]

Correct, 8 of the contigs hit model Neisseria gonorrhoeae 23S rRNA with mutation conferring resistance to azithromycin.

For the rRNA gene variant model the ORF_ID is named with the model and the query identifier .e.g., 3646_5918 | model_type_id: 40295 | pass_bit_score: 5300 | SNP: C2600T,A2145G,C2597T,A2045G,A2059G,C2611T | Neisseria gonorrhoeae 23S rRNA with mutation conferring resistance to azithromycin | QUERY: NODE_26_length_1251_cov_840.666388 NODE_26_length_1251_cov_840.666388

The Percentage Length of Reference Sequence calculation is incorrect, which we will fix. Cheers.