Authors

• Adam R. Rivers, US Department of Agriculture, Agricultural Research Service

• Qingpeng Zhang, US Department of Energy, Joint Genome Institute

• Susannah G. Tringe, US Department of Energy, Joint Genome Institute

Introduction

Vica is designed to identify highly divergent viruses and phage representing new families or orders in assembled metagenomic and metatranscriptomic data. Vica does this by combining information from across the spectrum of composition to homology. The current version of Vica uses three feature sets (5-mers, codon usage in all three frames, and minhash sketches from long kmers (k=24,31). The classifier uses a jointly trained deep neural network and logistic model implemented in Tensorflow. The software is designed to identify both DNA and RNA viruses and phage in metagenomes and metatranscriptomes.

Models

The current leases does not include trained models but we will be adding them in the future to allow for the rapid identification of viruses without model training.

Usage

This package can classify assembled data and train new classification models. Most users will only use the classification functionality in Vica. We will provide trained models for classifying contigs in future releases. classification can be easily invoked with the command:

vica classify -infile contigs.fasta -out classifications.txt -modeldir modeldir

The package also has a suite of tools to prepare data, train and evaluate new classification models. Many of the workflows for doing this can be evoked with the same sub-command interface:

vica split
vica get_features
vica train
vica evaluate

For details see the Tutorial.

Requirements

The package relies on a number of python dependencies that are resolved when the package is installed with PIP.

The non-python dependencies are:

• Bbtools > v37.75- https://jgi.doe.gov/data-and-tools/bbtools/

• Prodigal > v2.6.3 - https://github.com/hyattpd/Prodigal

• GNU Coreutils - http://www.gnu.org/software/coreutils/coreutils.html

Documentation

Documentation for the package is at http://vica.readthedocs.io/en/latest/

Package availability

• PyPi: https://pypi.python.org/pypi/vica

• Github: https://github.com/USDA-ARS-GBRU/vica

Copyright information

ViCA Copyright (c) 2018, The Regents of the University of California, through Lawrence Berkeley National Laboratory (subject to receipt of any required approvals from the U.S. Dept. of Energy). All rights reserved.

If you have questions about your rights to use or distribute this software, please contact Berkeley Lab’s Innovation & Partnerships Office at IPO@lbl.gov referring to “ Viral Classification Algorithm Using Supervised Learning (LBNL Ref 2017-125).”

NOTICE. This software was developed under funding from the U.S. Department of Energy. As such, the U.S. Government has been granted for itself and others acting on its behalf a paid-up, nonexclusive, irrevocable, worldwide license in the Software to reproduce, prepare derivative works, and perform publicly and display publicly. The U.S. Government is granted for itself and others acting on its behalf a paid-up, nonexclusive, irrevocable, worldwide license in the Software to reproduce, prepare derivative works, distribute copies to the public, perform publicly and display publicly, and to permit others to do so.