Comparative-study-of-PSSM-SMM-and-ANN-for-22125

22125 Algorithms in bioinformatics

• Paul Jeremy Simon (s202592)
• Maria Vinh Thuy Tien Ta (s134042)
• Huijiao Yang (s202360)

summary

We explored 3 methods' performance for MHC I binding prediction. In this process, we used the nest cross validation to train and evaluate SMM and ANN. And then we used 3 metrics, RMSE, AUC and Accuracy, to evaluate which method is the best one. Finally, we made a robustness test to check if the 3 methids are sensitive to distribution of target and metrics.