Mining Drug-Target Associations in Cancer: Analysis of Gene Expression and Drug Activity Correlations

Abstract

Cancer is a complex disease affecting millions of people worldwide, with over a hundred clinically approved drugs available. In order to improve therapy, treatment, and response, it is essential to draw better maps of the targets of cancer drugs and possible side interactors. This study presents a large-scale screening method to find associations of cancer drugs with human genes. The analysis is focused on the current collection of Food and Drug Administration (FDA)-approved drugs (which includes about one hundred chemicals). The approach integrates global gene-expression transcriptomic profiles with drug-activity profiles of a set of 60 human cell lines obtained for a collection of chemical compounds (small bioactive molecules). Using a standardized expression for each gene versus standardized activity for each drug, Pearson and Spearman correlations were calculated for all possible pairwise gene-drug combinations. These correlations were used to build a global bipartite network that includes 1007 gene-drug significant associations. The data are integrated into an open web-tool called GEDA (Gene Expression and Drug Activity) which includes a relational view of cancer drugs and genes, disclosing the putative indirect interactions found for FDA-approved drugs as well as the known targets of these drugs. The results also provide insight into the complex action of pharmaceuticals, presenting an alternative view to address predicted pleiotropic effects of the drugs.

Resistance to Tyrosine Kinase Inhibitors in Chronic Myeloid Leukemia—From Molecular Mechanisms to Clinical Relevance

“…Following this line of research, artificial intelligence (AI) and machine learning (ML) methods have been proven to be also very useful for integrating large-scale- omics data from cancer patients and for analyzing gene expression profiles in response to different drugs [ 231 ]. In this scenario, positive associations between gene expression and anticancer drug activity allowed the discovery of gene targets for the drugs tested [ 232 ]. Conversely, a negative association between gene expression and drug activity measured in these assays (for example, detecting high expression of a gene corresponding to decreased activity of a drug) indicated that such a gene/protein could be mediating resistance and low sensitivity to the drug.…”

Cancer drug resistance induced by EMT: novel therapeutic strategies

“…Genome and proteome-wide information associated to the drugs activity in human cells is essential to generate better maps of the molecular targets of each drug (De Las Rivas et al 2018 ). Construction of this type of drug-target interaction mapping has been successful in the field of cancer genomics thanks to the possibility of testing the activity of hundreds of cancer drugs in multiple human cancer cell lines (Arroyo et al 2020 ). Similar studies using genomic data combined with drugs activity are needed to elucidate at molecular level the genetic and somatic basis for inter-individual differences in drug response.…”

Design, Synthesis, and In Vitro/In Vivo Anti-Cancer Activities of Novel (20S)-10,11-Methylenedioxy-Camptothecin Heterocyclic Derivatives

“…Every year, more than 10 million new cancer cases and nearly seven million deaths are reported worldwide [ 1 ]. Chemotherapy is one of the most widely used modalities for cancer treatment, with over one hundred clinically approved drugs available [ 2 ]. Among them, camptothecin (CPT) [ 1 ] (1, Figure 1 ), a quinoline alkaloid as an inhibitor of topoisomerase I, is a well-known drug isolated from the Chinese tree Camptotheca by Wall et al in 1966 [ 3 ].…”