Metadata-Version: 2.1
Name: mhclovac
Version: 1.0.0
Summary: MHC binding prediction based on modeled physicochemical properties of peptides
Home-page: https://github.com/stefs304/mhclovac
Author: Stefan Stojanovic
Author-email: stefs304@gmail.com
License: UNKNOWN
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.5
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 2.7
Classifier: Programming Language :: Python :: 2
Classifier: License :: OSI Approved :: GNU General Public License v3 (GPLv3)
Classifier: Intended Audience :: Science/Research
Classifier: Intended Audience :: Healthcare Industry
Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
Classifier: Topic :: Scientific/Engineering :: Chemistry
Description-Content-Type: text/markdown
Requires-Dist: h5py
Requires-Dist: numpy
Requires-Dist: pandas
Requires-Dist: python-dateutil
Requires-Dist: scipy

# MHCLovac
MHC binding prediction based on modeled physicochemical properties of peptides

* [About](#about)
* [Installation](#installation)
* [Usage](#usage)

<hr>

<a name='about'></a>
### About 

MHCLovac is based on modeling each peptide as a **signal** of a certain 
physicochemical property. Each amino acid that 
 comprises a peptide encodes specific combination of physicochemical properties
  that determine physicochemical properties of the peptide at a given position.
  By modelling amino acids properties in a smart way, MHCLovac obtains a unique 
  physicochemical signal, or **profile** of each peptide 
  *(1.a)*. This 
is done for variety of physicochemical properties including hydrophobicity, 
  charge, isoelectric point, donor and acceptor hydrogen bonds etc. Finally, 
  by averaging out profiles of ligands known to bind certain HLA 
  type, an HLA-specific profile for each physicochemical property is obtained. 
  *(1.c)*

![physicochemical profile](https://raw.githubusercontent.com/stefs304/mhclovac/master/resources/plots/example.png)

  MHCLovac works by first building a *reference profile database*, containing 
  profiles specific to each HLA type. This database is compiled based on 
  experimentally confirmed ligands for each HLA type 
  ([resources/training_data](https://github.com/stefs304/mhclovac/tree/master/resources/training_data)) 
  and physicochemical propery schemas 
  ([resources/schemas](https://github.com/stefs304/mhclovac/tree/master/resources/schemas)).
  MHCLovac package includes `mhclovac-build` tool, used for compiling a 
  reference
  database and offers variety of options for choosing appropriate parameters.  
   Prebuilt databases for four species (human, chimpanzee, macaque, mouse) are 
   included in the package.

  Once the reference database is compiled, MHC binding prediction can be 
  preformed by scoring similarity of each potential ligand to precompiled 
  HLA-specific profile (for each physicochemical property included in the profile 
  database). Scoring is based on correlation coefficients, with the 
  availability of assigning weights to each physicochemical schema used. MHCLovac
   offers degree of flexibility with regard to scoring.

* Research and description of the tool will be available soon on 
[wiki](https://github.com/stefs304/mhclovac/wiki) pages.


<a name='installation'></a>
### Installation

Install from PyPI repository
```
pip install mhclovac
```

Download and install from git repository
```
git clone https://github.com/stefs304/mhclovac
cd mhclovac
pip install .
```

<a name='usage'></a>
### Usage
```
mhclovac --fasta example.fasta 
         --hla HLA-A*02:01 
         --peptide_length 9
         --reference_data human
         --output output.txt
```
To build reference profiles use `mhclovac-build` which is included in the 
package.
```
mhclovac-build --input_data binding_data.tsv
               --schemas h:hydrophobicity.csv i:isoelectric.csv
               --species human
               --output human.hdf5
```


