Predictive Modeling of Angiotensin I-Converting Enzyme Inhibitory Peptides Using Various Machine Learning Approaches

Predicting Antidiabetic Peptide Activity: A Machine Learning Perspective on Type 1 and Type 2 Diabetes

Diabetes mellitus (DM) presents a critical global health challenge, characterized by persistent hyperglycemia and associated with substantial economic and health-related burdens. This study employs advanced machine-learning techniques to improve the prediction and classification of antidiabetic peptides, with a particular focus on differentiating those effective against T1DM from those targeting T2DM. We integrate feature selection with analysis methods, including logistic regression, support vector machines (SVM), and adaptive boosting (AdaBoost), to classify antidiabetic peptides based on key features. Feature selection through the Lasso-penalized method identifies critical peptide characteristics that significantly influence antidiabetic activity, thereby establishing a robust foundation for future peptide design. A comprehensive evaluation of logistic regression, SVM, and AdaBoost shows that AdaBoost consistently outperforms the other methods, making it the most effective approach for classifying antidiabetic peptides. This research underscores the potential of machine learning in the systematic evaluation of bioactive peptides, contributing to the advancement of peptide-based therapies for diabetes management.

Machine learning methods for unveiling the potential of antioxidant short peptides in goat milk-derived proteins during in vitro gastrointestinal digestion

Milk serves as an important dietary source of bioactive peptides, offering notable benefits to individuals. Among the antioxidant short peptides (di- and tripeptides) generated from gastrointestinal digestion are characterized by enhanced bioavailability and bioaccessibility, while assessing them individually presents a labor-intensive and expensive challenge. Based on 4 distinct types of amino acid descriptors (physicochemical, 3D structural, quantum, and topological attributes) and genetic algorithms for feature selection, 1 and 4 machine learning predicted models separately for di- and tripeptides with ABTS radical scavenging capacity exhibited excellent fitting and prediction ability with random forest regression as machine learning algorithm. Intriguingly, the electronic properties of N-terminal amino acid were considered as only factor affecting the antioxidant capacity of dipeptides containing both tyrosine and tryptophan. Four peptides from the potential di- and tripeptides exhibited highly predicted values by the constructed predicted models. Subsequently, a total of 45 dipeptides and 52 tripeptides were screened by a customized workflow in goat milk during in vitro simulated digestion. In addition to 5 known antioxidant dipeptides, 9 peptides were quantified during digestion, falling within the range of 0.04 to 1.78 mg L-1. Particularly noteworthy was the promising in vivo functionality of antioxidant dipeptides with N-terminal tyrosine, supported by in silico assays. Overall, this investigation explored crucial molecular properties influencing antioxidant short peptides and high-throughput screening potential peptides with antioxidant activity from goat milk aided by machine learning, thereby facilitating the identification of novel bioactive peptides from milk-derived proteins and paving the way for understanding their metabolites during digestion.

pLM4ACE: A protein language model based predictor for antihypertensive peptide screening

Angiotensin-I converting enzyme (ACE) regulates the renin-angiotensin system and is a drug target in clinical treatment for hypertension. This study aims to develop a protein language model (pLM) with evolutionary scale modeling (ESM-2) embeddings that is trained on experimental data to screen peptides with strong ACE inhibitory activity. Twelve conventional peptide embedding approaches and five machine learning (ML) modeling methods were also tested for performance comparison. Among the 65 classifiers tested, logistic regression with ESM-2 embeddings showed the best performance, with balanced accuracy (BACC), Matthews correlation coefficient (MCC), and area under the curve of 0.883 ± 0.017, 0.77 ± 0.032, and 0.96 ± 0.009, respectively. Multilayer perceptron and support vector machine also exhibited great compatibility with ESM-2 embeddings. The ESM-2 embeddings showed superior performance in enhancing the prediction model compared to the 12 traditional embedding methods. A user-friendly webserver (https://sqzujiduce.us-east-1.awsapprunner.com) with the top three models is now freely available.

Bioaccessibility of novel antihypertensive short-chain peptides in goat milk using the INFOGEST static digestion model by effect-directed assays

Short-chain peptides (SCPs, 2-4 amino acids) offer potential health benefits. A customized workflow was designed to screen SCPs in goat milk during INFOGEST digestion in vitro and 186 SCPs were preliminarily identified. Based on a two-terminal position numbering method and genetic algorithm combined with a support vector machine, 22 SCPs with predicted IC50 values less than 10 μM were obtained using a quantitative structure-activity relationship (QSAR) model with satisfactory fitting and predictive capacity (R2, RMSE, Q2, and R2pre of 0.93, 0.27, 0.71, and 0.65, respectively). Four novel antihypertensive SCPs were confirmed by testing in vitro and molecular docking analysis, and their quantification results (0.06 to 1.53 mg L-1) suggested distinct metabolic fates. This study facilitated the discovery of unknown potential food-derived antihypertensive peptides and the understanding of bioaccessible peptides during digestion.

PeptideBERT: A Language Model Based on Transformers for Peptide Property Prediction

Recent advances in language models have enabled the protein modeling community with a powerful tool that uses transformers to represent protein sequences as text. This breakthrough enables a sequence-to-property prediction for peptides without relying on explicit structural data. Inspired by the recent progress in the field of large language models, we present PeptideBERT, a protein language model specifically tailored for predicting essential peptide properties such as hemolysis, solubility, and nonfouling. The PeptideBERT utilizes the ProtBERT pretrained transformer model with 12 attention heads and 12 hidden layers. Through fine-tuning the pretrained model for the three downstream tasks, our model is state of the art (SOTA) in predicting hemolysis, which is crucial for determining a peptide's potential to induce red blood cells as well as nonfouling properties. Leveraging primarily shorter sequences and a data set with negative samples predominantly associated with insoluble peptides, our model showcases remarkable performance.

Application of Molecular Simulation Methods in Food Science: Status and Prospects

Molecular simulation methods, such as molecular docking, molecular dynamic (MD) simulation, and quantum chemical (QC) calculation, have become popular as characterization and/or virtual screening tools because they can visually display interaction details that in vitro experiments can not capture and quickly screen bioactive compounds from large databases with millions of molecules. Currently, interdisciplinary research has expanded molecular simulation technology from computer aided drug design (CADD) to food science. More food scientists are supporting their hypotheses/results with this technology. To understand better the use of molecular simulation methods, it is necessary to systematically summarize the latest applications and usage trends of molecular simulation methods in the research field of food science. However, this type of review article is rare. To bridge this gap, we have comprehensively summarized the principle, combination usage, and application of molecular simulation methods in food science. We also analyzed the limitations and future trends and offered valuable strategies with the latest technologies to help food scientists use molecular simulation methods.

Novel angiotensin I-converting enzyme (ACE) inhibitory peptides from walnut protein isolate: Separation, identification and molecular docking study

Walnut protein isolate was hydrolyzed using alcalase® to obtain angiotensin-I-converting enzyme (ACE) inhibitory (ACEI) peptides. The components with high ACEI activity were successfully purified from walnut protein isolate hydrolysates (WPIH) by ultrafiltration and G-25 gel chromatography. The 1520 peptides were identified by high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS). Then the screening model of ACEI peptides was established by in silico approach. It was found that four ACEI active peptides (PPKP, YPQY, YLPP, and PKPP) were obtained with IC₅₀ values ranging from 506 to 89 μmol/L, among which PPKP had the highest ACEI activity (IC₅₀  = 89 ± 1 μmol/L). The four peptides mentioned above were novel, non-toxic, and resistant to gastrointestinal digestion. The molecular docking studies showed that the ACEI effect of ACEI peptide was mainly due to the interaction with residues of Gln281 and His353 in the ACE active pockets. In vivo availability of ACEI peptides showed that the probability of PPKP binding to ACE was 37.9% in the human body. Our studies suggest that the ACEI peptides derived from the WPIH can be considered functional foods that can prevent hypertension. PRACTICAL APPLICATIONS: Hypertension is a significant risk factor for cardiovascular and cerebrovascular disease, the leading cause of death worldwide. This study used a cost-effective method to isolate and identify potential ACEI peptides from the walnut meal. Since the walnut meal is often discarded in the processing of walnut products and thus pollutes the environment, the preparation of walnut meal into ACEI peptides can reduce the impact of hypertension on people and reduce environmental pollution. The experimental results show that walnut ACEI peptides are a safe and healthy nutritional product.

Food protein-derived bioactive peptides for the management of nutrition related chronic diseases

Dietary intervention via modifications of dietary pattern or supplementations of naturally derived bioactive compounds has been considered as an efficient approach in management of nutrition related chronic diseases. Food protein-derived bioactive peptide is representative of natural compounds which show the potential to prevent or mitigate nutrition related chronic diseases. In the past decades, substantial research has been conducted concentrating on the characterization, bioavailability, and activity assessment of bioactive peptides. Although various activities of bioactive peptides have been reported, the activity testes of most peptides were only conducted in cells and animal models. Some clinical trials of bioactive peptides were also reported but only limited to antihypertensive peptides, antidiabetic peptides and peptides modulating blood lipid profile. Hereby, clinical evidence of bioactive peptides in management of nutrition-related chronic diseases is summarized in this chapter, which aims at providing implications for the clinical studies of bioactive peptides in the future.

Advance in dietary polyphenols as dipeptidyl peptidase-IV inhibitors to alleviate type 2 diabetes mellitus: aspects from structure-activity relationship and characterization methods

Dietary polyphenols with great antidiabetic effects are the most abundant components in edible products. Dietary polyphenols have attracted attention as dipeptidyl peptidase-IV (DPP-IV) inhibitors and indirectly improve insulin secretion. The DPP-IV inhibitory activities of dietary polyphenols depend on their structural diversity. Screening methods that can be used to rapidly and accurately identify potential polyphenol DPP-IV inhibitors are urgently needed. This review focuses on the relationship between the structures of dietary polyphenols and their DPP-IV inhibitory effects. Different characterization methods used for polyphenols as DPP-IV inhibitors have been summarized and compared. We conclude that the position and number of hydroxyl groups, methoxy groups, glycosylated groups, and the extent of conjugation influence the efficiency of inhibition of DPP-IV. Various combinations of methods, such as in-vitro enzymatic inhibition, ex-vivo/in-vivo enzymatic inhibition, cell-based in situ, and in-silico virtual screening, are used to evaluate the DPP-IV inhibitory effects of dietary polyphenols. Further investigations of polyphenol DPP-IV inhibitors will improve the bioaccessibility and bioavailability of these bioactive compounds. Exploration of (i) dietary polyphenols derived from multiple targets, that can prevent diabetes, and (ii) actual binding interactions via multispectral analysis, to understand the binding interactions in the complexes, is required.

Study on the domain selective inhibition of angiotensin-converting enzyme (ACE) by food-derived tyrosine-containing dipeptides

In this article, the selective inhibition of several tyrosine-containing dipeptides on N and C domain of ACE (angiotensin-converting enzyme) was studied, and the interaction mode of ACE and inhibitors was simulated by molecular docking. MTT assay was used to detect the effect of dipeptide on human umbilical vein endothelial cells (HUVEC). The results showed that the food-derived dipeptides AY (Ala-Tyr), LY (Leu-Tyr), and IY (Ile-Tyr) containing tyrosine at the C-terminal were favorable structures for selective inhibition of ACE C-domain. These dipeptides showed competitive and mixed inhibition patterns, while the dipeptides EY (Glu-Tyr), RY (Arg-Tyr), FY (Phe-Tyr), and SY (Ser-Tyr) showed noncompetitive inhibition. Food-derived dipeptides containing tyrosine have no cytotoxicity on HUVEC cells, which provides a basis for the application of food-derived tyrosine dipeptides as antihypertensive peptides. This study provides a theoretical basis for exploring the selective inhibition mechanism of ACE inhibitory peptides containing tyrosine residue. PRACTICAL APPLICATIONS: Angiotensin-converting enzyme (ACE) is a two-domain dipeptidyl carboxypeptidase, which is a key enzyme to regulate blood pressure. ACE has two active sites, C- and N-domain, which have high catalytic activity. Although the amino acid sequences of the two active sites have 60% similarity, there are some differences in structure and function. The action mechanism of ACE domain should be clarified, and the structure-activity relationship between inhibitors and ACE domain has not been systematically studied. The aim of this study was to identify the selective inhibitory effect of food-derived tyrosine dipeptides on the domain of ACE. This provides a new idea for finding new antihypertensive drugs with less side effects.