A Multi-center Study on the Reproducibility of Drug-Response Assays in Mammalian Cell Lines

• DOI: 10.1016/j.cels.2019.06.005
• Publication Date: 2019-07-10T14:33:51Z
• Authors: Mario Niepel; Marc Hafner; Caitlin E. Mills; Kartik Subramanian; Elizabeth H. Will...; Mirra Chung; Benjamin Gaudio; Anne Marie Barrette; Alan D. Stern; Bin Hu; James E. Korkola; Joe W. Gray; Marc R. Birtwistle; Laura M. Heiser; Peter K. Sorger; Caroline E. Shamu; Gomathi Jayaraman; Evren U. Azeloglu; Ravi Iyengar; Eric A. Sobie; Gordon B. Mills; Tiera Liby; Jacob D. Jaffe; Maria Alimova; Desiree Davison; Xiaodong Lu; Todd R. Golub; Aravind Subramanian; Brandon Shelley; Clive N. Svendsen; Avi Ma’ayan; Mario Medvedovic; Heidi S. Feiler; Rebecca Smith; Kaylyn Devlin
• Keywords: Drug response

Highlights•Implementing FAIR data standards requires identification of experimental confounders•Five labs performed the same experiment on mammalian cells and compared results•Several factors affecting reproducibility were explored•Biological context had an unexpected impact robustness cell-based assaysSummaryEvidence that some high-impact biomedical results cannot be repeated has stimulated interest in practices generate findable, accessible, interoperable, reusable (FAIR) data. Multiple papers have identified specific examples irreproducibility, but practical ways to make more reproducible not been widely studied. Here, five research centers NIH LINCS Program Consortium investigate a prototypical perturbational assay: quantifying responsiveness cultured anti-cancer drugs. Such assays are important for drug development, studying cellular networks, patient stratification. While many computational intra- inter-center reproducibility, most difficult identify control those with strong dependency biological context. These often vary magnitude being analyzed growth conditions. We provide such context-sensitive factors, thereby improving both theory practice assays.Graphical abstract