OMArk Workflow¶

OMArk (Orthology MARKer) is a proteome quality assessment tool that evaluates completeness based on the presence of conserved genes from the target lineage.

Overview¶

OMArk provides an orthology-based assessment of proteome quality by comparing your protein set against a database of conserved orthologous groups. It identifies complete, fragmented, missing, and potentially contaminant sequences.

How OMArk Works¶

1. OMAmer Mapping: Maps protein sequences to hierarchical orthologous groups (HOGs)
2. Lineage Assignment: Determines the taxonomic placement of each protein
3. Completeness Assessment: Evaluates expected gene presence based on lineage
4. Consistency Check: Identifies potential contamination or misassignments

Key Features¶

• Lineage-Specific: Uses taxonomically appropriate conserved gene sets
• Contamination Detection: Identifies sequences inconsistent with target lineage
• Fragmentation Analysis: Distinguishes complete vs. partial genes
• Placement Verification: Validates taxonomic assignments

Running OMArk¶

Basic Usage¶

nextflow run main.nf \
  --csvFile proteomes.csv \
  --run_omark \
  --host mysql-server.example.com \
  --port 3306 \
  --user_r ensro \
  --outdir results

Required Inputs¶

Create a CSV file with these columns:

dbname,species_id,taxon_id
danio_rerio_core_110_11,1,7955
xenopus_tropicalis_core_110_10,1,8364
gallus_gallus_core_110_7,1,9031

Configuration Parameters¶

Custom Database¶

nextflow run main.nf \
  --csvFile proteomes.csv \
  --run_omark \
  --omamer_database /data/omamer/LUCA.h5 \
  --host mysql-server.example.com \
  --user_r ensro

Understanding Results¶

Output File¶

OMArk generates a detailed summary file: {sample}_omark_proteins_detailed_summary.txt

Example Output¶

# OMArk Analysis Summary
# Database: danio_rerio_core_110_11
# Taxon ID: 7955 (Danio rerio)
# Date: 2024-01-15

Completeness Assessment:
========================
Expected HOGs: 3,640
Complete: 3,521 (96.7%)
Fragmented: 87 (2.4%)
Missing: 32 (0.9%)

Consistency Analysis:
====================
Consistent: 3,598 (98.8%)
Inconsistent: 10 (0.3%)

Result Categories¶

Completeness Scores¶

Overall Completeness = Complete / Expected × 100

Consistency Scores¶

Consistency = Consistent / Total × 100

OMArk vs. BUSCO¶

Both tools assess proteome completeness but use different approaches:

Interpreting Results¶

High Completeness, High Consistency¶

Complete: 97.5%
Consistent: 99.2%

✅ Excellent Quality - Proteome is complete and clean - No significant contamination - Ready for downstream analyses

High Completeness, Low Consistency¶

Complete: 96.8%
Consistent: 89.5%

⚠️ Contamination Issues - Many genes present but from wrong lineage - Possible contamination or mixed sample - Check for foreign sequences

Action: - Review inconsistent proteins - Screen for contamination - Verify sample purity

Low Completeness, High Consistency¶

Complete: 82.3%
Consistent: 98.9%

⚠️ Annotation Gaps - Correct lineage but missing genes - Annotation incomplete or gene loss - Assembly may have gaps

Action: - Review annotation pipeline - Check assembly completeness - Consider re-annotation

Low Completeness, Low Consistency¶

Complete: 78.5%
Consistent: 87.2%

❌ Multiple Issues - Both completeness and contamination problems - Severe quality issues - Requires thorough investigation

Action: - Comprehensive QC review - Check sample identity - Validate assembly and annotation

Use Cases¶

1. Contamination Screening¶

OMArk excels at detecting contamination:

# Screen multiple assemblies
nextflow run main.nf \
  --csvFile new_assemblies.csv \
  --run_omark \
  --host mysql-server.example.com \
  --user_r ensro

Look for: - Low consistency scores (<95%) - Inconsistent proteins from distantly related lineages - Patterns suggesting systematic contamination

2. Pre-Publication QC¶

Validate proteome quality before public release:

# Run both OMArk and BUSCO
nextflow run main.nf \
  --csvFile release_genomes.csv \
  --run_omark \
  --run_busco_core \
  --busco_mode protein \
  --host mysql-server.example.com \
  --user_r ensro

3. Comparative Proteomics¶

Assess relative quality across multiple species:

dbname,species_id,taxon_id
species_a_core_110_1,1,12345
species_b_core_110_1,1,12346
species_c_core_110_1,1,12347

nextflow run main.nf \
  --csvFile comparative_set.csv \
  --run_omark \
  --host mysql-server.example.com \
  --user_r ensro

4. Annotation Pipeline Validation¶

Test annotation pipeline improvements:

# Compare v1 vs. v2 annotations
cat > comparison.csv << EOF
dbname,species_id,taxon_id
species_v1_core,1,9606
species_v2_core,1,9606
EOF

nextflow run main.nf \
  --csvFile comparison.csv \
  --run_omark \
  --host mysql-server.example.com \
  --user_r ensro

Output Files¶

Directory Structure¶

results/
└── omark/
    ├── sample1_omark_proteins_detailed_summary.txt
    ├── sample2_omark_proteins_detailed_summary.txt
    └── sample3_omark_proteins_detailed_summary.txt

File Contents¶

The detailed summary includes:

1. Header: Metadata (database, taxon, date)
2. Completeness Metrics: HOG presence/absence
3. Consistency Metrics: Lineage validation
4. Fragmentation Details: Partial gene information
5. Inconsistent Proteins: List of potentially contaminating sequences

Troubleshooting¶

Low Completeness with BUSCO OK¶

Problem: OMArk shows low completeness but BUSCO is good

Possible causes: - Different gene sets used (OMArk uses broader HOGs) - Lineage-specific gene expansions/losses - Different sensitivity thresholds

Solution: Review both results; some divergence is normal.

High Inconsistency Rate¶

Problem: >10% inconsistent proteins

Possible causes: 1. Contamination: Foreign DNA in sample 2. Wrong taxon ID: Incorrect species assignment 3. Horizontal gene transfer: Natural (e.g., in bacteria) 4. Symbionts: Co-sequenced organisms

Solutions: - Check sample purity - Verify taxon ID is correct - Screen for contamination with tools like BlobTools - Review inconsistent sequences manually

Missing OMAmer Database¶

Problem: Pipeline cannot find OMAmer database

Solution: Download and specify:

# Download from OMArk project
wget https://omabrowser.org/All/LUCA.h5

# Use in pipeline
nextflow run main.nf \
  --csvFile proteomes.csv \
  --run_omark \
  --omamer_database /data/LUCA.h5

Very Long Runtime¶

Problem: OMArk taking too long

Possible causes: - Very large proteome (>50,000 proteins) - Slow I/O on database file

Solutions: - Use local SSD for OMAmer database - Split large proteomes if possible - Increase computational resources

Best Practices¶

Advanced Topics¶

Custom OMAmer Databases¶

For specialized analyses, you can create custom HOG databases:

1. Extract relevant clade from OMA
2. Build OMAmer index
3. Point pipeline to custom database

Integration with Contamination Tools¶

Combine OMArk with other contamination detection:

# 1. Run OMArk
nextflow run main.nf --csvFile genomes.csv --run_omark

# 2. Extract inconsistent proteins
grep "Inconsistent" results/omark/*_detailed_summary.txt

# 3. Further screen with BlobTools or similar

Batch Analysis¶

Process hundreds of proteomes:

# Generate CSV for all databases
mysql -h mysql-server.example.com -u ensro -e "
  SELECT CONCAT(db_name, ',1,', meta_value) 
  FROM information_schema.schemata 
  JOIN meta ON db_name = database()
  WHERE db_name LIKE '%_core_%' 
  AND meta_key = 'species.taxonomy_id'
" > all_cores.csv

# Run OMArk on all
nextflow run main.nf \
  --csvFile all_cores.csv \
  --run_omark \
  --host mysql-server.example.com \
  --user_r ensro

Performance Considerations¶

Resource Requirements¶

Optimization¶

# Increase parallelization
nextflow run main.nf \
  --csvFile proteomes.csv \
  --run_omark \
  --max_cpus 32 \
  --max_memory 64.GB

References¶

Primary Citation¶

• Nevers et al. (2022). OMArk: proteome quality assessment using the OMA database. bioRxiv. DOI: 10.1101/2022.11.25.517970

Additional Resources¶

• OMArk GitHub
• OMAmer Documentation
• OMA Browser

Next Steps¶

• BUSCO Workflow - Complementary completeness assessment
• Ensembl Stats - Generate database statistics
• Output Documentation - Detailed output specifications
• Troubleshooting - Common issues and solutions