Agent skill
bio-metagenomics-functional-profiling
Profile functional potential of metagenomes using HUMAnN3 and similar tools. Use when obtaining pathway abundances, gene family counts, or functional annotations from metagenomic data.
Install this agent skill to your Project
npx add-skill https://github.com/FreedomIntelligence/OpenClaw-Medical-Skills/tree/main/skills/bio-metagenomics-functional-profiling
SKILL.md
Version Compatibility
Reference examples tested with: HUMAnN 3.8+, MetaPhlAn 4.1+, matplotlib 3.8+, pandas 2.2+, scanpy 1.10+, scipy 1.12+, seaborn 0.13+
Before using code patterns, verify installed versions match. If versions differ:
- Python:
pip show <package>thenhelp(module.function)to check signatures - CLI:
<tool> --versionthen<tool> --helpto confirm flags
If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.
Functional Profiling
"What metabolic pathways are present in my metagenome?" → Profile functional potential of metagenomic samples to obtain pathway abundances and gene family counts using translated search against UniRef and MetaCyc.
- CLI:
humann --input reads.fastq --output results/(HUMAnN3)
Profile the functional potential of metagenomic samples using HUMAnN3 to get pathway and gene family abundances.
HUMAnN3 Workflow
Installation
# Install via conda (recommended)
conda create -n humann -c bioconda humann
conda activate humann
# Download databases
humann_databases --download chocophlan full /path/to/databases
humann_databases --download uniref uniref90_diamond /path/to/databases
# Update config with database paths
humann_config --update database_folders nucleotide /path/to/databases/chocophlan
humann_config --update database_folders protein /path/to/databases/uniref
Basic Usage
# Run HUMAnN3 on a single sample
humann --input sample.fastq.gz --output sample_humann
# With MetaPhlAn taxonomic profile (faster)
humann --input sample.fastq.gz \
--taxonomic-profile sample_metaphlan.txt \
--output sample_humann
# Paired-end reads (concatenate first)
cat sample_R1.fq.gz sample_R2.fq.gz > sample_concat.fq.gz
humann --input sample_concat.fq.gz --output sample_humann
Output Files
sample_humann/
├── sample_genefamilies.tsv # Gene family abundances (UniRef90)
├── sample_pathabundance.tsv # MetaCyc pathway abundances
├── sample_pathcoverage.tsv # Pathway coverage (0-1)
└── sample_humann_temp/ # Intermediate files
Output Format
Gene Families
# Gene Family sample_Abundance-RPKs
UniRef90_A0A000|g__Bacteroides.s__Bacteroides_vulgatus 123.45
UniRef90_A0A001|unclassified 67.89
UNMAPPED 1000.0
Pathway Abundance
# Pathway sample_Abundance
PWY-5100: pyruvate fermentation 456.78
PWY-5100|g__Bacteroides.s__Bacteroides_vulgatus 234.56
PWY-5100|unclassified 222.22
Batch Processing
# Process multiple samples
for fq in *.fastq.gz; do
sample=$(basename $fq .fastq.gz)
humann --input $fq --output ${sample}_humann --threads 8
done
# Join tables across samples
humann_join_tables -i . -o merged_genefamilies.tsv --file_name genefamilies
humann_join_tables -i . -o merged_pathabundance.tsv --file_name pathabundance
Normalization
# Normalize to relative abundance
humann_renorm_table -i merged_genefamilies.tsv \
-o genefamilies_relab.tsv \
-u relab
# Normalize to copies per million (CPM)
humann_renorm_table -i merged_pathabundance.tsv \
-o pathabundance_cpm.tsv \
-u cpm
Regroup Gene Families
# Regroup to different functional categories
# EC numbers
humann_regroup_table -i genefamilies.tsv \
-g uniref90_level4ec \
-o genefamilies_ec.tsv
# KEGG Orthologs
humann_regroup_table -i genefamilies.tsv \
-g uniref90_ko \
-o genefamilies_ko.tsv
# GO terms
humann_regroup_table -i genefamilies.tsv \
-g uniref90_go \
-o genefamilies_go.tsv
# Pfam domains
humann_regroup_table -i genefamilies.tsv \
-g uniref90_pfam \
-o genefamilies_pfam.tsv
Stratification
Split by Organism
# Unstratify (remove organism info, sum across species)
humann_split_stratified_table -i merged_pathabundance.tsv \
-o .
# Creates: merged_pathabundance_unstratified.tsv
# merged_pathabundance_stratified.tsv
Species Contributions
import pandas as pd
df = pd.read_csv('merged_pathabundance.tsv', sep='\t', index_col=0)
unstratified = df[~df.index.str.contains('\\|')]
stratified = df[df.index.str.contains('\\|')]
def get_species_contrib(pathway, df):
'''Get species contributions to a pathway'''
mask = df.index.str.startswith(pathway + '|')
return df[mask]
contrib = get_species_contrib('PWY-5100', stratified)
Quality Control
# Check unmapped and unintegrated
humann_barplot -i merged_pathabundance.tsv \
-o pathabundance_barplot.png \
--focal-feature UNMAPPED
Key QC Metrics
| Metric | Good | Concerning |
|---|---|---|
| UNMAPPED (gene families) | <30% | >50% |
| UNINTEGRATED (pathways) | <40% | >60% |
| Pathway coverage | >0.5 | <0.3 |
Differential Analysis
LEfSe Format
# Format for LEfSe
humann_join_tables -i . -o merged.tsv --file_name pathabundance
humann_renorm_table -i merged.tsv -o merged_relab.tsv -u relab
Python Analysis
Goal: Identify differentially abundant metabolic pathways between conditions from HUMAnN3 output.
Approach: Load unstratified pathway abundances, split samples by condition using metadata, run Mann-Whitney U tests per pathway, and apply FDR correction.
import pandas as pd
from scipy import stats
df = pd.read_csv('pathabundance_cpm.tsv', sep='\t', index_col=0)
metadata = pd.read_csv('metadata.tsv', sep='\t', index_col=0)
group1 = metadata[metadata['condition'] == 'healthy'].index
group2 = metadata[metadata['condition'] == 'disease'].index
results = []
for pathway in df.index:
if '|' not in pathway and pathway != 'UNMAPPED':
vals1 = df.loc[pathway, group1]
vals2 = df.loc[pathway, group2]
stat, pval = stats.mannwhitneyu(vals1, vals2)
fc = vals2.mean() / (vals1.mean() + 1e-10)
results.append({'pathway': pathway, 'pvalue': pval, 'fold_change': fc})
results_df = pd.DataFrame(results)
results_df['padj'] = stats.false_discovery_control(results_df['pvalue'])
Visualization
import matplotlib.pyplot as plt
import seaborn as sns
df = pd.read_csv('pathabundance_relab.tsv', sep='\t', index_col=0)
df = df[~df.index.str.contains('\\|')]
df = df.drop(['UNMAPPED', 'UNINTEGRATED'], errors='ignore')
top = df.mean(axis=1).nlargest(20).index
plt.figure(figsize=(12, 8))
sns.heatmap(df.loc[top].T, cmap='viridis', xticklabels=True)
plt.tight_layout()
plt.savefig('pathway_heatmap.png')
Related Skills
- metagenomics/metaphlan-profiling - Taxonomic profiling (input for HUMAnN)
- metagenomics/kraken-classification - Alternative taxonomy
- metagenomics/metagenome-visualization - Visualization methods
- pathway-analysis/kegg-pathways - KEGG pathway interpretation
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
vcf-annotator
Annotate VCF variants with VEP, ClinVar, gnomAD frequencies, and ancestry-aware context. Generates prioritised variant reports.
chemist-analyst
Analyzes events through chemistry lens using molecular structure, reaction mechanisms, thermodynamics, kinetics, and analytical techniques (spectroscopy, chromatography, mass spectrometry). Provides insights on chemical processes, material properties, reaction pathways, synthesis, and analytical methods. Use when: Chemical reactions, material analysis, synthesis planning, process optimization, environmental chemistry. Evaluates: Molecular structure, reaction mechanisms, yield, selectivity, safety, environmental impact.
bio-alignment-io
Read, write, and convert multiple sequence alignment files using Biopython Bio.AlignIO. Supports Clustal, PHYLIP, Stockholm, FASTA, Nexus, and other alignment formats for phylogenetics and conservation analysis. Use when reading, writing, or converting alignment file formats.
sleep-analyzer
分析睡眠数据、识别睡眠模式、评估睡眠质量,并提供个性化睡眠改善建议。支持与其他健康数据的关联分析。
metabolomics-workbench-database
Access NIH Metabolomics Workbench via REST API (4,200+ studies). Query metabolites, RefMet nomenclature, MS/NMR data, m/z searches, study metadata, for metabolomics and biomarker discovery.
bio-hi-c-analysis-matrix-operations
Balance, normalize, and transform Hi-C contact matrices using cooler and cooltools. Apply iterative correction (ICE), compute expected values, and generate observed/expected matrices. Use when normalizing or transforming Hi-C matrices.
Didn't find tool you were looking for?