Differentiating amino acids from nanopore sequencing

Amino acids nanopore sequencing is a significant breakthrough in the fields of molecular biology, biochemistry, and medical diagnostics.The tool's high sensitivity, specificity, and real-time analytic capability make it essential for accurately identifying amino acids.A new nanopore, known as Msp-NTA-Ni, has recently advanced the bounds by allowing accurate differentiation of all 20 proteinogenic amino acids and their post-translational modifications (PTMs).Utilizing the data produced by this nanopore, our research conducted a thorough examination of five features, pinpointing the most useful pairs for the purpose of classification.Subsequently, we undertake an elaborate process that encompasses the training, fine-tuning, and comparative evaluation of multiple machine learning models, such as Random Forest, CatBoost, and SVM.The results of our research indicate that the Random Forest model surpasses the current benchmarks, obtaining a validation accuracy of 99.04%.Moreover, our research emphasizes the crucial significance of particular combinations of features, such as the mean and standard deviation, in improving the performance of the model, despite some limitations in differentiating between certain pairs of amino acids.