I am trying to upload an annotation file to NCBI. NCBI will run PGAP - their prokaryote annotation tool - but not if I have additional annotations. Since I do, from my own source, I am trying to combine them into a single upload file because NCBI will not take multiple. So, PGAP produces a .sqn file when I run it locally and it concludes, I use asn2gb to convert this into a .tbl file that I can combine with the .tbl file I produce through a custom script, and then tbl2asn creates the new .sqn file for upload. However, this file leads to a large series of BadProteinStart (and bad codon start) errors that are sourced from the PGAP annotation.
I believe this to be related to running both asn2gb and then tbl2asn causing a translation table error, but I haven't found a location I can set it to specify the prokaryote translation table (gcode=11). I have tried running tbl2asn with "-j gcode=11" but this was no different.
Does anyone have any experience with this? I can post the errors themselves, the discrepancy report, the error summary evaluation, etc. but I think everything relevant is laid out here already.
I will note that my custom annotations do not cause these errors. Further, running PGAP and converting it to tbl and back causes these errors on its own, without input from my annotations.
This might be because I was changing the header format that came from asn2gb to match what I was using for tbl2asn. This made sense because I was combining two tbl files and the information I was removing didn't seem significant, but was apparently an issue.
What had said ">Feature lcl|1" from PGAP was being replaced with ">Feature 1" as in my custom tbl. Now, I have not made that change and it appears to resolve the issue.