Abstract The RET oncogene is a receptor tyrosine kinase, mainly expressed in neural crest-derived tissues, that plays a role in cell growth and differentiation. Since the first identification of inherited mutations in RET as the cause of Multiple endocrine neoplasia type 2 (MEN2), somatic RET alterations have been identified as actionable drivers in subtypes of lung and thyroid, among other cancer types. This enabled the development of several RET targeted treatments, such as pralsetinib and selpercatinib, showing meaningful clinical benefit. To understand the full scope of RET alteration frequency and heterogeneity we used the AACR-GENIE database. We found that the frequency of RET alterations pan-cancer is 3%, of which two-thirds are missense mutations. The major cancer types with RET alterations are medullary thyroid, papillary thyroid and NSCLC, as expected, but also breast, colorectal and melanoma. An important distinction between NSCLC and thyroid cancer compared to colorectal, breast and melanoma is the fraction of missense vs. fusion events, with 30%-40% of samples in the former group containing RET fusions. This is in contrast with >80% of samples in the latter group containing RET missense mutations. Moreover, the heterogeneity of RET alterations is large, with over 1,000 unique types of missense mutations identified. The vast majority of mutations are classified as variants of uncertain significance (VUS). To better understand the potential activity of these mutations, we assayed over 200 different missense mutations and fusions using a high-throughput cell-based functional assay. Results were analyzed using a novel machine learning model consisting of a multi-scale deep convolutional neural net, followed by a tree-based regressor. Our results show that the assay correctly predicted the activity of both known activating alterations as well as of loss of function mutations. Interestingly, only ~3% of over 150 VUS tested were found to be activating. Furthermore, we show that the active alterations are sensitive to the FDA-approved targeted agents pralsetinib and selpercatinib, with varying IC50 values. Taken together, these results uncover the full extent of actionable RET alterations, differentiating between active and non-active variants, as well as the effectiveness of RET inhibitors. Citation Format: Shay Rotkopf, DIkla Haham, Natalie Fillipov-Levy, Lea Birnbaum, Elinor Dehan, Ilona Kifer, Zohar Barbash, Gabi Tarcic. A systematic analysis of RET mutations and their sensitivity to different RET inhibitors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1162.