Abstract 3106: Integrated in vitro and machine learning approaches for targeted drug screening in gliomas with concurrent NF1 and ATRX loss

Abstract

Abstract Neurofibromatosis type 1 (NF1) frequently precipitates the development of brain and nerve tumors, with subsets of associated solid tumors demonstrating an aggressive phenotype often characterized by alpha thalassaemia mental retardation X-linked (ATRX) loss. In response, we executed an in vitro high-throughput drug screening utilizing a 10,000-compound library in an NF1-associated ATRX mutant glioblastoma cell line (JHH-NF1-GBM1) and in U251 cells with ATRX knockout, mimicking the concurrent loss of ATRX and NF1. The screening protocol aimed to pinpoint compounds selectively targeting vulnerabilities arising from the co-occurrence of ATRX and NF1 loss. To enhance the precision of our results, we employed an unsupervised-supervised machine learning approach. Unsupervised models identified distinctive chemical groups associated with heightened sensitivity in ATRX/NF1-deficient cells. Subsequently, supervised models were trained, and their uncertainty estimates were utilized to further narrow down the selection of hit compounds. We also utilized pre-trained models to predict ADME and toxicity properties, facilitating a further filtration. This multi-step in vitro screening, coupled with in silico analyses, successfully identified promising compounds, subsequently validated through biophysical characterization. The integration of machine learning in our screening pipeline enhances the identification of potential targeted therapies for aggressive tumors (gliomas) characterized by ATRX/NF1 loss, offering a comprehensive and efficient approach to drug discovery in the pursuit of more effective treatments. Citation Format: Swati Dubey, Somesh Mohapatra, Ming Yuan, Charles G. Eberhart, Fausto J. Rodriguez. Integrated in vitro and machine learning approaches for targeted drug screening in gliomas with concurrent NF1 and ATRX loss [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3106.