Abstract Cell shapes have long served as hallmarks for cancer diagnostics and therapeutics evaluations. This work converges public available image and transcriptome profiles to build a drug-gene-shape networks, thus facilitate the usage of cell shape features as drug targets during the development of novel cancer therapeutics. We took advantage of a previous work (Sailem et. al., 2017) of image-omits modeling where 504 shape-correlated genes were identified for 20 cell shape features across 18 breast cancer cell lines. Transcriptome profiles from Broad Institute Clue.io database were used to identify small molecular compounds that can significantly impact those shape-correlated genes. As a result, 22 compound families of at least two members were identified as candidates that can significantly impact the shape-correlating gene sets for 13 shape features. Our work re-opens the doors for using cell shape as drug target and considering the impact on cell shapes during drug repurposing and design of combination therapies for breast cancer. Compound families with the most significant correlations with gene sets regulating average cell areaCompound famileAverage scoremTOR inhibitors94.975Tubulin inhibitors94.615Adrenergic receptor antagonist94.545PKC inhibitor-93.580MEK inhibitor-94.137HDAC inhibitor-94.153 Citation Format: Zheng Yin, Chris Bakal, Stephen T. Wong. Predicting drug effects on breast cancer cell shapes by converging image and transcriptomic profiles [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6570.