ACE-inhibitory Peptide Predictor

ACE-inhibitory peptide predictor (ACEiPP) is an efficient prediction tool to predict, design, and screen Angiotensin-Converting Enzyme (EC 3.4.15.1) inhibitory peptides (ACEiPs) by using optimized amino acid descriptors (AADs) and long-short term memory neural network. Here, we optimized a combined feature matrix VVSFZL37 consisting of six sets of AADs (VSW, VHSE, ST-scales, and Z-scales, FASGAI, and Lin's scales) and discussed the optimal amino acid profiles of ACEiPs. ACEiPP effectively learns the key sequence/structure characteristics of ACEiPs transformed by VVSFZL37 and significantly improves the prediction performance (ACC = 0.9479 and MCC = 0.9876) compared with the existing predictors.

ACEiPP mainly provides five services: (i) Predict unknown sequences in a single/batch manner; (ii) Screen predicted ACEiPs (pre-ACEiPs) from 10 theoretical peptide libraries, which are constructed by using different enzymes to simulate the hydrolysis of 21,249 proteins in 60 species; (iii) Screen potential multifunctional bioactive peptides (MBPs), e.g. related to cardiovascular diseases from the theoretical MBP library, which is constructed by predicting 30 types of food-derived BPs other than ACEiPs from DFBP (http://www.cqudfbp.net/); (iv) View the sequence characteristics of experimental ACEiPs and preACEiPs including length, mass, N-/C-terminals, and amino acid compositions; (v) Download datasets of ACEiPs, non-ACEiPs, preACEiPs, and preMBPs.