VHIP is a machine-learning model that predict virus-microbe interactions (i.e., infection or non-infection) from genomic sequences of viruses and microbes of interest. It leverages virus-microbe coevolution signal that are extracted from genomic sequences. VHIP was trained on lab-verified virus-microbe pairs collected from literature and the NCBI virus database.
You can find more information about the philosophy and performance of VHIP here: https://www.biorxiv.org/content/10.1101/2023.11.03.565433v1
This module will be made available as a conda environment. To be updated once it is up on conda.
Also will be available with pip install virushostinteractionpredictor.
The inputs for VHIP are sequences for the viruses and hosts of interest. Each file should represent an unique virus/host species. In addition, a blastn between the virus and host sequences, and between viruses and host spacers are needed. If there are no results from the blastn, then the files can be empty.
An example is included in the example folder. It includes virus and microbes sequences, how they should be organized (separate folders), and example blastn files. Make sure you are able to run VHIP on the example before applying to a new dataset.