A novel DNA weighted summation system for cancer diagnosis

Precise and non-invasive cancer diagnosis remains a persistent pursuit for clinicians. DNA circuit, owing to its biocompatibility and capability for complex information processing, has been explored for cancer diagnostics. Here, we established a cancer diagnosis platform based on a novel DNA weighted summation system. Using mRNA expression profiles from The Cancer Genome Atlas, we developed diagnostic models for prostate and colorectal cancers. An effective amplification method was then optimized and employed to quasi-linearly amplify and convert trace mRNA in blood, followed by molecular computation via our DNA weighted summation system. This approach overcomes a key limitation of previous DNA computing platforms that were restricted to integer weights, enabling continuous-weight computation and achieving precise disease diagnosis within a streamlined workflow. In clinical validation with 100 blood samples per cancer type, the platform achieved robust diagnostic accuracies of 85.0% for CRC and 86.0% for PCa. As a generalizable molecular computing framework, this platform may be extended to diverse clinical applications, including cancer staging, prognosis assessment, and therapeutic response prediction.