agentleFS
dailycafidailycafi/metabolism-skills/skills/systems-biology/model-curation/SKILL.md

bio-systems-biology-model-curation

Skill7 starsChanged 8 months agoMITView on GitHub

Validate and curate genome-scale metabolic models using memote quality scores, gap filling, and SBML compliance checks. Use when: user wants to check model quality, run memote scoring, fill gaps in a metabolic model, fix dead-end metabolites, or prepare a GEM for publication. Triggers: model quality, memote score, gap filling, model curation, SBML validation, dead-end metabolites, blocked reactions, mass/charge balance, model debugging, metabolic model QC, improve draft model, annotation completeness.

  • Installs packages

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before using code patterns, verify installed versions match.

    line 16
  2. Distinctive:

    If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API …

    line 20
  3. Distinctive:

    Approach: Run memote snapshot to score the model against SBML community standards, then use the Python API to inspect individual test …

    line 33

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Version Compatibility

Reference examples tested with: COBRApy 0.29+

Before using code patterns, verify installed versions match. If versions differ:

  • Python: pip show <package> then help(module.function) to check signatures
  • CLI: <tool> --version then <tool> --help to 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.

Model Curation

"Validate and improve the quality of my metabolic model" → Score a genome-scale model against SBML community standards using memote, then gap-fill blocked reactions and fix stoichiometric inconsistencies using COBRApy to ensure biologically meaningful predictions.

  • CLI: memote report snapshot for quality scoring
  • Python: cobra.flux_analysis.gapfilling.gapfill() for gap-filling

Memote Quality Assessment

Goal: Evaluate the quality and standards compliance of a genome-scale metabolic model to identify areas needing curation.

Approach: Run memote snapshot to score the model against SBML community standards, then use the Python API to inspect individual test failures and guide manual fixes.

# Install memote
pip install memote

# Run full quality report
memote report snapshot model.xml --filename report.html

# Quick score
memote run model.xml

# Continuous integration testing
memote run --pytest-args "--tb=short" model.xml

Memote Python API

import memote
import cobra

model = cobra.io.read_sbml_model('model.xml')

# Run all tests
result = memote.suite.api.run(model)

# Get score breakdown
scores = memote.suite.api.snapshot(model)
print(f"Total score: {scores['score']['total_score']:.2%}")

# Detailed test results
for test_name, test_result in scores['tests'].items():
    if not test_result['passed']:
        print(f"Failed: {test_name}")

Gap-Filling

import cobra
from cobra.flux_analysis import gapfill

model = cobra.io.read_sbml_model('model.xml')

# Load universal reaction database
universal = cobra.io.read_sbml_model('universal_model.xml')

# Find reactions to add for growth
# demand: reaction to optimize (usually biomass exchange)
# iterations: number of alternative solutions
solution = gapfill(model, universal,
                   demand=model.reactions.BIOMASS,
                   iterations=5)

# solution contains list of reaction sets to add
for i, rxn_set in enumerate(solution):
    print(f'Solution {i+1}: {[r.id for r in rxn_set]}')

# Add first solution
for rxn in solution[0]:
    model.add_reactions([rxn])

Identify Dead-End Metabolites

def find_dead_end_metabolites(model):
    '''Find metabolites that cannot be produced or consumed

    Dead-end metabolites indicate:
    - Missing reactions in the network
    - Incorrect reaction stoichiometry
    - Incomplete pathways
    '''
    dead_ends = []

    for met in model.metabolites:
        producing = [r for r in met.reactions if r.get_coefficient(met) > 0]
        consuming = [r for r in met.reactions if r.get_coefficient(met) < 0]

        if not producing or not consuming:
            dead_ends.append({
                'metabolite': met.id,
                'name': met.name,
                'producers': len(producing),
                'consumers': len(consuming)
            })

    return dead_ends


dead_ends = find_dead_end_metabolites(model)
print(f'Found {len(dead_ends)} dead-end metabolites')

Check Mass and Charge Balance

def check_reaction_balance(reaction):
    '''Check if reaction is mass and charge balanced

    Unbalanced reactions indicate:
    - Missing metabolites
    - Wrong stoichiometry
    - Proton accounting issues
    '''
    mass_balance = {}
    charge_balance = 0

    for met, coef in reaction.metabolites.items():
        # Check mass
        if met.formula:
            for element, count in met.elements.items():
                mass_balance[element] = mass_balance.get(element, 0) + coef * count

        # Check charge
        if met.charge is not None:
            charge_balance += coef * met.charge

    is_balanced = all(abs(v) < 1e-6 for v in mass_balance.values())
    is_charge_balanced = abs(charge_balance) < 1e-6

    return {
        'mass_balanced': is_balanced,
        'charge_balanced': is_charge_balanced,
        'mass_imbalance': {k: v for k, v in mass_balance.items() if abs(v) > 1e-6}
    }


# Check all reactions
unbalanced = []
for rxn in model.reactions:
    result = check_reaction_balance(rxn)
    if not result['mass_balanced']:
        unbalanced.append((rxn.id, result['mass_imbalance']))

Fix Gene-Protein-Reaction Rules

def standardize_gpr(model):
    '''Standardize gene-protein-reaction rules

    GPR format: (gene1 and gene2) or gene3
    - 'and' = protein complex (all genes required)
    - 'or' = isozymes (any gene sufficient)
    '''
    for rxn in model.reactions:
        if rxn.gene_reaction_rule:
            # Standardize formatting
            rule = rxn.gene_reaction_rule
            rule = rule.replace(' AND ', ' and ')
            rule = rule.replace(' OR ', ' or ')
            rxn.gene_reaction_rule = rule


def identify_orphan_reactions(model):
    '''Find reactions without gene associations

    Orphan reactions may be:
    - Spontaneous reactions
    - Unannotated genes
    - Transport reactions (often orphan)
    '''
    orphans = [r for r in model.reactions if not r.genes]

    # Classify orphans
    exchange = [r for r in orphans if r in model.exchanges]
    transport = [r for r in orphans if 'transport' in r.name.lower() or 't_' in r.id.lower()]
    other = [r for r in orphans if r not in exchange and r not in transport]

    return {
        'exchange': len(exchange),
        'transport': len(transport),
        'other': len(other),
        'total': len(orphans)
    }

Annotation Standards

def add_standard_annotations(model):
    '''Add standard database annotations

    Required annotations for SBML compliance:
    - KEGG IDs for reactions and metabolites
    - ChEBI IDs for metabolites
    - BiGG IDs if applicable
    '''
    for met in model.metabolites:
        if not hasattr(met, 'annotation'):
            met.annotation = {}

        # Add SBO term for metabolite
        met.annotation['sbo'] = 'SBO:0000247'  # Simple chemical

    for rxn in model.reactions:
        if not hasattr(rxn, 'annotation'):
            rxn.annotation = {}

        # Add SBO term based on reaction type
        if rxn in model.exchanges:
            rxn.annotation['sbo'] = 'SBO:0000627'  # Exchange
        else:
            rxn.annotation['sbo'] = 'SBO:0000176'  # Biochemical reaction

Related Skills

  • systems-biology/metabolic-reconstruction - Generate draft models
  • systems-biology/flux-balance-analysis - Test curated models
  • pathway-analysis/kegg-pathways - Add KEGG annotations

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.