Marco Loog

Percolate: an exponential family JIVE model to design DNA-based predictors of drug response

Percolate combines gene expression with other genomic data to create robust DNA-based predictors of drug response, outperforming individual data types.

Identifying commonalities between cell lines and tumors at the single cell level using Sobolev Alignment of deep generative models

This paper introduces Sobolev Alignment, a method to identify shared biological processes between cancer cell lines and tumors using deep generative models and kernel methods, showing its effectiveness on synthetic and real datasets.

Predicting clinical drug response from model systems by non-linear subspace-based transfer learning

The paper develops TRANSACT, a transfer learning framework for predicting clinical drug response, outperforming existing methods on multiple datasets.

PRECISE: A domain adaptation approach to transfer predictors of drug response from pre-clinical models to tumors

PRECISE transfers drug response predictors from pre-clinical models to human tumors using domain adaptation, improving biomarker-drug associations in human tumors.

Synthetic-evolution reveals that phosphoregulation of the mitotic kinesin-5 Cin8 is constrained

This study shows that phosphoregulation of the mitotic kinesin-5 Cin8 requires site-specific and precise phosphorylation sites to achieve optimal regulation.