A novel diagnostic signature constructed based on serum-circulating mA-related miRNAs for cancer detection.

Emerging data have revealed that mA modification and its regulators are key participants in tumorigenesis and progression. The cell-free miRNAs have been demonstrated to circulate stably in the serum, making them biomarker candidates for cancer diagnosis. Here we included 16,902 serum samples to develop a diagnostic signature named m1A-miRNA signature based on serum circulating mA-related miRNAs for cancer detection. The m1A-miRNA signature presented excellent accuracy, and the area under the curve (AUC) was 0.991 in the training cohort. The diagnostic capability of the m1A-miRNA signature was not affected by sexual distinction, age and non-cancer disease. As far as distinguishing cancer types is concerned, the signature exerted superior ability in identifying the types of glioblastoma multiforme, gastric cancer and lung cancer. We also found that the m1A-miRNA signature showed a satisfactory AUC in the early diagnosis of pan-cancer. Additionally, the accuracy of the m1A-miRNA signature was further verified by clinical samples.