@@ -7,7 +7,7 @@ The [adaptive immune system](https://en.wikipedia.org/wiki/Adaptive_immune_syste
MHCflurry currently supports peptide / [MHC class I](https://en.wikipedia.org/wiki/MHC_class_I) affinity prediction using one model per MHC allele. The predictors may be trained on data that has been augmented with data imputed based on other alleles (see [Rubinsteyn 2016](http://biorxiv.org/content/early/2016/06/07/054775)). We anticipate adding additional models, including pan-allele and class II predictors.
You can fit MHCflurry models to your own data or download trained models that we provide. Our models are trained on data from [IEDB](http://www.iedb.org/home_v3.php) and [Kim 2014](http://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-15-241). See [here](downloads-generation/data_combined_iedb_kim2014) for details on the training data preparation. The steps we use to train predictors on this data, including hyperparameter selection using cross validation, are [here](downloads-generation/models_class1_allele_specific_single).
You can fit MHCflurry models to your own data or download trained models that we provide. Our models are trained on data from [IEDB](http://www.iedb.org/home_v3.php) and [Kim 2014](http://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-15-241). See [here](https://github.com/hammerlab/mhcflurry/tree/master/downloads-generation/data_combined_iedb_kim2014) for details on the training data preparation. The steps we use to train predictors on this data, including hyperparameter selection using cross validation, are [here](https://github.com/hammerlab/mhcflurry/tree/master/downloads-generation/models_class1_allele_specific_single).
The MHCflurry predictors are implemented in Python using [keras](https://keras.io).