To set up APE-Gen2.0 locally, download the repo using git clone.
Before you move on to anything else, am important step must be performed: Download the frequency matrices generated using MHCFlurry2.0 through the link below:
https://rice.box.com/s/duxshqxtkykg7u9y3b0tl79j8byc2dp2
These will be crucial for the softawre to run. After these are downloaded, go ahead and put those on the helper_files folder. Then you're all set for the installation!
We recommend that APE-Gen2.0 is set up with a docker configuration, so the latest version of docker should be installed on your system. Setting APE-Gen2.0 through a python virtual env is definitely plausible (check the environment.ymland the provided Dockerfile for all needed packages), but we'll leave that to you if you want to set it up this way.
Just running:
bash build_image.sh
should build the docker image After this, just running:
bash start_image.sh
gets you into the container, APE-Gen2.0 can be run from there.
To run APE-Gen2.0 for existing MHC templates in the database, you should be good to go (Ignore the warning message about the MODELLER invalid key). However, for a new MHC not found in the database, you will need to provide a valid MODELLER key. To do that, go to:
cd /conda/envs/apegen/lib/modeller-10.4/modlib/modeller/
and modify the lines of config.py with a valid MODELLER key.
For now, peptide sequence (along with positions with PTM) + an MHC allotype should work! (more inputs to follow...)
python New_APE-Gen.py LLGIGSLTV HLA-A*02:01 --verbose
python New_APE-Gen.py LLGIGSLTV HLA-A*02:01 --verbose --score_with_openmm
python New_APE-Gen.py LLGIGpSLTV HLA-A*02:01 --verbose --score_with_openmm
python New_APE-Gen.py LLGpSGpSLTV HLA-A*02:01 --verbose --score_with_openmm
Please contact the team should you have any issues with running APE-Gen2.0!
@article{Fasoulis2024,
title = {APE-Gen2.0: Expanding Rapid Class I Peptide–Major Histocompatibility Complex Modeling to Post-Translational Modifications and Noncanonical Peptide Geometries},
volume = {64},
ISSN = {1549-960X},
url = {http://dx.doi.org/10.1021/acs.jcim.3c01667},
DOI = {10.1021/acs.jcim.3c01667},
number = {5},
journal = {Journal of Chemical Information and Modeling},
publisher = {American Chemical Society (ACS)},
author = {Fasoulis, Romanos and Rigo, Mauricio M. and Lizée, Gregory and Antunes, Dinler A. and Kavraki, Lydia E.},
year = {2024},
month = feb,
pages = {1730–1750}
}