INeoP usage instructions

dbPepNeo 2.0

Database of Neoantigen Peptides from Mass Spectrometry and TCR Recognition

Home

BLASTdb

DeepCNN-Ineo

Document

Download

Usage instructions

Our tool can be used to predict the immunogenicity of neoantigen.

1. For antigen, 11297 experimental IEDB epitopes were used for training, involving nine main HLA supertypes. We recommend using peptide of length 9 AAs .

2. Predictive score value greater than 0.8 is suggested as Positive-High (immunogenicity), the score between 0.5-0.8 is suggested as Positive-Low, the score less than 0.5 is suggested as Negative-High. Examples can be clicked to test it.

DeepCNN-Ineo was only a preliminary test on a small scope of fixed datasets, for tumor neoantigens identified from upstream approaches. We've collected as many neoantigen peptides as we can. However, experimentally verified immunogenicity neoantigens are relatively precious and therefore scarce. To ameliorate the effects and increase the reliability of the model, we newly added binding affinity values to DeepCNN-Ineo, considering both the binding between HLA-peptide pairs and the potential immunogenicity of pHLA. Double filtering can increase the reliability of DeepCNN-Ineo prediction of neoantigen immunogenicity. Users are free to choose this feature.