This document summarizes three papers on deep learning approaches for analyzing omics data: 1) A study that used denoising autoencoders to extract features from breast cancer gene expression data and found the features were linked to clinical characteristics and survival outcomes. 2) A study that used stacked denoising autoencoders to classify breast cancer samples and identify predictive genes related to diagnosis. 3) A study that integrated gene expression and miRNA data with autoencoders to identify survival subgroups in liver cancer patients, which were validated in additional cohorts and found to activate distinct pathways.