RAPTOR Quick Reference Cheat Sheet

Print this for quick command reference!

⚡ Installation

# PyPI (quick)
pip install raptor-rnaseq

# Conda (with tools)
conda env create -f environment-full.yml
conda activate raptor-full

# Check version
raptor --version

🔥 Most Common Commands

Complete Workflow

# 1. Quality Check
raptor qc -c counts.csv -m metadata.csv -o qc/

# 2. Profile Data
raptor profile -c counts.csv -m metadata.csv -o profile/

# 3. Get Recommendation
raptor recommend -p profile/profile.json -o recommend/

# 4. Optimize Thresholds
raptor optimize -d de_results.csv -m fdr-control --fdr-target 0.05

# 5. Ensemble Analysis
raptor ensemble-compare --deseq2 d.csv --edger e.csv --limma l.csv -o ensemble/

📋 Module Quick Reference

Module 2: Quality Assessment

# Basic
raptor qc -c counts.csv -m metadata.csv

# With plots
raptor qc -c counts.csv -m metadata.csv --plot --plot-output qc.pdf

# Python
from raptor import quick_quality_check
report = quick_quality_check('counts.csv', 'metadata.csv')
print(f"Outliers: {report.outliers}")

Detects: Outliers using 6 methods (MAD, Isolation Forest, LOF, PCA, Clustering, Statistical)

Module 3: Data Profiler

# Basic
raptor profile -c counts.csv -m metadata.csv -g condition

# Python
from raptor import profile_data_quick
profile = profile_data_quick('counts.csv', 'metadata.csv', group_column='condition')
print(f"BCV: {profile.features['bcv']:.3f}")  # Most important!

Extracts: 32 features including BCV (key metric for pipeline selection)

BCV Guide:

< 0.2: Low variation → limma-voom
0.2-0.4: Moderate (typical) → DESeq2, edgeR
0.4: High variation → edgeR_robust

Module 4: Recommender

# ML-based (recommended)
raptor recommend -p profile.json -m ml

# Both methods
raptor recommend -p profile.json -m both --verbose-explanation

# Python
from raptor import recommend_pipeline
rec = recommend_pipeline(profile_file='profile.json', method='ml')
print(f"Use: {rec.pipeline_name} (confidence: {rec.confidence:.2f})")

Returns: Pipeline recommendation with confidence score and reasoning

Module 7: DE Import

# Import DESeq2
raptor import-de -i deseq2.csv -m deseq2 -o imported/

# Import edgeR
raptor import-de -i edger.csv -m edger

# Import limma
raptor import-de -i limma.csv -m limma

# Compare all
raptor compare-de deseq2.csv edger.csv limma.csv -o comparison/ --venn-diagram

# Python
from raptor import import_deseq2, compare_de_results
deseq2 = import_deseq2('deseq2.csv')
edger = import_edger('edger.csv')
comparison = compare_de_results(deseq2=deseq2, edger=edger)

Standardizes: Any DE format into common schema

Module 8: Parameter Optimization

# Ground Truth (simulated data)
raptor optimize -d de.csv -m ground-truth -g truth.csv

# FDR Control (most common)
raptor optimize -d de.csv -m fdr-control --fdr-target 0.05

# Stability (cross-validation)
raptor optimize -d de.csv -m stability --counts counts.csv --metadata metadata.csv

# Reproducibility (independent cohorts)
raptor optimize -d de.csv -m reproducibility --counts c1.csv --cohort2 c2.csv

# Python
from raptor import optimize_with_fdr_control
result = optimize_with_fdr_control(de_result, fdr_target=0.05)
print(f"Optimal thresholds: FDR={result.optimal_threshold['padj']}, log2FC={result.optimal_threshold['lfc']}")

Optimizes: FDR and log2FC thresholds using 4 validated methods

Module 9: Ensemble Analysis

# Single method (Fisher's)
raptor ensemble -m fisher --deseq2 d.csv --edger e.csv --limma l.csv -o ens/

# Compare all 5 methods
raptor ensemble-compare --deseq2 d.csv --edger e.csv --limma l.csv -o comparison/

# Python
from raptor import ensemble_fisher, ensemble_brown, ensemble_rra

# Fisher's Method
fisher = ensemble_fisher({
    'deseq2': deseq2_result,
    'edger': edger_result,
    'limma': limma_result
})

# Brown's Method (recommended - accounts for correlation)
brown = ensemble_brown({
    'deseq2': deseq2_result,
    'edger': edger_result,
    'limma': limma_result
})

print(f"Consensus genes: {len(brown.consensus_genes)}")

Combines: Multiple DE methods using 5 statistical approaches

🔧 Common Parameters

Parameter	Flag	Description	Example
Counts	`-c`	Count matrix	`-c counts.csv`
Metadata	`-m`	Sample metadata	`-m metadata.csv`
Output	`-o`	Output directory	`-o results/`
Group	`-g`	Group column	`-g condition`
Threshold	`--threshold`	Significance threshold	`--threshold 0.05`
Method	`-m`	Analysis method	`-m ml`

📊 Pipeline Quick Reference

Pipeline	Command	Memory	Time	Best For
salmon ⭐	`raptor pipeline run -n salmon`	8 GB	10-20 min	Recommended
kallisto	`raptor pipeline run -n kallisto`	4 GB	5-10 min	Speed
star_featurecounts	`raptor pipeline run -n star_featurecounts`	32 GB	40-70 min	Publication genes
star_rsem	`raptor pipeline run -n star_rsem`	32 GB	60-120 min	Isoforms

# List available
raptor pipeline list

# Run salmon
raptor pipeline run -n salmon -s samples.csv -g salmon_index/ -o results/ -t 8

🐍 Python API Quick Reference

Import Everything

from raptor import (
    # Module 2: QC
    quick_quality_check, DataQualityAssessor,
    
    # Module 3: Profiler
    profile_data_quick, RNAseqDataProfiler,
    
    # Module 4: Recommender
    recommend_pipeline, MLRecommender,
    
    # Module 7: DE Import
    import_deseq2, import_edger, import_limma, compare_de_results,
    
    # Module 8: Optimization
    optimize_with_ground_truth, optimize_with_fdr_control,
    optimize_with_stability, optimize_with_reproducibility,
    
    # Module 9: Ensemble
    ensemble_fisher, ensemble_brown, ensemble_rra,
    ensemble_voting, ensemble_weighted
)

Minimal Example

from raptor import quick_quality_check, profile_data_quick, ensemble_fisher

# QC
qc = quick_quality_check('counts.csv', 'metadata.csv')

# Profile
profile = profile_data_quick('counts.csv', 'metadata.csv', group_column='condition')

# Ensemble
result = ensemble_fisher({
    'deseq2': deseq2_result,
    'edger': edger_result
})

📁 File Formats

Count Matrix (counts.csv)

gene_id,sample_1,sample_2,sample_3,sample_4
ENSG00001,1523,1872,2103,2987
ENSG00002,456,523,678,789

Rows = genes
Columns = samples
Values = integer counts

Metadata (metadata.csv)

sample_id,condition,batch,replicate
sample_1,control,1,1
sample_2,control,1,2
sample_3,treatment,1,1
sample_4,treatment,2,2

Must have sample_id column
Must match count matrix columns
Include grouping variable (condition)

Sample Sheet for Pipelines (samples.csv)

sample_id,fastq_1,fastq_2,condition
sample_1,s1_R1.fq.gz,s1_R2.fq.gz,control
sample_2,s2_R1.fq.gz,s2_R2.fq.gz,treatment

⚡ Quick Decision Flowchart

Do you have count matrix?
  ├─ YES → Start with Module 2 (QC)
  └─ NO  → Run Module 5 (Pipeline) or use quick-count

Is data quality good?
  ├─ YES → Continue to Module 3 (Profile)
  └─ NO  → Remove outliers, check batch effects

What pipeline should I use?
  → Module 4 (Recommender) - Use ML method!

Have DE results from DESeq2/edgeR/limma?
  → Module 7 (Import) → Module 9 (Ensemble)

Need to optimize thresholds?
  → Module 8 (FDR Control method most common)

Want robust results?
  → Module 9 (Ensemble - Brown's method recommended)

🆘 Common Errors & Fixes

Error: "Too few genes"

# Check count matrix
head -n 5 counts.csv
wc -l counts.csv  # Should have >100 genes

# Filter low counts
python -c "
import pandas as pd
counts = pd.read_csv('counts.csv', index_col=0)
filtered = counts[counts.sum(axis=1) > 10]
filtered.to_csv('counts_filtered.csv')
"

Error: "Sample mismatch"

# Check sample IDs
head -n 1 counts.csv
cut -d',' -f1 metadata.csv

# They must match exactly!

Error: "Module not found"

# Reinstall
pip uninstall raptor-rnaseq
pip install raptor-rnaseq

# Verify
python -c "import raptor; print(raptor.__version__)"

📚 Get Help

# General help
raptor --help

# Module-specific help
raptor qc --help
raptor profile --help
raptor recommend --help
raptor optimize --help
raptor ensemble --help

# Check installation
raptor validate-installation

# Version
raptor --version

Online:

📖 Docs: docs/
🐛 Issues: GitHub Issues
💬 Discussions: GitHub Discussions
📧 Email: ayehbolouki1988@gmail.com

🎯 Workflow Templates

Template 1: Quick QC Only

raptor qc -c counts.csv -m metadata.csv

Template 2: QC + Profile

raptor qc -c counts.csv -m metadata.csv
raptor profile -c counts.csv -m metadata.csv

Template 3: Complete Analysis

raptor qc -c counts.csv -m metadata.csv -o qc/
raptor profile -c counts.csv -m metadata.csv -o profile/
raptor recommend -p profile/profile.json -o recommend/
raptor optimize -d de_results.csv -m fdr-control -o opt/
raptor ensemble-compare --deseq2 d.csv --edger e.csv --limma l.csv -o ens/

Template 4: Ensemble Only

raptor import-de -i deseq2.csv -m deseq2 -o imported/
raptor import-de -i edger.csv -m edger -o imported/
raptor ensemble-compare \
    --deseq2 imported/deseq2_standardized.csv \
    --edger imported/edger_standardized.csv \
    -o ensemble_results/

💡 Pro Tips

Always run QC first - outliers will mess up everything downstream
BCV is key - Check profile.bcv to understand your data
Use ML recommender - It's trained on real data
Brown's method for ensemble - Accounts for correlation between DESeq2/edgeR
FDR control for optimization - Most practical method
Keep intermediate files - Useful for troubleshooting
Document parameters - Save commands in script for reproducibility

Version: 2.2.0
Updated: March 2026

Save this file for quick reference!

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

RAPTOR Quick Reference Cheat Sheet

⚡ Installation

🔥 Most Common Commands

Complete Workflow

📋 Module Quick Reference

Module 2: Quality Assessment

Module 3: Data Profiler

Module 4: Recommender

Module 7: DE Import

Module 8: Parameter Optimization

Module 9: Ensemble Analysis

🔧 Common Parameters

📊 Pipeline Quick Reference

🐍 Python API Quick Reference

Import Everything

Minimal Example

📁 File Formats

Count Matrix (counts.csv)

Metadata (metadata.csv)

Sample Sheet for Pipelines (samples.csv)

⚡ Quick Decision Flowchart

🆘 Common Errors & Fixes

Error: "Too few genes"

Error: "Sample mismatch"

Error: "Module not found"

📚 Get Help

🎯 Workflow Templates

Template 1: Quick QC Only

Template 2: QC + Profile

Template 3: Complete Analysis

Template 4: Ensemble Only

💡 Pro Tips

FilesExpand file tree

RAPTOR_QUICK_REFERENCE.md

Latest commit

History

RAPTOR_QUICK_REFERENCE.md

File metadata and controls

RAPTOR Quick Reference Cheat Sheet

⚡ Installation

🔥 Most Common Commands

Complete Workflow

📋 Module Quick Reference

Module 2: Quality Assessment

Module 3: Data Profiler

Module 4: Recommender

Module 7: DE Import

Module 8: Parameter Optimization

Module 9: Ensemble Analysis

🔧 Common Parameters

📊 Pipeline Quick Reference

🐍 Python API Quick Reference

Import Everything

Minimal Example

📁 File Formats

Count Matrix (counts.csv)

Metadata (metadata.csv)

Sample Sheet for Pipelines (samples.csv)

⚡ Quick Decision Flowchart

🆘 Common Errors & Fixes

Error: "Too few genes"

Error: "Sample mismatch"

Error: "Module not found"

📚 Get Help

🎯 Workflow Templates

Template 1: Quick QC Only

Template 2: QC + Profile

Template 3: Complete Analysis

Template 4: Ensemble Only

💡 Pro Tips