NetMHCIIpan 4.0 server predicts binding of peptides to MHC class II molecules. The predictions are available for the three human MHC class II isotypes HLA-DR, HLA-DP and HLA-DQ, as well as mouse molecules (H-2).
Submission is accepted in two formats - as a list of peptides or as a protein sequence in FASTA format. A comprehensive list of MHC molecules is available for prediction, alternatively the user can upload their MHC protein sequence of interest.
The prediction values are given likelihood for MHC antigen presentation and %Rank score. The percentile rank for a peptide is generated by comparing its score against the scores of 100,000 random natural peptides. For example, if a peptide is assigned a rank of 1%, it means that its predicted affinity is among the top 1% scores for the specified molecule.
If selected by the user, also predicted binding affinity and corresponding %Rank values can be reported.
Strong and weak binding peptides are identified based on %Rank, with customizable thresholds. You may sort the output based on predicted binding affinity and filter out non-binders. -->
The NetMHCIIpan-4.0 server predicts peptide binding to any MHC II molecule of known sequence using Artificial Neural Networks (ANNs). It is trained on an extensive dataset of over 500.000 measurements of Binding Affinity (BA) and Eluted Ligand mass spectrometry (EL), covering the three human MHC class II isotypes HLA-DR, HLA-DQ, HLA-DP, as well as the mouse molecules (H-2). The introduction of EL data extends the number of MHC II molecules covered, since BA data covers 59 molecules and EL data covers 74. As mentioned, the network can predict for any MHC II of known sequence, which the user can specify as FASTA format. The network can predict for peptides of any length.
The output of the model is a prediction score for the likelihood of a peptide to be naturally presented by and MHC II receptor of choice. The output also includes %rank score, which normalizes prediction score by comparing to prediction of a set of random peptides. Optionally, the model also outputs BA prediction and %rank scores.
New in this version: The two output neuron architechture introduced in NetMHCpan-4.0 permits the inclusion of EL data, and the new training algorithm NNAlign_MA extends training data to ligands of ambiguous allele assignments. The model also, optionally, encodes ligand context.
Refer to the instruction page for more details.
The project is a collaboration between CBS, and LIAI.
View the version history of this server. All previous versions are available online, for comparison and reference.