papanikos / vhulk

viral Host UnveiLing Kit (vHULK)

vHULK

Viral Host UnveiLing Kit - A tool-kit for phage host prediction

Pre-print publication

A preview of vHULK's publication is available in BioRxiv as the original manuscript is under review.

Phage Host Prediction using high level features and neural networks

Metagenomics and sequencing techiniques have greatly improved in these last five years and, as a consequence, the amount of data from microbial communities is astronomic. An import part of the microbial community are phages, which have their own ecological roles in the environment. Besides that, they have also been given a possible human relevant (clinical) role as terminators of multidrug resistant bacterial infections. A lot of basic research still need to be done in the Phage therapy field, and part of this research involves gathering knowledge from new phages present in the environment as well as about their relationship with clinical relevant bacterial pathogens.

Having this scenario in mind, we have developed v.HULK. An user-friendly tool for prediction of phage hosts given their complete or partial genome in FASTA format. Our tool outputs an ensamble prediction at the genus or species level based on scores of four different neural network models. Each model was trained with more than 4,000 genomes whose phage-host relatonship was known. v.HULK also outputs a mesure of entropy for each final prediction, which we have demonstrated to be correlated with prediction's accuracy. The user might understand this value as additional information of how certain v.HULK is about a particular prediction. We also suspect that phage with higher entropy values may have a broad host-range. But that hypothesis is to be tested later. Accuracy results in test datasets were >99% for predictions at the genus level and >98% at the species level. vHULK currently supports predictions for 52 different prokaryotic host species and 61 different genera.

Target species

Acinetobacter_baumannii, Aeromonas_salmonicida, Arthrobacter_globiformis, Bacillus_cereus, Bacillus_megaterium, Bacillus_pumilus, Bacillus_subtilis, Bacillus_thuringiensis, Brucella_abortus, Campylobacter_jejuni, Caulobacter_crescentus, Cellulophaga_baltica, Citrobacter_freundii, Clostridium_difficile, Clostridium_perfringens, Cronobacter_sakazakii, Enterococcus_faecalis, Erwinia_amylovora, Escherichia_coli, Gordonia_terrae, Helicobacter_pylori, Klebsiella_pneumoniae, Lactococcus_lactis, Listeria_monocytogenes, Microbacterium_foliorum, Moraxella_catarrhalis, Mycobacterium_smegmatis, Paenibacillus_larvae, Pectobacterium_atrosepticum, Propionibacterium_acnes, Proteus_mirabilis, Pseudomonas_aeruginosa, Pseudomonas_fluorescens, Pseudomonas_syringae, Ralstonia_solanacearum, Rhodococcus_erythropolis, Salmonella_enterica, Salmonella_typhimurium, Shigella_flexneri, Shigella_sonnei, Staphylococcus_aureus, Staphylococcus_epidermidis, Streptococcus_pneumoniae, Streptococcus_thermophilus, Streptomyces_griseus, Streptomyces_lividans, Streptomyces_venezuelae, Sulfolobus_islandicus, Vibrio_anguillarum, Vibrio_cholerae, Vibrio_parahaemolyticus, Yersinia_enterocolitica

Target genera

Achromobacter, Acidianus, Acinetobacter, Aeromonas, Arthrobacter, Bacillus, Brevibacillus, Brucella, Burkholderia, Campylobacter, Caulobacter, Cellulophaga, Citrobacter, Clostridium, Corynebacterium, Cronobacter, Dickeya, Enterobacter, Enterococcus, Erwinia, Escherichia, Flavobacterium, Gordonia, Haloarcula, Halorubrum, Helicobacter, Klebsiella, Lactobacillus, Lactococcus, Leuconostoc, Listeria, Mannheimia, Microbacterium, Moraxella, Mycobacterium, Paenibacillus, Pectobacterium, Prochlorococcus, Propionibacterium, Proteus, Pseudoalteromonas, Pseudomonas, Ralstonia, Rhizobium, Rhodococcus, Ruegeria, Salmonella, Serratia, Shigella, Sinorhizobium, Staphylococcus, Stenotrophomonas, Streptococcus, Streptomyces, Sulfolobus, Synechococcus, Thermus, Vibrio, Xanthomonas, Xylella, Yersinia

Scripts

Main script:

• vHULK_v0.1.py - Deep learning host prediction from phage genomes

Auxiliary script:

• download_and_set_models.py - Used only at the first use for downloading and setting models

Dependencies

All scripts from this project were coded in Python 3.6. So, first of all, make sure you have it installed and updated.
v.HULK's main scrip (vHULK.py) requires Prokka and HMMER tools as dependencies.

• Prokka - Rapid Prokaryotic genome annotation.
• HMMER Tools - Biosequence analysis using profile hidden Markov models

These Python libraries are required:

• Numpy, Pandas, Scipy - Efficiently handling arrays and scientific computing for Python
• Biopython - Handling biological sequences
• Tensorflow - Google's Deep Neural Networks libraries

To install these dependencies, there are usually two ways: pip or conda.
We strongly recomend the creation of a specific conda environment containing the installed libraries and tools. In this way, everything will be set with the following commands:

conda create -n vHULK -c biobuilds perl=5.22 python=3.6
conda install -n vHULK -c bioconda prokka hmmer
conda install -n vHULK -c bioconda numpy pandas scipy biopython tensorflow=2.2.0
conda activate vHULK

Obs: Some people have been facing issues with newer versions of Prokka. To avoid such problems, we froze the specific versions above which we know for sure they work with vHULK.

Installing

Getting vHULK ready to run is as simple as cloning this Github project or download and extract it to a directory inside your computer:

git clone https://github.com/LaboratorioBioinformatica/vHULK

Quick start

Inside the directory where vHULK was extracted (or cloned), you will need to download and set the models. This is required only once and it is simple. Just run:

python download_and_set_models.py

All set!
Now, to run vHULK type:

python vHULK_v0.1.py -i input_directory -t num_threads

Change 'input_directory' to the folder where bins or genomes are stored in fasta format and 'num_threads' to the number of CPU cores to be used. Several threads should be used to speed up prokka and hmm searches.
Results will be stored in the 'Results' folder inside the input directory.
Obs: You need to execute the scripts from the directory where vHULK was extracted, i.e., v.HULK's root folder.

Running the example datasets

We provide a folder with example datasets containing mocking bins of RefSeq viral and bacterial genomes.
To try these examples, run:

python vHULK_v0.1.py -i test_input/ -t 4

Obs: It will take about 2 minutes to generate your nice and accurate predictions.

Input

vHULK is ready to accept whole or partial phage genomes. Just keep in mind that vHULK's predictions are based in high level annotated features, i.e., features that depend on gene annotation. So, very small contigs with few or no entire genes will not provide features for vHULK to work with. In general, if you have a partial genome larger than 10 kbp you should be fine.
Only FASTA nucleotide files are accepted, and vHULK will understand each individual file as one phage genome. Therefore, single-sequence FASTA files will be understood as whole or partial phage genomes and multiFASTA as a metagenomic BIN. Our tests indicate that prediction's accuracy is not affected in fragmented genomes (BINs).

Output

vHULK main output file is "results.csv", which will be inside input_folder/results after execution is finished. If more than one fasta file is inside the input_folder, vHULK will understand that there are more than one bin or genome for prediction. Therefore, each line of the CSV file will correspond to a bin or genome. Identifiers will by in the first column.
Output example:

BIN/genome,pred_genus_relu,score_genus_relu,pred_genus_softmax,score_genus_softmax,pred_species_relu,score_species_relu,pred_species_softmax,score_species_softmax,final_prediction,entropy
MK801680_BIN_staphylococcus_aureus,Staphylococcus,1.0,Staphylococcus,0.9999988,Staphylococcus_aureus,1.0,Staphylococcus_aureus,0.9999976,Staphylococcus_aureus,1.9657731e-05
ZC4_test,Mycobacterium,0.31396204,Streptomyces,0.9495423,None,0,Streptomyces_griseus,0.4496213,Streptomyces,0.31331635
MK570225_enterococcus_faecalis,Enterococcus,1.0,Enterococcus,0.99999917,None,0,Enterococcus_faecalis,0.9999571,Enterococcus_faecalis,1.4299788e-05
MK524521_mycobacterium_smegmatis,Mycobacterium,1.0,Mycobacterium,1.0,Mycobacterium_smegmatis,1.0,Mycobacterium_smegmatis,1.0,Mycobacterium_smegmatis,1.465061e-11
MN41915_vibrio_cholerae,None,0,Vibrio,1.0,None,0,Vibrio_cholerae,0.99995685,Vibrio_cholerae,1.9252913e-11
MN689520_lactococcus_lactis,Lactococcus,1.0,Lactococcus,1.0,Lactococcus_lactis,1.0,Lactococcus_lactis,0.9999994,Lactococcus_lactis,4.9965405e-07

Note that there is a header in the output file, and it is self explanatory. Nonetheless:

• 1st column: ID of the submitted sequence
• 2nd to 9th columns: Host prediction and respective score for each one of the four vHULK's internal models
• 10th column: Unified host prediction generated by an heuristic decision tree based on score values
• 11th column: Entropy value for predictions of the Softmax model at the genus level. It can be used as a proxy of confidence for a particular prediction, i.e., values close to 0 correlate with higher accuracies and values close to 4 otherwise.