no signal peptides found

[Majd Abdulghani]

Hi Kristoffer, 

Thank you for such a fantastic tool and great documentation.

When I try to import SignalP results, I encounter the following issue:
> analyzeSignalP(ssSwitchList,
+                pathToSignalPresultFile = "external_tools/signalp/prediction_results.txt")
Error in analyzeSignalP(ssSwitchList, pathToSignalPresultFile = "external_tools/signalp/prediction_results.txt") :
  No signial peptides were found

I used the standalone version and this is the command I ran:

signalp6 --fastafile isoformSwitchAnalyzeR_isoform_AA_complete.fasta \

--organism eukarya --output_dir external_tools/signalp \

--format none --mode fast

In the ISAR documentation, you said that when running the standalone version, -f summary should be used, but no such option exits. Based on the SignalP-6.0 documentation, I determined that '-f none' produces the same output as the webserver with the 'short' output format. Here is the SignalP standalone documentation:

- `--format`, `-fmt`, can take the values `txt`, `png`, `eps`, `all`, `none`.

It defines what output files are created for individual sequences.

`txt` produces a tabular `.gff` file with the per-position predictions for each sequence.

`png`, `eps`, `all` additionally produce probability plots in the requested format.

`none` only writes the summary prediction files.

For larger prediction jobs, plotting will slow down the processing speed significantly.

I attached my output file, which does have signal peptides, and I double checked that they exist in the SwitchList.

I would really appreciate your help with this!

Thank you,

Majd

[Majd Abdulghani]

I also tried the webserver (short, eukarya, fast) and fed the attached results file to ISAR, but still got the same error, even though there are signal peptides identified. 

[Kristoffer Vitting-Seerup]

This error just means that there were not detected any signal peptides in the isoforms your analyzed. I can see how the error message could be more informative and I apologize.

[Majd Abdulghani]

Hi Kristoffer,

Thank you so much for getting back to me quickly! And please don't apologize- IsoformSwitchAnalyzeR is amazing!

I get that's what the error could indicate, that's why I checked that some of the signal peptides are present in the SwitchList before posting the question. 

Do you know why signal peptides could be identified by SignalP that come from isoforms in the SwitchList but the error still gets raised?

[Kristoffer Vitting-Seerup]

Unless you think something very strange is going on the most likely explanation is simply that you are analyzing isoforms that are all "just" in the cytosol and hence have no signal peptides.

If you want, you can predict sub-cellular locations via deeploc2 and try importing that into IsoformSwitchAnalyzeR as a sanity check.

[Majd Abdulghani]

Sorry, I think there might be some confusion on my end. I ran the amino acid file through SignalP 6.0 and the results showed there are signal peptides present in many of the isoforms, so the issue is not that there aren't any signal peptides in the results (Which is what I think you're suggesting). I double checked that the isoforms identified as having a signal peptide are present in the switchAnalyzeRlist. But when I ran analyzeSignalP(), I got the message that no signal peptides were found. Am I missing something?

Thank you so much for your help!

[Kristoffer Vitting-Seerup]

Ahh - yes I can see there are some in the SingalP output.

Unfortunatly I don't currently have the bandwidth to investigate this problem. You are more than welcome to make a pull request if you find a bug and else I will keep it in mind for the next time I'm working on IsoformSwitchAnalyzeR.

Thanks for reporting this problem :-)

[Majd Abdulghani]

Totally understand. Good luck with everything!

[Majd Abdulghani]

In case someone else ran into the same issue: the output from SignalP 5.0 worked. Just don't use SignalP 6.0.