MHCBoost

Gradient boosted trees based predictor for MHC Class I epitope binding prediction. MHCBoost is also part of BERT, a very powerful deimmunization workflow.

How to install

Python 3.7+ is a requirement!

1. git clone https://github.com/Zethson/MHCBoost
2. python setup.py install
3. mhcboost

Usage

The CLI - Command Line Interface

> mhcboost
-p, --dataset_to_predict_path             <arg> file to perform prediction on OR peptide sequence
-o, --predicted_dataset_path              <arg> filepath to save the predicted output file to
-a, --allele                              <arg> allele to perform prediction on
-l, --peptide_length                      <arg> epitope peptide length - usually 9
optional -t, --training_dataset_path      <arg> file for classifier training
optional -s, --silent                     suppresses learning output

Examples

Simply provide the answers to the questions asked by our tool.

> mhcboost 

Alternatively, provide input parameters when starting the tool.

> mhcboost -p examples/example_input.txt -o /home/mypc/Desktop/output.txt -a A*02:01 -l 9

Supported Alleles

MHCBoost supports 65 alleles. HLA-A01:01
HLA-A02:01
HLA-A02:02
HLA-A02:03
HLA-A02:06
HLA-A02:11
HLA-A02:12
HLA-A02:16
HLA-A02:19
HLA-A02:50
HLA-A03:01
HLA-A11:01
HLA-A23:01
HLA-A24:02
HLA-A24:03
HLA-A25:01
HLA-A26:01
HLA-A26:02
HLA-A26:03
HLA-A29:02
HLA-A30:01
HLA-A30:02
HLA-A31:01
HLA-A32:01
HLA-A32:07
HLA-A32:15
HLA-A33:01
HLA-A66:01
HLA-A68:01
HLA-A68:02
HLA-A68:23
HLA-A69:01
HLA-A80:01
HLA-B07:02
HLA-B08:01
HLA-B08:02
HLA-B08:03
HLA-B14:02
HLA-B15:01
HLA-B15:02
HLA-B15:03
HLA-B15:09
HLA-B15:17
HLA-B18:01
HLA-B27:05
HLA-B35:01
HLA-B38:01
HLA-B39:01
HLA-B40:01
HLA-B40:02
HLA-B44:02
HLA-B44:03
HLA-B45:01
HLA-B46:01
HLA-B48:01
HLA-B51:01
HLA-B53:01
HLA-B57:01
HLA-B58:01
HLA-B73:01
HLA-B83:01
HLA-C05:01
HLA-C06:02
HLA-C14:02
HLA-C15:02

Performance

MHCBoost has an 5-fold crossvalidated average AUC of 0.899 on the IEDB dataset. The performance on each allele was compared to the state of the art NetMHCPan. Please refer to Results

License

MIT

Authors

Team iGEM 2018 Tübingen
Lukas Heumos
Steffen Lemke
Alexander Röhl