DocumentationDocumentationExamples
Examples Gallery
Explore real-world examples of AlignAIR in action. See input sequences, commands, and expected outputs for different use cases.
Example Categories
Basic Analysis
Single sequence processing
Batch Processing
Large dataset analysis
Custom Parameters
Optimized configurations
Advanced
Complex workflows
Example 1: Basic Heavy Chain Analysis
Single sequence V(D)J assignment with default parameters
📥 Input
Input File (sequences.csv):
sequence_id,sequence seq_001,CAGGTGCAGCTGGTGGAGTCTGGGGGAGGCTTGGTAAAGCCT... seq_002,GAGGTGCAGCTGGTGGAGTCTGGGGGAGGCTTGGTAAAGCCT...
Command:
python app.py run \ --model-checkpoint=/app/pretrained_models/IGH_S5F_576 \ --chain-type=heavy \ --sequences=/data/input/sequences.csv \ --save-path=/data/output/results
📤 Expected Output
Output File (results.csv):
sequence_id,v_call,d_call,j_call,productive seq_001,IGHV1-2*01,IGHD3-3*01,IGHJ4*01,True seq_002,IGHV1-3*01,IGHD2-2*01,IGHJ6*01,True
Processing Details
• 2 sequences processed
• Default thresholds used (V:0.75, D:0.3, J:0.8)
• Both sequences are productive
• Default thresholds used (V:0.75, D:0.3, J:0.8)
• Both sequences are productive
Example 2: Light Chain with Custom Thresholds
High-stringency analysis for clean data
📥 Setup
Use Case:
High-quality light chain sequences from flow-sorted B cells. Using stricter thresholds for precise allele calling.
Command:
python app.py run \ --model-checkpoint=/app/pretrained_models/IGL_S5F_576 \ --chain-type=light \ --sequences=/data/input/light_chains.csv \ --save-path=/data/output/light_results \ --v-allele-threshold=0.9 \ --j-allele-threshold=0.85 \ --airr-format
📊 Results Analysis
Threshold Impact:
V calls with default (0.75):1,850 calls
V calls with strict (0.9):1,650 calls
Higher confidence, fewer ambiguous calls
Output Format:
Full AIRR Schema with standardized column names for downstream analysis pipelines.
Example 3: Large Dataset Processing
Optimized parameters for 100K+ sequences
⚡ Performance Setup
Optimized Command:
python app.py run \ --model-checkpoint=/app/pretrained_models/IGH_S5F_576 \ --chain-type=heavy \ --sequences=/data/input/large_dataset.csv \ --save-path=/data/output/batch_results \ --batch-size=4096 \ --fix-orientation
Performance Tips:
Increased batch size to 4096
Enabled orientation fixing
GPU memory: 16GB+
📈 Benchmark Results
Processing Times:
100K sequences45 minutes
500K sequences3.5 hours
Resource Usage:
12GB
GPU Memory
95%
GPU Utilization
🚀 Quick Start Templates
Standard Heavy Chain
python app.py run \ --model-checkpoint=/app/pretrained_models/IGH_S5F_576 \ --chain-type=heavy \ --sequences=/data/input/sequences.csv \ --save-path=/data/output
High-Quality Light Chain
python app.py run \ --model-checkpoint=/app/pretrained_models/IGL_S5F_576 \ --chain-type=light \ --sequences=/data/input/light_chains.csv \ --save-path=/data/output \ --v-allele-threshold=0.85 \ --j-allele-threshold=0.9 \ --airr-format
Ready to Try These Examples?
Use these examples as starting points for your own AlignAIR analyses. Modify parameters based on your specific data and requirements.
© 2025 AlignAIR. All rights reserved.•Advancing computational biology through AI