This demo shows how to predict the binding strength of peptides to different Major Histocompatibility Class II (MHC-II) haplotypes.
First load the library:
This demo assumes NetMHCIIpan is installed. Installation cannot be done netmhc2pan
, as one needs to request a download link at https://www.cbs.dtu.dk/services/NetMHCIIpan/.
To install netmhc2pan
, use install_netmhc2pan
with the download link:
install_netmhc2pan("https://www.cbs.dtu.dk/download/33A6B0AC-0F2E-11E9-B4D1-8ABCB9CD16B5/")
The installation of netmhc2pan
is checked, with the goal of producing a helpful error message:
tryCatch(
check_netmhc2pan_installation(),
error = function(e) print(e)
)
#> <simpleError in check_netmhc2pan_installation(): NetMHCIIpan binary not found at
#> /home/richel/.local/share/netMHCIIpan-3.2/netMHCIIpan
#>
#> Tip 1: from R, run 'netmhc2pan::install_netmhc2pan()'
#> with a (non-expired) download URL
#> Tip 2: request a download URL at the NetMHCIIpan download page at
#>
#> https://services.healthtech.dtu.dk/service.php?NetMHCIIpan-3.2
#> >
Now, lets use the sequence of an example protein:
Now, we need to select an MHC-II haplotype. There are many alleles, so first we count the number of haplotypes:
if (is_netmhc2pan_installed()) {
mhc_haplotypes <- get_netmhc2pan_alleles()
length(mhc_haplotypes)
}
Now, we simply pick the first haplotype:
Now, we can predict how strong our peptide binds to our allele, by obtaining the Inhibitory Concentration 50% (IC50) value in nanomolars (nM), of which lower values indicate stronger binders:
if (is_netmhc2pan_installed()) {
knitr::kable(
predict_ic50(
peptides = sequence,
mhc_haplotype = mhc_haplotype
)
)
}
To investigate if whole a protein is immunogenic, we need to obtain the IC50 values for all its cleaved products. As the MHC-II molecules prefers 13 amino acids residues, we can get the IC50 values for each residue as such: