PHERI

PHERI predicts bacterial host genera for bacteriophages from whole-genome sequence data to enable identification of phage host specificity in metagenomic and environmental samples.

Key Features:

• Host Prediction: Predicts the bacterial host genus for bacteriophages by analyzing whole-genome sequence data in FASTA format.
• Machine Learning Approach: Employs machine learning algorithms to infer host specificity, reporting approximately 97% precision for eight genera: Arthrobacter, Escherichia, Gordonia, Lactococcus, Mycobacterium, Pseudomonas, Staphylococcus, and Streptococcus.
• Protein Sequence Identification: Identifies and highlights protein sequences that are important for host selection to provide insights into host–phage interactions.
• Data Preprocessing: Operates on preprocessed sequence data stored in a designated folder to support downstream analysis.

Scientific Applications:

• Medical Therapeutics: Supports identification of phages targeting multi‑drug‑resistant bacterial strains to inform phage therapy development and selection.
• Environmental and Industrial Applications: Aids study of phage–bacteria dynamics in environments relevant to food safety, bioremediation, and industrial microbiology.

Methodology:

Analyzes whole-genome sequences in FASTA format using machine learning algorithms for genus-level host prediction, identifies protein sequences implicated in host selection, and processes inputs from a preprocessed data folder.