Help running pipeline that creates GoSlim terms for Seastar Protien seq

[hannahnowers]

I am using this pipeline(https://github.com/sr320/workflow-annotation.git) to annotate seastar protein sequences from a .fa file. I am having trouble figuring out why it won't create the Go-slim and Go-ontology tables. From my first run of the pipeline, in my output directory, I can see three files called annotation_full_go.tsv, derm-protein.blast.tsv, postprocess_uniprot_go.py but it's not finishing through to create the full output with the Go-Slim terms and the Go ontology.

This is the error message I am getting:

File "/usr/lib/python3.12/urllib/request.py", line 639, in http_error_default
raise HTTPError(req.full_url, code, msg, hdrs, fp)
urllib.error.HTTPError: HTTP Error 403: Forbidden

Do you have any suggestions for how I could fix this?

[github-actions]

Thanks for posting an issue!

[kubu4]

When you have a moment, can you please provide a link to the input file(s) you're using?

Also, please provide the command(s) you used to initiate the anlalysis.

[hannahnowers]

This is the input file I am using: https://raw.githubusercontent.com/hannahnowers/Seastar-capstone/refs/heads/main/data/derm-protein.fa

Let me know if that link doesn't work or if I need to send a different type!

Here is the command I am using: nohup bash workflow-annotation/blast2slim.sh -i data/derm-protein.fa --protein > blast_run.log 2>&1 &

Thank you for your help!

[kubu4]

Okay, I may have tracked this down.

I'm assuming you're working on Raven?

I found this in your log file:

/home/shared/8TB_HDD_03/hnowers/Seastar-capstone$ tail -n 100 blast_run.log 
nohup: ignoring input
[INFO] Output directory: output/run_20260303_201239
[INFO] Running BLASTP...
Warning: [blastp] Examining 5 or more matches is recommended
deps ok
Traceback (most recent call last):
  File "/home/shared/8TB_HDD_03/hnowers/Seastar-capstone/output/run_20260303_201239/postprocess_uniprot_go.py", line 105, in <module>
    urllib.request.urlretrieve("http://purl.obolibrary.org/obo/go/go-basic.obo", obo_basic.as_posix())
  File "/usr/lib/python3.12/urllib/request.py", line 240, in urlretrieve
    with contextlib.closing(urlopen(url, data)) as fp:

I visited Gene Ontology website. The link to download the go-basic.obo is:

https://purl.obolibrary.org/obo/go/go-basic.obo

Notice that the link I pasted above is prefixed with https, while the link in the URL referenced in blast2slim.sh is prefixed with http.

That could explain the error you got:

urllib.error.HTTPError: HTTP Error 403: Forbidden

Unfortunately, I'm unable to edit that script directly.

I've submitted a pull request. Once @sr320 has merged that in, you'll need to incorporate the changes in your copy of the repo on Raven. Once you've done that, then you can try re-running and see if that resolves the issue.

[kubu4]

incorporate the changes in your copy of the repo on Raven

You can do this like so:

1. cd /home/shared/8TB_HDD_03/hnowers/Seastar-capstone/workflow-annotation

2. git pull

[sr320]

I have merged pull request

[hannahnowers]

Thank you!

[sr320]

@kubu4 if you have not already done so can you run the fasta @hannahnowers provided and see if you get any errors / get results?

[kubu4]

I ran this:

bash blast2slim.sh \
-i derm-protein.fa
--protein
> blast_run.log 2>&1

No errors, but also no results.

• blast_run.log, derm-protein.blast.tsv and annotation_full_go.tsv are empty

---

FYI, here's conda environment setup:

mamba create \
-n sr320-workflow-annotation \
python=3.11 \
diamond \
requests \
pandas \
goatools \
matplotlib \
-c bioconda \
-c conda-forge \
-y

[sr320]

readme! did you read the readme? --protein

[kubu4]

did you read the readme

Ha! No. I was just replicating what Hannah ran.

I'll look through it.

[kubu4]

I didn't even look at the FastA file. Don't know what I'm working with.

I'll spend more time figuring out what all of this is and then re-run it.

Would be a cool feature if code could throw an error when incorrect flag is used.

[kubu4]

did you read the readme? --protein

I guess I'm confused. The FastA is a protein FastA.

Looks like I made a rookie mistake and left out a continuation slash:

bash blast2slim.sh \
-i derm-protein.fa
--protein
> blast_run.log 2>&1

Fixed and am re-running:

bash blast2slim.sh \
-i derm-protein.fa \
--protein \
> blast_run.log 2>&1

And, that is working. derm-protein.blast.tsv has results in it. Will report back when it's finished.

[hannahnowers]

Just for some additional context, I had a semi-successful run of the pipeline last night. I incorporated the changes to the pipeline, but had a similar thing happen where it never created the go-slim files.

In my directory for the run, I only had go-basic.obo and not the goslim_generic.obo. I used a web link to manually download the goslim_generic.obo file into my run directory. I then ran it just from the post-processing script and was able to generate the annotation_with_goslim.tsv and summary_stats.json. This is what it looks like when I open the summary stats:

And these are the files it generated in my run folder. They don't quite match what was shown in the readme for the annotation pipeline. My output is missing summary.md and goslim_chart.png.

[kubu4]

@sr320 - Looking through the initial BLAST results table, I notice that there are multiple hits per query. Is that desired?

Looking through the code, I noticed the -max_hsps 1 option isn't set - which would limit the BLAST results to a single hit per query. Don't know if this makes any difference downstream yet, but figured I'd point it out.