The application discloses a tumor driving
gene super-sensitive quantitative detection method based on digital PCR and belongs to the technical field of tumor
gene detection in
molecular biology. The application realizes accurate calculation of the average number of molecules in a microdroplet, the
absolute concentration of
nucleic acid and the frequency of variant alleles by means of directional pretreatment of clinical samples, targeted
nucleic acid purification, multiple
chemical modification of tumor driving
gene primers and probes, microfluidic vibration atomization nanoscale microdroplet partitioning, six-color
fluorescence channel multiple digital PCR amplification and
machine learning assisted microdroplet intelligent
typing. The application breaks through the
bottleneck of existing digital PCR detection technology, avoids conventional processes and algorithms, realizes 0.001% VAF super-sensitive detection, has a quantitative relative deviation of less than or equal to 1%, simultaneously detects eight driving genes in a
single tube, is suitable for multiple clinical samples such as
plasma, FFPE tissue and
body fluid, and is excellent in detection sensitivity, specificity and
repeatability, and is suitable for early tumor screening,
efficacy monitoring, recurrence early warning and targeted
drug guidance.