Algorithms for Drug Sensitivity Prediction

Abstract

Precision medicine entails the design of therapies that are matched for each individual patient. Thus, predictive modeling of drug responses for specific patients constitutes a significant challenge for personalized therapy. In this article, we consider a review of approaches that have been proposed to tackle the drug sensitivity prediction problem especially with respect to personalized cancer therapy. We first discuss modeling approaches that are based on genomic characterizations alone and further the discussion by including modeling techniques that integrate both genomic and functional information. A comparative analysis of the prediction performance of four representative algorithms, elastic net, random forest, kernelized Bayesian multi-task learning and deep learning, reflecting the broad classes of regularized linear, ensemble, kernelized and neural network-based models, respectively, has been included in the paper. The review also considers the challenges that need to be addressed for successful implementation of the algorithms in clinical practice.

Toward Explainable Anticancer Compound Sensitivity Prediction via Multimodal Attention-Based Convolutional Encoders

“…This varied clinical response has led to the promise of personalized (or precision) medicine in cancer, where molecular biomarkers, e.g., the expression of specific genes, obtained from a patient's tumor profiling may be used to choose a personalized therapy. These challenges highlight a need across both pharmaceutical and healthcare industries for multimodal quantitative methods that can jointly exploit disparate sources of knowledge with the goal of characterizing the link between the molecular structure of compounds, the genetic and epigenetic alterations of the biological samples and drug response (De Niz et al, 2016). In this work, we present a multimodal attention-based convolutional encoder that enables us to tackle the aforementioned challenges.…”

Methodological challenges in translational drug response modeling in cancer: A systematic analysis with FORESEE

“…Within the context of a permanently growing interest in precision medicine over the last years, where therapies are intended to be tailored to specific characteristics of individual patients, the study of drug sensitivity prediction for complex diseases, such as cancer, has experienced a tremendous boost [1]. The availability of both the computational power to work with complex algorithms and large-scale pharmacological data sets gave rise to various drug sensitivity studies.…”

“…While these efforts are great in comparing and improving distinct components of a model, such as batch effect correction methods [14], feature selection methods [15] or regression algorithms [1], they often lack the consideration of the complete modeling workflow and the interplay of the individual pipeline components. Moreover, published methods are inclined to be biased towards the authors' fields of expertise, which hinders a fair and objective benchmarking of existing methods.…”

Methodological Challenges in Translational Drug Response Modeling in Cancer

“…Within the context of a permanently growing interest in precision medicine over the last years, where therapies are intended to be tailored to specific characteristics of individual patients, the study of drug sensitivity prediction for complex diseases, such as cancer, has experienced a tremendous boost (De Niz et al, 2016). The availability of both the computational power to work with complex algorithms and large-scale pharmacological datasets gave rise to various drug sensitivity studies.…”

“…In order to keep an overview of the versatile field of drug sensitivity prediction, efforts were made to systematically compare existing approaches, for example in collaborative projects, such as the DREAM challenges by Costello et al (2014). Albeit these efforts are great in comparing and improving distinct components of a model, such as batch effect correction methods (Lazar et al, 2012), feature selection methods (Saeys et al, 2007) or regression algorithms (De Niz et al, 2016), they often lack the consideration of the complete modeling workflow and the interplay of the individual pipeline components. Moreover, published methods are inclined to be biased towards the authors' fields of expertise, which hinders a fair and objective benchmarking of existing methods.…”