PHERI - Phage Host Exploration pipeline

Abstract

Antibiotic resistance is becoming a common problem in health care, veterinary medicine, agriculture or food industry. Multi-resistant bacterial strains occur in all regions of the world. One of the possible future solutions is the use of bacteriophages in therapy.Bacteriophages are the most abundant form of life in the biosphere, so it is highly likely that we can purify a specific phage against each target bacterium. A standard identification and consistent characterization of individual bacteriophages include host-specificity of viruses.Unfortunately, these routine methods are also considerably time consuming. With the advent of new modern sequencing methods, scientists are able to obtain multiple phage sequences from samples and identify more phages. However there appeared a problem with unknown host specificity of identified phages. The solution to this problem may be to use a bioinformatic approach in the form of prediction software capable to determine a bacterial host based on the phage whole-genome sequence. The result of our research is the machine learning algorithm based tool called PHERI. PHERI predicts suitable bacterial host genus for purification of individual viruses from different samples. In addition, the tool can identify and highlight protein sequences that are important for host selection.

Phylogenomic analyses and host range prediction of cluster P mycobacteriophages

“… Putative host genera of the 40 cluster P bacteriophages included in this study ( Supplementary Table 1 ) as predicted by PHERI ( Baláž et al 2020 ). …”

“…Following the best practices suggested by Versoza and Pfeifer (2022) , both exploratory and confirmatory methods were used to computationally predict host ranges for 40 closely related cluster P mycobacteriophages ( Supplementary Table 1 ). First, the exploratory tool PHERI v.0.2 ( Baláž et al 2020 ) was used to predict bacterial host genera. Among the currently available exploratory host range prediction tools, PHERI was the most user-friendly and well-documented, making it ideally suited for course-based undergraduate research experiences.…”

The Rationale for Using Bacteriophage to Treat and Prevent Periprosthetic Joint Infections

“…The use of phages could also enable to combat current unculturable bacteria, on the condition that they can easily be identified through genomic approach. It has been suggested that machine learning approaches can be utilized to either identify, or generate through synthetic genomics, based on the genomic information provided on the bacterial target ( Leite et al, 2018 ; Martorell-Marugán et al, 2019 ; Baláž et al, 2020 ; Pirnay, 2020 ).…”

Computational Prediction of Bacteriophage Host Ranges

“…BacteriophageHostPrediction [41] uses more than 200 features-ranging from genomic sequences (such as nucleotide and codon frequencies and GC-content), to protein sequences (such as amino acid frequency), to protein secondary structure (such as α-helix and β-sheet frequencies), and to physicochemical properties (such as molecular weight and isoelectric point)-to represent receptor-binding proteins which play a crucial role in determining host specificity by recognizing receptors on the surface of the bacterial host [47]. At a higher level of sequence representation, PHERI [48] infers bacterial hosts from bacteriophage sequences through annotated protein sequence clusters.…”