smmMatrix: Get Prediction Matrix
In EpitopePrediction: Predict Binding Between Peptides and MHC Molecules

Description Usage Arguments Value Examples

View source: R/peptide-mhc.R

Loads the prediction matrix for a given MHC molecule at the given peptide length. Matrices are loaded from the package folder upon first load and are then cached in memory for further uses. The SMM method predicts binding by summing the matrix entries for each amino acid per position and then adding and MHC-dependant number. The resulting number is the log IC50 value (IC50=half maximal inhibitory concentration). Hence, low numbers mean strong binding and high numbers mean weak binding.

1	smmMatrix(mhc = "HLA-A-02:01", l = 9)

`mhc`	name of the MHC molecule.
`l`	the peptide length.

A list with components M (the matrix) and c (a number to be added to each prediction made with this matrix).

## load prediction matrix for HLA-A-02:01 9-mers
M <- smmMatrix( "HLA-A-02:01", 9 )$M
## How well does Leucine bind at each position?
M["L",]
## Which amino acid is most preferred at the C-terminal position?
names( which.min( M[,9] ) )

## Naming of the MHC molecule is permissive
for( mhc in c("HLA-A0201","HLA-A-0201","HLA-A*0201","HLA-A*02:01") ){
   names( which.min( smmMatrix( "HLA-A-02:01", 9 )$M[,9] ) )
}