David López-Paz

The compositional perturbation autoencoder separates drug, dose and cell-type effects in single-cell transcriptomes, predicts unseen combinations, and generates 5,329 unmeasured genetic combinations in silico.

This paper introduces Poincare maps, a hyperbolic geometry-based method for analyzing complex hierarchies in single-cell data, achieving state-of-the-art results in visualization, clustering, and lineage detection.

The paper introduces Back-to-Back regression to disentangle correlated factors from multivariate observations, showing it outperforms existing methods in causal identification.