A Deep Ensemble Predictor for Identifying Anti-Hypertensive Peptides Using Pretrained Protein Embedding.

Area of Science:

• Biochemistry and Bioinformatics
• Cardiovascular Research
• Computational Biology

Background:

• Hypertension (HT) is a major cause of cardiovascular diseases and associated with severe health complications like renal failure.
• Current methods for identifying anti-hypertension peptides (AHTPs) are often time-consuming and laborious.
• Computational approaches offer a promising alternative for efficient AHTP classification.

Purpose of the Study:

• To develop a novel computational model for predicting anti-hypertension peptides (AHTPs).
• To enhance the efficiency and accuracy of AHTP identification compared to existing methods.
• To provide a valuable tool for advancing hypertension therapy.

Main Methods:

• Developed a comprehensive feature representation algorithm using a pretrained model and convolutional neural network.
• Constructed a deep ensemble model for predicting AHTPs.
• Validated the predictor on benchmark and independent datasets.

Main Results:

• The developed predictor demonstrated superior performance on an independent test set compared to recent methods.
• The model effectively identified anti-hypertension peptides, showcasing its practical utility.
• Comparative analysis confirmed the model's advancement in AHTP classification.

Conclusions:

• The deep ensemble model offers a powerful and efficient computational tool for AHTP identification.
• This approach can significantly contribute to the development of novel hypertension therapies.
• The study provides valuable insights into the classification of AHTPs, aiding future research.