Dr.VAE: Drug Response Variational Autoencoder
This work addresses drug response prediction for biomedical applications, presenting an incremental improvement with novel model variants.
The authors tackled drug response prediction by developing two VAE-based models that learn latent gene state representations before and after drug application, resulting in performance improvements of 3-11% AUROC and 2-30% AUPR over benchmarks.
We present two deep generative models based on Variational Autoencoders to improve the accuracy of drug response prediction. Our models, Perturbation Variational Autoencoder and its semi-supervised extension, Drug Response Variational Autoencoder (Dr.VAE), learn latent representation of the underlying gene states before and after drug application that depend on: (i) drug-induced biological change of each gene and (ii) overall treatment response outcome. Our VAE-based models outperform the current published benchmarks in the field by anywhere from 3 to 11% AUROC and 2 to 30% AUPR. In addition, we found that better reconstruction accuracy does not necessarily lead to improvement in classification accuracy and that jointly trained models perform better than models that minimize reconstruction error independently.