Abstract IA012: Leveraging multiomics and machine learning towards a stepwise approach to multi-cancer screening

Abstract

Abstract Early cancer discovery is challenging due to heterogeneity of different cancers and even within cancers from the same organ. In order to tackle this problem, Freenome has built a multiomics discovery platform that looks for signals along the entire central dogma-DNA, methylation, RNA, protein, immunoprofiling, extracellular vesicles, circulating cells, among others. In order to featurize, train, and build the best models that are generalizable and robust, we have created our own machine learning platform incorporating different computational biological signals. All this is aligned with a step-wise investigation of cancers that can most benefit the population and make a difference to patients, starting with colorectal cancer and advanced adenoma. During this talk, we will discuss the clinical, scientific, and computational strategy that we think is important to create the best products to benefit the most patients. Citation Format: Jimmy Lin. Leveraging multiomics and machine learning towards a stepwise approach to multi-cancer screening [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Translating Cancer Evolution and Data Science: The Next Frontier; 2023 Dec 3-6; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(3 Suppl_2):Abstract nr IA012.