Cancer diagnosis using cell-free DNA (cfDNA) can significantly improve treatment and survival but has several technical limitations. Here, we show that tumor-associated mutations create neomers, DNA sequences 11-18bp in length that are absent in the human genome, that can accurately detect cancer subtypes and features. We show that we can detect twenty-one different tumor-types with higher accuracy than state-of-the-art methods using a neomer-based classifier. Refinement of this classifier via supervised learning identified additional cancer features with even greater precision. We also demonstrate that neomers can precisely diagnose cancer from cfDNA in liquid biopsy samples. Finally, we show that neomers can be used to detect cancer-associated non-coding mutations affecting gene regulatory activity. Combined, our results identify a novel, sensitive, specific and straightforward cancer diagnostic tool.

Introduction: Cancer diagnosis using cell-free DNA (cfDNA) has the potential to improve treatment and survival but has several technical limitations. Methods: Here, we used neomers, short DNA sequences (13-17bp in length) that are largely absent from the healthy human genome, but appear in the tumor genome due to somatic mutations, to detect cancer at early stages from cfDNA. cfDNA was extracted from 1mL of plasma, and utilized from Whole Genome Sequencing at 5X coverage. Results: First, we analyzed over 2,500 cancer whole-genome sequences (WGS), to show that a neomer-based classifier can distinguish twenty-one different tumor types with higher accuracy than state-of-the-art methods. Refinement of this classifier via manually selected neomer sets, identified additional cancer features with greater precision. We next generated cfDNA WGS data from >600 patients and non-cancerous controls across four different types of cancers (prostate, lung, ovarian, colorectal), and demonstrated that neomers can precisely identify cancer from a limited amount of plasma for all four cancer types, with exampled sensitivities of 93.73% and 73.47% for lung and ovarian cancers, respectively. Additionally, using both luciferase and massively parallel reporter assays (MPRAs), we show that our neomer-driven approach can detect cancer-associated non-coding mutations that affect gene regulatory activity. Conclusions: Our results show that neomers provide a novel, sensitive, specific and potentially broadly applicable cancer detection tool. This approach has several advantages over existing strategies for liquid biopsies: 1) it is based on standard next-generation sequencing which is already widely available in the clinic; 2) it requires low amounts of starting material and is completely tumor agnostic; 3) it searches the entire genome and is not restricted to known driver mutations. 4) it allows the detection of early-stage disease. Citation Format: Ofer Yizhar Barnea, Ilias Georgakopoulos-Soares, Nadav Ahituv, Jocelyn Chapman, Martin Hemberg, Ioannis Mouratidis, Konstantinos Syrigos, Nikolaos Syrigos, Ioannis Vathiotis, Mayank Mahajan, Emmanouil Panagiotou, Andriani Charpidou, Mark Kvale, Candace S. Chan, Ryder Easterlin. Leveraging sequences missing from the human genome to detect cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 991.