Colorectal cancer (CRC) is one of the deadliest cancers. At diagnosis, nearly 60% of CRC cases are advanced with regional or distant metastases, and only about 40% of patients with advanced disease live beyond 5 years. Colonoscopy remains the gold standard for early-stage diagnosis of invasive CRC lesions and high-risk adenomas. Unfortunately, although timely colonoscopy screening can reduce CRC mortality, because the procedure is medically complex and costly, participation rates are very low. To remove barriers for access and improve overall participation, specific and sensitive tests that are simpler and less expensive are needed. Emerging liquid biopsy approaches for the detection of circulating tumor DNA (ctDNA) fragments in a sample of blood could address this need. Dan Landau and his team at Weill Cornell Medicine and the New York Genome Center have recently developed an ultra-sensitive technique for ctDNA detection in blood called MRDetect, a machine learning-powered approach for early cancer detection that relies on the cumulative signal provided by thousands of tumor-specific mutations. Members of the Landau laboratory will collaborate with scientists in the laboratory of Claus Andersen at Aarhus University in Denmark to analyze samples from high-risk premalignant adenoma patients using MRDetect. Through this combined effort, they hope to a significantly advance liquid biopsy ctDNA detection technology such that it can accurately determine the extent of plasma burden of ctDNA in these early-stage patients. This study will demonstrate the feasibility of liquid biopsy as a noninvasive screening/early detection for CRC. In addition it will allow the team to refine their algorithm for analyzing ctDNA in diagnostic screens that can detect cancer in cases in which the patient’s tumor risk is unknown and, if cancer is present, the overall tumor burden is very low.