apclusterv 1.1.2

No project description provided

Apclusterv: Clustering viral genomes with Affinity Propagation

Apclusterv is a novel clustering software for viral genomes. The input genomes can be either complete genomes or contigs from metagenomic assembly. The program is based on protein-protein alignment and written in python

Dependencies:

python>=3.7
pandas
numpy
networkx >= 2.8.4
scipy >=1.8.1
scikit-learn >= 1.1.2
MCL
diamond >= 0.9.34
prodigal >= 2.6.3
R>=3.6.1

Installation:

Suppose you are in a conda environment, you need to install MCL, prodigal (for ORF prediction),diamond (for alignment) and R(if not already installed, we just need stats library in r-base)

conda install diamond -c bioconda 
conda install mcl -c bioconda
conda install prodigal -c bioconda
conda install r-base 

pip install apclusterv 

Getting Started:

option 1. start with contigs

step1. preduct ORFs from the DNA file with the following command:

prepare  contig_dna_fasta 

(contig_dna_fasta is the path to the dna sequences for clustering)

step2. execute clustering with the following command:

apclusterv -contig contig_dna_fasta 

option 2. if you already have protein sequences from the contigs, you can run apclusterv by the proteins and a protein-contig map file.

An example of protein file and mapfile are data/experiment1.faa and data/experiment1.csv

apclusterv -protein experiment1.faa -csv experiment1.csv

Help message and parameter setting

apclusterv -h

Results

The program will create tmp/ directory. The clustering result is tmp/cluster_result.i.r.csv (cluster_result.3.4.csv by default) Simulation profile used in the manuscript is in data/profile.csv RI and ARI for evalation script is data/eval.py