Prediction of Compound Cytotoxicity Based on Compound Structures and Cell Line Molecular Characteristics

Abstract

In parallel to developments in Next-Generation Sequencing for cancer patient therapy decision making, personalized approaches to chemotherapy selection are also becoming desired. In an ideal situation, an individual's genomic, transcriptomic, and tumor-specific in-vitro response to chemical perturbation would be combined, and the US National Cancer Institute NCI-60 project has systematically screened a large chemical library against a variety of cell lines from various tumor types. Therefore, chemoinformatics approaches to make effective use of this data and identify the chemical and biological factors are of value. In this work, we investigate the impact of both chemical and biological descriptions of tumor response to chemical inhibition, and assess how well modeling approaches can predict tumor inhibition response on external datasets. We find that external datasets in both the classification and regression problems are reasonably well addressed, with the impact of chemical description outweighing the contribution from transcriptome or genome descriptions of tumors.

Comparison of Cellular Morphological Descriptors and Molecular Fingerprints for the Prediction of Cytotoxicity- and Proliferation-Related Assays

“…Previous models in cytotoxicity predictions have used many machine learning approaches such as Random Forests, 20,21 Bayesian learning, 22 and deep learning; 23 these have been trained on features such as physicochemical descriptors and molecular fingerprints as well as cell line descriptors of mRNA expression data. 24 Previously, assays developed based on morphology screens have been shown to identify similar sets of compounds compared to a standard cytotoxicity assay. 25 In recent years, models have started to also use other highdimensional readouts for toxicity prediction, such as the combination between molecular fingerprints and gene expression data.…”

“…Such deviations could arise because the in vitro dose may be completely irrelevant to target organ exposure in vivo (among other possible factors). Previous models in cytotoxicity predictions have used many machine learning approaches such as Random Forests, , Bayesian learning, and deep learning; these have been trained on features such as physicochemical descriptors and molecular fingerprints as well as cell line descriptors of mRNA expression data . Previously, assays developed based on morphology screens have been shown to identify similar sets of compounds compared to a standard cytotoxicity assay .…”

A systematic evaluation of deep learning methods for the prediction of drug synergy in cancer

“…This makes it a powerful approach to develop models that are able to predict drug synergy based on drug combination screening experiments and other relevant data. Several ML models for drug synergy prediction have been described in the literature [11,[17][18][19][20][21]. Many of these studies used tree-based ML methods, such as random forests (RFs) [17,18,20] or gradient boosting [18,19,21].…”

A systematic evaluation of deep learning methods for the prediction of drug synergy in cancer

“…Several ML models for drug synergy prediction have been described in the literature [8,[11][12][13][14][15]. Many of these studies used tree-based ML methods, such as random forests (RFs) [11,12,14] or gradient boosting [12,13,15].…”

“…This makes it a powerful 2/22 approach to develop models that are able to predict drug synergy based on drug combination screening experiments and other relevant data. Several ML models for drug synergy prediction have been described in the literature [8,[11][12][13][14][15]. Many of these studies used tree-based ML methods, such as random forests (RFs) [11,12,14] or gradient boosting [12,13,15].…”