Visualizing drug effects over time in live animals using optical pharmacodynamics

Prediction of drug combination responses is a research question of growing importance for cancer and other complex diseases. Current machine learning approaches generally consider predicting either drug combination synergy summaries or single combination dose-response values, which fail to appropriately model the continuous nature of the underlying dose-response combination surface and can lead to inconsistencies when a synergy score or a dose-response matrix is reconstructed from separate predictions. We propose a novel prediction method, comboKR, that directly predicts the continuous drug combination response surface for a drug combination. The method is based on a powerful input–output kernel regression technique and functional modelling of the response surface. ComboKR belongs to the family of functional output regression methods, where the prediction target is a function, in our case, a non-linear parametric surface. Our method thus avoids predicting discretized forms of the target as scalars, vectors or matrices, and therefore provides better interpolation and extrapolation along the surfaces. As an important part of our approach, we develop a novel normalisation between response surfaces that standardises the heterogeneous experimental designs used to measure the dose-responses, and thus allows training the method with data measured in different laboratories. Our experiments on two predictive scenarios and using two combination datasets highlight the suitability of the proposed approach especially in the traditionally challenging setting of predicting combination responses for new drugs not available in the training data.