Unable to use prospr with multiple issues

Hello,

Thank you for releasing this! I’m excited to use it, but have have been running into a lot of issues in trying to.

I attempted using it both with and without a docker container, including building a docker image based on the updated code on the master branch of this repo. I also attempted downloading the plmDCA package from FTP site mentioned on the README.

All of my attempts ended in failure.

Using the off-the-shelf docker container also doesn’t work:

docker run -it -v /home/dangerzone/data/:/data prospr/prospr build example_protein

Which resulted in the following error.

FileNotFoundError: [Errno 2] No such file or directory: '/usr/local/bin/hhmake -i /data/example_protein/example_protein.a3m -o /data/example_protein/example_protein.hhm': '/usr/local/bin/hhmake -i /data/example_protein/example_protein.a3m -o /data/example_protein/example_protein.hhm'

This error is simple enough to fix (and I think it is fixed in the version of hhblits.py in the master branch), but I ran into basically every one of the other open issue tickets in this Github repo. Most of them having to do with the plmDCA dependency.

I have two concrete questions: Are there any plans to release an updated docker image? Is it possible to not use plmDCA and it’s dependencies? https://github.com/KIT-MBS/pydca might be a good starting point.

[yamule]

https://github.com/yamule/prospr/tree/sep I share the Prospr files I’m using. It accepts input files separately, therefore you may run it with pydca.

[wbillings]

We are sorry about all of the dependency issues. We plan to release a complete update within the coming months that will hopefully improve usability by removing MATLAB and other problematic requirements. It will also include a heavily restructured ProSPr network, updated model files, and improved results.