Problem with Proteinortho.pl

Question

[JV]

Since I can't find a specific "issues" page or contact adress for Proteinortho, I'll try posting my problem here...

I am currently using proteinortho v.5.16 for orhtolog detection. I ran it on a set of ~100 genomes and it worked fine. Then I realized that some of the input files were incomplete, and I replaced them (deleting only the blast-results for comparison pairs that include those input files). This has always worked with past versions of proteinortho and saved me the time of having to redo ALL blast-calulations (Proteinortho only redoes the blasts-combinations that are missing --> the ones I deleted).

Now however, I keep getting an error at the clustering step, terminating it because it threw an instance of 'char const*':

**Step 3**
Clustering by similarity (Proteinortho mode)
Reading projectA_tryagain_nopoff.blast-graph
97 species
375016 paired proteins
10950803 bidirectional edges
Clustering: 0.26%terminate called after throwing an instance of 'char const*'
Aborted (core dumped)
Adding singles...
[OUTPUT] -> written to projectA_tryagain_nopoff.proteinortho
Writing graph...
Cleaning edge list...
Removed 154480 / 11353468 edges
Done.
[OUTPUT] -> written to projectA_tryagain_nopoff.proteinortho-graph

It DOES generate an ortholog-table (the result I want to get), but I have no idea if i can/should trust the results.

Anybody encounter this problem with proteinortho and/or know what it means? How can I avoid this error?

Answer 1

[paulmjklemm]

So although this is allready 22 months old, i will give an update for all the people that still stumble across this:

There is a newer version of proteinortho available at gitlab: https://gitlab.com/paulklemm_PHD/proteinortho this should solve this problem. Also available via (bio)conda and brew. So just download the new proteinortho and this should be fixed.

Greetings

[lieven.sterck]

why not simply go for OrthoMCL or Orthofinder or such, they seem to do something very similar and are well maintained.