kegg-pathways-completeness tool

This tool computes the completeness of KEGG pathway modules for a given set of KEGG Orthologues (KOs) based on their presence/absence.

The current version includes 570 KEGG modules (updated 01/04/2026).

Please, read the Theory & Background section for a detailed explanation.

Table of Contents

• Installation
• Prerequisites
• Quick Start
• Detailed Usage
  • give_completeness
  • plot_modules_graphs
• Module Data Files
• Output Files
• Theory & Background
• Updating Module Data
• Complete Workflow
• Citation

Installation

The tool is available via PyPI, Bioconda, and Docker.

Install with pip

pip install kegg-pathways-completeness

Install with bioconda

conda install -c bioconda kegg-pathways-completeness

See bioconda recipe for details.

Docker

docker pull quay.io/biocontainers/kegg-pathways-completeness

Install from source (for development)

git clone https://github.com/EBI-Metagenomics/kegg-pathways-completeness-tool.git
cd kegg-pathways-completeness-tool
pip install -e .

Prerequisites

• Python: 3.8 or higher
• graphviz: Required for pathway visualization (install via system package manager)
• HMMER (optional): For annotating protein sequences with KOs

Quick Start

Tool uses pre-generated files modules_table.tsv and graphs.pkl described in Module Data Files.

Option 1: From a list of KOs

Input format (example): File with KO identifiers

K00001,K00002,K00003

command:

give_completeness \
  --input-list kos_list.txt \
  --list-separator ',' \
  --outprefix my_analysis

Option 2: From per-contig KO annotations

Input format (example): Tab-separated file with contig names and KOs

contig_1    K00001    K00002    K00003
contig_2    K00004    K00005

command:

give_completeness \
  --input ko_annotations.tsv \
  --outprefix my_analysis 

Detailed Usage

give_completeness

Calculate KEGG pathway module completeness from KO annotations.

Required Arguments

Input (choose one):

• -i, --input <FILE>: Tab-separated file with contig names and KOs (example)
• -l, --input-list <FILE>: List of KOs, separated by delimiter (example)

Module data:

• -t, --modules-table <FILE>: Module information in TSV format (columns: module, definition, name, class)
  • Default: Uses packaged kegg_pathways_completeness/pathways_data/modules_table.tsv
• -g, --graphs <FILE>: Custom graphs file (default: uses packaged kegg_pathways_completeness/pathways_data/graphs.pkl)

Optional Arguments

• -s, --list-separator <CHAR>: Separator for --input-list (default: ,)
• -o, --outdir <DIR>: Output directory (default: current directory)
• -r, --outprefix <PREFIX>: Prefix for output files (default: summary.kegg)
• -m, --add-per-contig: Generate per-contig completeness table
• -w, --include-weights: Include KO weights in output (e.g., K00942(0.25))
• -p, --plot-pathways: Generate pathway visualization plots
• -v, --verbose: Enable verbose logging

Examples

# Basic usage with KO list
give_completeness \
  --input-list kos.txt \
  --modules-table kegg_pathways_completeness/pathways_data/modules_table.tsv \
  --graphs kegg_pathways_completeness/pathways_data/graphs.pkl \
  --outprefix sample1

# Full analysis with per-contig results, weights, and plots
give_completeness \
  --input ko_annotations.tsv \
  --outprefix sample1 \
  --add-per-contig \
  --include-weights \
  --plot-pathways \
  --outdir results/

# Using custom module data
give_completeness \
  --input ko_annotations.tsv \
  --modules-table custom_modules.tsv \
  --graphs custom_graphs.pkl \
  --outdir custom_analysis

plot_modules_graphs

Generate pathway visualization with KOs highlighted.

Note: Requires graphviz to be installed.

Required Arguments

Input (choose one):

• -i, --input-completeness <FILE>: Completeness output from give_completeness
• -m, --modules <ID> [<ID> ...]: Module IDs to plot (can be specified multiple times)
• -l, --modules-file <FILE>: File containing module IDs (one per line)

Graphs:

• -g, --graphs <FILE>: Graphs pickle file (default: pathways_data/graphs.pkl)

Optional Arguments

• -s, --file-separator <CHAR>: Separator in modules file (default: newline)
• -o, --outdir <DIR>: Output directory (default: pathways_plots)
• --use-pydot: Use pydot instead of graphviz backend

Examples

# Plot from completeness results
plot_modules_graphs \
  -i sample1_pathways.tsv \
  -g kegg_pathways_completeness/pathways_data/graphs.pkl \
  -o pathway_plots

# Plot specific modules
plot_modules_graphs \
  -m M00001 M00002 M00050 \
  -g kegg_pathways_completeness/pathways_data/graphs.pkl

# Plot modules from file
plot_modules_graphs \
  -l modules_of_interest.txt \
  -g kegg_pathways_completeness/pathways_data/graphs.pkl

# Use pydot backend
plot_modules_graphs \
  -i sample1_pathways.tsv \
  -g kegg_pathways_completeness/pathways_data/graphs.pkl \
  --use-pydot

Output:

• PNG images with pathways (present KOs in red)
• DOT source files (when using --use-pydot)

More visualization examples: test output plots

Module Data Files

The package includes pre-generated data files in pathways_data/:

modules_table.tsv

Unified TSV file with all module information.

Columns:

• module: Module ID (e.g., M00001)
• definition: KEGG module definition in KO notation
• name: Module name/description
• class: Module classification/category

File: modules_table.tsv

graphs.pkl

Pre-parsed NetworkX directed graphs for all modules. Each pathway definition has been converted to a graph structure for completeness calculation.

File: graphs.pkl

Output Files

Pathway completeness table (*_pathways.tsv)

Main output with completeness scores for all detected pathways.

Columns:

• module_accession: Module ID
• completeness: Completeness percentage (0-100)
• pathway_name: Module name
• pathway_class: Module classification
• matching_ko: KOs found in the pathway
• missing_ko: KOs required but not found

Example: test_kos_pathways.tsv

Per-contig completeness (*_contigs.tsv)

Generated with -m/--add-per-contig flag. Same format as above but with contig name as first column.

Example: test_pathway_contigs.tsv

Weighted output (*.with_weights.tsv)

Generated with -w/--include-weights flag. Includes weight values for each KO in parentheses (e.g., K00942(0.25) means weight = 0.25).

Example: test_weights_pathways.with_weights.tsv

Pathway plots (pathways_plots/)

Generated with -p/--plot-pathways flag. Contains:

• PNG images with pathway graphs
• Present KOs highlighted in red
• Missing KOs in black

Example directory: pathways_plots/

Theory & Background

How KEGG modules are represented

KEGG provides pathway definitions as logical expressions of KOs.

Example: (K00844,K12407) (K01810,K06859,K13810) (K00850,K16370) K00918

Notation:

• Space = AND (all components required)
• Comma = OR (any one component required)
• Plus (+) = Essential component
• Minus (-) = Optional component
• Double minus (–) = Missing optional (replaced with K00000 with 0 weight)
• Newline = Mediator (multi-line definitions use AND between lines)

Examples:

• M00014 - module with missing optionals (graph)
• M00031 - module with mediators (graph)

Pathway to graph conversion

Each KEGG module definition is converted into a directed graph using NetworkX:

• Start node: 0
• End node: 1
• Edges: Represent KOs with assigned weights

Completeness calculation

Algorithm:

1. Each edge in the graph has a weight based on its importance (calculated from pathway structure)
2. For a given set of KOs:
   • Present KOs → edge weight = original weight
   • Missing KOs → edge weight = 0
3. Find the path from node 0 to node 1 with minimum (current_weight / original_weight) ratio
4. Calculate completeness:

completeness = (path_weight / max_path_weight) × 100%

Note on mediators: Some modules have multi-line definitions where each line represents a mediator component. All mediators are connected with AND operators. The complete list of modules with mediators is in definition_separated.txt.

Updating Module Data

To update module data to the latest KEGG version, see the update documentation.

The update process includes:

1. Fetching latest module definitions from KEGG API
2. Generating the unified modules_table.tsv
3. Creating NetworkX graphs from module definitions
4. Validating and testing the updated data

Complete Workflow

From raw sequences to pathway completeness

# Step 1: Annotate protein sequences using HMMER
# Download KEGG profiles database (KOfam) from KEGG
hmmscan --domtblout hmmer_output.tbl \
  --cpu 4 \
  profiles.hmm \
  sequences.faa

# Step 2: Parse HMMER output to extract KO annotations per contig
parse_hmmer_table \
  -i hmmer_output.tbl \
  -f sequences.faa \
  -t hmmscan \
  -o ko_annotations.tsv

# Step 3: Calculate pathway completeness
give_completeness \
  -i ko_annotations.tsv \
  -t kegg_pathways_completeness/pathways_data/modules_table.tsv \
  -r my_sample \
  -m \
  -w \
  -p

# Step 4 (optional): Visualize specific modules
plot_modules_graphs \
  -i my_sample_pathways.tsv \
  -g kegg_pathways_completeness/pathways_data/graphs.pkl \
  -o pathway_plots

See detailed documentation about hmmer usage and parsing.

---

Citation

If you use this tool in your research, please cite

Richardson L, Allen B, Baldi G, Beracochea M, Bileschi ML, Burdett T, et al. MGnify: the microbiome sequence data analysis resource in 2023 [Internet]. Vol. 51, Nucleic Acids Research. Oxford University Press (OUP); 2022. p. D753–9. Available from: http://dx.doi.org/10.1093/nar/gkac1080.

Issues & Contributions: Report bugs or request features on GitHub Issues

License: Apache License 2.0