README.md

In Degoot123/PeptideMHC2Prediction: Prediction of Peptide:MHC-II Molecules Interactions

PeptideMHC2Prediction

An R package implementing biophysical methods for prediction of peptide and MHC-II molecules interactions, developed by Degoot and colleagues (https://www.frontiersin.org/articles/10.3389/fimmu.2018.01410/full). It can predict the probability of binding affinity between peptide of any length that greater than or equal 9 and 35 HLA-II molecules belonging to the three MHC-II loci of human genes: DRB, DP, and DQ.

License

This package is licensed under a CC BY-NC-SA 4.0 license. It is free to use for non-commercial purposes.

Installation

# Install the package from github
# devtools::install_github("Degoot/PeptideMHC2Prediction")

Example

Load the package

 ## library(PeptideMHC2Prediction)

To predict probability of binding affinity between peptide, for example say,

  pep <- "ADAGYAPATPAAAGA"

and an MHC-II molecule, for examples say DRB1_0401, based on trans-allelic method, you can use the following:

##predictBinding(pep, "DRB1_0401", "DRB", 2)

Or using the intra-allelic method as follows:

##predictBinding(pep, "DRB1_0401", "DRB", 1)

the package also allows to make predictions for set of peptides and MHC-II molecules, together organized in form of a dataframe of the three columns. The columns of the dataframe must be in the following order: th peptide, the name of MHC-II moleclue, and the locus. For example, let us assume that we have a dataframe called dataset, which has the following structure.

| Peptide | MHC | Locus | |:---------------:|:----------:|:-----:| | AAAGAEAGKATTEEQ | DRB1_0301 | DRB | | AFALVLLFCALASSC | DRB3_0101 | DRB | | AAAGAEAGKATTEEQ | DRB1_1501 | DRB | | SGLVWGQKYFKGNFQ | DPB10401 | DP | | SSNPTILSEGNSFTA | DQB10201 | DQ |

Then you call the function predictBindingSet to get prediction of the above dataset as follows:

##predictBindingSet(dataset, opt = 1)

The parameter opt specifies the prediction method, 1 for intra-allelic method and 2 for trans-allelic method. The output of function predictBindingSet is a dataframe of four columns, which has the below structure:

| Peptide | MHC | Probability | Binary Value | |:---------------:|:----------:|:-----------:|:------------:| | AAAGAEAGKATTEEQ | DRB1_0301 | 0.176 | 0 | | AFALVLLFCALASSC | DRB3_0101 | 0.549 | 1 | | AAAGAEAGKATTEEQ | DRB1_1501 | 0.679 | 1 | | SGLVWGQKYFKGNFQ | DPB10401 | 0.300 | 0 | | SSNPTILSEGNSFTA | DQB10201 | 0.026 | 0 |

References

Abdoelnaser M. Degoot, Chirove Faraimunashe, and Ndifon Wilfred (2018) Trans-Allelic Model for Prediction of Peptide:MHC-II Interactions, Frontiers in Immunology. URL:https://www.frontiersin.org/article/10.3389/fimmu.2018.01410.

Degoot123/PeptideMHC2Prediction documentation built on May 21, 2019, 10:15 a.m.