Journal: Digital Discovery
Article Title: Identification of drug candidates against glioblastoma with machine learning and high-throughput screening of heterogeneous cellular models
doi: 10.1039/d5dd00190k
Figure Lengend Snippet: High-throughput phenotypic assay screening of heterogeneous, patient-derived GBM cell lines. (A) High content phenotypic assay screening workflow on GBM cells. Five libraries were screened at multiple concentrations (C3L, 786 compounds at [0.003, 0.03, 0.3, 3] µm; KCGS, 187 compounds at [0.1, 1] µm; LOPAC, 1277 compounds at [0.5, 3] µm; Prestwick, 1276 compounds at [1, 10] µm; TargetMol-330, 330 compounds at [0.01, 0.1, 1, 10] µm). (B) Representative images of multiplex cell painting assay screen performed across the three main GBM cellular subtypes: classical (GCGR-E13, GCGR-E28), proneural (GCGR-E31, GCGR-E34), and mesenchymal (GCGR-E21, GCGR-E57). The cell painting stains shown are Hoechst (blue), phalloidin/WGA (green), and endoplasmic reticulum (red). (C) Density plots over all six GBM cell lines of the percentage of nuclei counts after compound treatment with respect to the average DMSO negative control (100%) per plate. After removal of duplicates, the dataset contains 3561 unique compounds. Colours correspond to each library (panel A), and only the highest concentration per library is shown. Cell line sensitivity to the compounds is variable across the cell lines; the bimodality displayed by the TargetMol library reveals a larger fraction of compounds that affect cell viability as compared to the other four libraries. (D) Density plots of the TargetMol library screen at all concentrations of cell line GCGR-E57 (top) and of cell line GCGR-E21 (bottom). The black dotted line represents the average DMSO negative control nuclei count. As the concentration increases, a larger number of compounds affects cell viability.
Article Snippet: The density of compounds' cell survival against DMSO control at the maximum concentration per library (C3L – 3 μM, KCGS – 1 μM, LOPAC – 3 μM, Prestwick – 10 μM, and TargetMol – 10 μM) shows a bimodal distribution for the TargetMol library, with more treatments that affect the cells' viability.
Techniques: Sample Throughput, Phenotypic Assay, Screening, Derivative Assay, Multiplex Assay, Cell Painting, Negative Control, Concentration Assay, Sensitivity, Fluorescence Microscopy