Overview of the proposed methodology for predicting AHTPs that involved the following steps: (i) construction of benchmarking and independent dataset; (ii) extraction of 9 different feature encodings that characterize those peptides in different ways and generation of 51 feature descriptors; (iii) generation of 51-dimensional feature vector using feature representation learning scheme; (iv) ranking the 51-dimension feature vector using RF algorithm; (v) generation of the optimal meta-predictor model using sequential forward search; (vi) construction of the final prediction model by integrating four meta-predictors that separates the input into putative AHTPs and non-AHTPs.
Reference
mAHTPred: a sequence-based meta predictor for improving the prediction of anti-hypertensive peptides using effective feature representation (Bioinformatics ). [Please cite this paper if you find mATHPred useful in your research]
