vHULK, a New Tool for Bacteriophage Host Prediction Based on Annotated Genomic Features and Neural Networks

Area of Science:

• Virology
• Bioinformatics
• Computational Biology

Background:

• Experimental determination of bacteriophage hosts is time-consuming.
• There is a significant need for accurate computational methods for phage host prediction.

Purpose of the Study:

• To develop and evaluate vHULK, a novel computational tool for predicting bacteriophage hosts.
• To improve the accuracy and efficiency of identifying bacteriophage-host relationships.

Main Methods:

• vHULK was developed using 9504 phage genome features, including alignment significance scores of predicted proteins against a viral protein family database.
• Two neural network models were trained to predict bacteriophage hosts at the genus and species levels.

Main Results:

• vHULK achieved 83% precision and 79% recall at the genus level, and 71% precision and 67% recall at the species level on reduced redundancy test sets.
• Comparative analysis on a dataset of 2153 phage genomes showed vHULK outperformed three other host prediction tools at both genus and species levels.

Conclusions:

• vHULK demonstrates superior performance compared to existing tools for bacteriophage host prediction.
• The developed program represents a significant advancement in the state-of-the-art for computational phage host identification.