Prediction of B-cell epitopes from amino acid sequences using deep neural networks. Supported on Linux and Mac.

8 GB RAM should be available. With 8GB even processing protein sequences longer than 6000 amino acids and/or multiple hundreds of sequences shouldn't be problematic.

1. Download and install Conda.

2. Create a Conda environment with Python 3.7

   conda create -n epidope python=3.7

3. Activate the Conda environment. You will need to activate the Conda environment in each terminal in which you want to use epidope.

   conda activate epidope

4. Install epidope via conda

   conda install -c flomock -c conda-forge -c pytorch epidope h5py=2.10 jsonnet 

   Note: While installation with conda, the loading bar of epidope is not working. So depending on your internet connection, it can take from a few seconds too minutes until you see any progress.

5. Install other dependencies

   pip install allennlp==0.9.0

Example

epidope -i /path_to/multifasta.fa -o ./results/ -e /known/epitopes.txt

Options:

We also provide a Docker image for EpiDope.
Simply pull and run a ready-to-use image from Dockerhub:

docker run -t --rm -v /path/to/input/files:/in -v /path/to/output:/out \
flomock/epidope:v0.2 -i /in/proteins.fasta -o /out/epidope_results

(you need to mount files/folders that you want to access in the Docker via -v)

Or if you want you can build the image yourself locally from the Dockerfile in this repo:

docker build -t epidope .

Run as non-root user under linux:

docker run -t --rm -v /path/to/input/files:/in -v /path/to/output:/out -u `id -u $USER`:`id -g $USER` \
flomock/epidope:v0.2 -i /in/proteins.fasta -o /out/epidope_results

Run docker with a different memory allocation see System requirements (default is 2GB for linux and mac):
(e.g. 8GB)

docker run -t --rm -v -m=8g /path/to/input/files:/in -v /path/to/output:/out \
flomock/epidope:v0.2 -i /in/proteins.fasta -o /out/epidope_results

If you are interested, you find most of the code which was used to create this tool under:

https://github.com/flomock/epitop_pred

Maximilian Collatz, Florian Mock, Emanuel Barth, Martin Hölzer, Konrad Sachse, Manja Marz, EpiDope: A Deep Neural Network for linear B-cell epitope prediction, Bioinformatics, , btaa773, https://doi.org/10.1093/bioinformatics/btaa773