Stop codons in ORFs

[deminatanja]

• phold version: 0.2.0
• Python version: 3.11.9
• Operating System: Linux

Description

I have used phold to annotate some virus sequences obtained from a metagenome. The run was alright and there is an expected output. However, ORFs were predicted so that they contain stop codons and don't always start with M. Please explain if I am missing something here, but I would expect not to have stop codons within ORFs.

What I Did

basic command used:

phold run -i xx.fasta -o xx_output_phold -d /path-to-db/ -t 12

[gbouras13]

Hi @deminatanja ,

Do you mean stop codons within the ORFs (or just at the end?) If you see them within, please provide me an example as this is very much not desired behaviour. Phold (the way you ran it) will use prodigal-gv, which takes into account alternate translation tables (4,15) by default.

With the start codons, while M (AUG) is by far the most dominant, prokaryotes can use other start codons - see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5622118/ (GUG / V and UUG / L). So sometimes these are predicted by Prodigal-gv.

George

[deminatanja]

Hi @gbouras13,

Thanks for the answer! I think I found the problem: the annotation phold_per_cds_predictions.tsv table has coordinates different from phold.gbk: all first coordinates are shifted by + one nt, e.g. if the coordinates in .tsv are 2-184, they are written as 3..184 in .gbk That's why I see odd things like stop codons within ORFs when I visualize the annotations from phold.gbk. I could just change the starting coordinates by -1 in the phold.gbk output, but it is probably worth checking what has gone wrong within the run. Translations are correct in .gbk. Is there some troubleshooting I could try?

Best,
Tatiana

[gbouras13]

Hi @deminatanja - thanks for picking this up!

I will be making some major updates to phold in the next few weeks (hopefully) - so I will make sure I fix this error then for the next release. For now, please just change the coordinates for you purposes.

George