A Prediction Model for Uncoating Receptor Usage in Human Enteroviruses Based on Amino Acid Sequences and a Naive...

Area of Science:

• Bioinformatics
• Virology
• Structural Biology

Background:

• Human enteroviruses utilize specific host cell receptors for entry and replication.
• Identifying these uncoating receptors is crucial for understanding viral pathogenesis and developing targeted therapies.
• Previous methods for receptor identification were limited by data availability and complexity.

Purpose of the Study:

• To develop a predictive bioinformatics algorithm for identifying human enterovirus uncoating receptors.
• To leverage amino acid sequences and physicochemical properties for accurate receptor prediction.
• To classify enterovirus serotypes and predict their receptor usage.

Main Methods:

• Construction of a bioinformatics prediction algorithm using Naive Bayes and network analysis.
• Utilizing amino acid sequences of receptor-binding sites and their physicochemical properties as model features.
• Classification of human enterovirus serotypes into training, validation, and prediction datasets based on available receptor and 3D structural data.

Main Results:

• The prediction model achieved 100% accuracy on both training and validation datasets.
• Analysis of 56 enterovirus serotypes revealed that most use known receptors (e.g., SCARB2, CAR, ICAM-1).
• A subset of enteroviruses may share novel, currently unidentified uncoating receptors.

Conclusions:

• Amino acid sequences and physicochemical properties accurately predict human enterovirus uncoating receptors.
• The algorithm reflects three-dimensional structural features of receptor-binding sites.
• This predictive approach offers valuable insights for enterovirus research, vaccine design, and antiviral drug development.