A Method for the Recognition of High Resolution Melting Curve Based on Machine Learning

摘要

High Resolution Melting (HRM)analysis technology has been widely used in the study of genotyping and has caused widespread attention. However, the recognition of melting curves requires manual interactions, resulting in a decrease in the accuracy and efficiency of genotyping. This paper presents a new method for the recognition of high melting curve based on machine learning. We validated this method on 57 known serotypes of Streptococcus pneumoniae data sets and achieved 100% accuracy on the test set with good robustness. Then, we used the Kras gene fragment and its 7 mutant to further verify the effectiveness of the algorithm and demonstrated 100% prediction accuracy. The method proposed in this paper achieves efficient, accurate, and automated genotyping, and has broad application prospects in the fields of medicine and biology.