Our study uncovered many exciting biological findings—be sure to check it out to learn more!
It was a pleasure collaborating with such an amazing team, and I truly appreciate the support from Dr. Ayse Dincer, Dr. Joseph Janizek, Prof. Kamila Naxerova and Prof. Su-In Lee @suinlee.bsky.social!

4/4

By linking latent variables to mutation burden and survival, we discovered that mutation burden is closely associated with cell-cycle-related genes, and pathways for DNA-mismatch repair and MHC class II antigen presentation are consistently associated with patient survival.

3/4

We utilized explainable AI to identify the genes and pathways that define the latent spaces. We found that genes that are universally important across cancer types control immune cell activation, and cancer-type-specific genes and pathways define molecular disease subtypes.

2/4

We developed an unsupervised deep learning framework for the generation of low-dimensional latent spaces and applied it to gene expression data from 50,211 transcriptomes across 18 human cancers.

1/4