CompleConta

Estimation of completeness/contamination with simple taxonomic classification of bacterial genomes (bins)

Introduction:

Comple teness and Conta mination estimation using EggNOG-profiles

This tool was developed for PhenDB, where it provides a quick and rough estimation of both completeness, contamination and heterogeneity (similar to CheckM) and inference of taxonomic information that can be gathered from those marker genes (like in Amphora). The analysis includes a rudimental lowest common ancestor (LCA) approach with 90% majority rule. Currently the databases are build from bactNOG data of the EggNOG 4.5 release and taxonomy level 2 of EggNOG 5 (Bacteria). Please note, that the taxonomic tree only contains bacteria, which were used to create the EggNOG database. To run CompleConta you need to perform a classification with the bactNOG database, which might be time consuming. We suggest to use HMMER 3.1b2 or 3.2.1 for that purpose.

Dependencies:

• BioPython tested with v1.7x
• MUSCLE Aligner tested with v3.8.31
• ncbi-blast+ v2.3.0 or higher

Usage:

usage: compleconta.py [-h] [--margin MARGIN] [--majority MAJORITY]
                      [--rank RANK] [--aai AAI] [-o TAXONOMY_OUTPUT]
                      [--threads N_BLAST_THREADS] [--muscle MUSCLE_EXECUTABLE]
                      [--blast BLAST_EXECUTABLE] [--database DATABASE]
                      input.faa input.faa.out

Completeness and Contamination estimation using EggNOG-profiles

positional arguments:
  input.faa             genome proteome file
  input.faa.out         tab separated file with bactNOG classification of
                        proteome input file

optional arguments:
  -h, --help            show this help message and exit
  --margin MARGIN       Taxonomy: fraction margin for hits taken relative to
                        bitscore of best hit in blastp (default: 0.9)
  --majority MAJORITY   Taxonomy: majority threshold for fraction of paths
                        that support the lowest common ancestor (default: 0.9)
  --rank RANK           Taxonomy: lowest standard rank to be reported. 0:
                        species, 1: genus, ... 5: phylum (default: 1)
  --aai AAI             Contamination: amino acid identity to which the
                        multiple marker genes are considered to be strain
                        heterogenic (default: 0.9)
  -o TAXONOMY_OUTPUT    Taxonomy: file to write additional taxonomic
                        information for each marker gene, if not set,
                        information is omitted (default: None)
  --threads N_BLAST_THREADS
                        Taxonomy: number of parallel blastp jobs run (default:
                        5)
  --muscle MUSCLE_EXECUTABLE
                        Path to the muscle executable (default: None)
  --blast BLAST_EXECUTABLE
                        Path to the blast executable (makeblastdb) (default:
                        None)
  --database DATABASE   database which was used for annotation (default: auto)

Positional argument 1 should be the fasta format proteome file, positional argument 2 the corresponding classification output as tab-separated file where column 1 is the sequence identifier of the fasta file and column 2 the ENOG cluster id to which it was classified. Example files can be found in the example directory.

Output:

The output is written to stdout and provides Completeness, Contamination, Strain Heterogeneity (all in fractions), ncbi taxon ID, taxon name and taxon rank.

The output of the example for EggNOG 5.0 is*:

Comp.   Cont.   St. Het.    ncbi_taxid  taxon_name  taxon_rank
1.0000  0.0000  0.0000  85015   Nocardioidaceae family

*Note that the result may differ with other classification tools or databases used.

The optional output file contains detailed information about the marker genes that were used and their taxonomic information. Note that the majority cutoff for proceding to a lower level is 0.9, the ranks below are still reported in the output file, but not used for the final LCA.

LCA path and percentage of marker genes assignment:
Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 0.94    Nocardioidaceae 0.94

LCA per sequence of identified marker genes:i
WP_008360965.1  COG0048 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 0.99 Streptomycetales 0.31   Streptomycetaceae 0.31  Streptomyces 0.29
WP_008362892.1  COG0081 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00    Pimelobacter 0.50
WP_040755997.1  COG0096 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00    Nocardioides 0.67
WP_008363218.1  COG0130 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00
WP_008363208.1  COG0184 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00    Nocardioides 0.50
WP_008361063.1  COG0203 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00
WP_008363328.1  COG0228 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00
WP_008362879.1  COG0244 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00
WP_040755263.1  COG0261 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00
WP_008363690.1  COG0268 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00
WP_008359372.1  COG0292 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00    Nocardioides 0.50
WP_008363321.1  COG0335 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00    Nocardioides 0.50
WP_008355442.1  COG0359 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00
WP_008355448.1  COG0360 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00
WP_008358493.1  COG0536 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00
WP_008358688.1  COG0691 Bacteria 1.00   Actinobacteria 1.00 Actinobacteria 1.00 Propionibacteriales 1.00    Nocardioidaceae 1.00

Setup:

At the moment, there is no special setup required.

# Clone repository
git clone https://github.com/phyden/compleconta
# Change to directory
cd compleconta
# Run to display useage
./compleconta.py -h

Both the reduced taxonomy files (names.dmp and nodes.dmp) and the databases which were created from the bactNOG raw alignments are located in the data folder. The tool is ready to run, and will create the indices for the database files on execution if non existent. The script to prepare the database from EggNOG 4.5 is provided: prepare_blast_database.sh. To include other databases this script requires slight adaptions.

Please file an issue or contact the author if you need assistance.