%0 Journal Article
%T Network pharmacology modeling identifies synergistic Aurora B and ZAK interaction in triple-negative breast cancer
%A Abhishekh Gupta
%A Agnieszka Szwajda
%A Alok Jaiswal
%A Bhagwan Yadav
%A Denes Turei
%A Jani Saarela
%A Jing Tang
%A Julio Saez-Rodriguez
%A Krister Wennerberg
%A Liye He
%A Matti Kankainen
%A Prson Gautam
%A Sanna Timonen
%A Tero Aittokallio
%A Wenyu Wang
%A Yevhen Akimov
%J Archive of "NPJ Systems Biology and Applications".
%D 2019
%R 10.1038/s41540-019-0098-z
%X Network pharmacology modeling for MDA-MB-231 cancer cells. a Schematic outline of the computational¨Cexperimental approach to predicting and validating effective drug combinations and their underlying target interactions. The TIMMA algorithm takes as input single-drug sensitivity profiles and drug-target interaction profiles (here, among 41 kinase inhibitors and 385 kinase targets), and utilizes min¨Cmax averaging rules to search a target subset that is most predictive of the observed single-drug sensitivities in the given cells (see Methods). A drug combination is then treated as a combination of the selected targets, the combined effect of which can be quantitatively predicted based on the set relationships between the target profiles of the drugs. The outcome of the TIMMA model consists of a list of predicted drug synergy scores and a drug combination network for further experimental validation. b The drug combination network predicted for MDA-MB-231 cancer cells. The network consists of drugs (rectangular nodes) and their kinase targets (oval nodes). An effective drug combination can be inferred by checking whether the removal of them breaks the network into disjoint components (e.g., BI2536¨Cdasatinib combination and dasatinib¨Cmidostaurin combination). The EPHA5 and MAK target nodes contain multiple kinases that are unique to dasatinib and alvocidib, respectively, but indistinguishable by their target profiles. c The predicted drug combinations and their target interactions were confirmed using pairwise drug combination screen (left) and double knock-down siRNA screen (right) using cell viability assay (CellTiter-Glo). Drug combinations with predicted synergy score higher than the average (0.3485) were classified as high synergy group. d The double knock-downs that involved a predicted target of dasatinib showed a stronger cell viability inhibition compared to the other target pairs (right), which may explain the stronger synergies observed in the dasatinib-involving drug combinations compared to non-dasatinib combinations (left). Statistical significance was evaluated using Wilcoxon rank sum test (two-sided
%K Cancer
%K Computational biology and bioinformatics
%U https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6614366/