🧬
βš—οΈ
πŸ”¬
πŸ§ͺ

|

The comprehensive benchmarking platform for antibody de novo sequencing.

|

Our platform provides the largest high-quality dataset to date, comprising 1,638,248 peptide-spectrum matches (PSMs) from 131 monoclonal antibodies (mAbs) across six species and 11 proteases, as well as eight mAbs with known sequences for evaluating full-length reconstruction. In addition, it offers an unbiased evaluation infrastructure that ensures fair comparisons. Through a unified training dataset, we benchmarked leading sequencing and assembly tools across six metric categories, spanning peptide sequencing (accuracy, robustness, efficiency, error types) and assembly (coverage depth, assembly score).

πŸ’‰
πŸ”¬

Antibody-specific Dataset

πŸ“‹

High-Quality PSM Dataset

Comprehensive dataset comprising 1,638,248 high-quality peptide-spectrum matches from 131 monoclonal antibodies across multiple species and proteases

131
mAbs
1.6M+
PSMs
Explore Dataset
πŸ”—

Full-Length Validation Set

Eight monoclonal antibodies with known sequences for evaluating full-length sequence reconstruction and assembly performance

8
Known mAbs
100%
Validated
View Sequences
🧬
βš—οΈ
πŸ”¬

Comprehensive Benchmarking Features

🎯

Accuracy

Comprehensive accuracy metrics including amino acid and peptide precision/recall, PTM identification accuracy, and AUC performance

β€’ AA Precision, AA Recall
β€’ Peptide Precision, Peptide Recall
β€’ PTM Precision, PTM Recall
β€’ AUC
View Metrics
⚑

Efficiency

Performance evaluation focusing on computational efficiency and resource utilization for practical deployment

β€’ Computing Time
β€’ Resources
View Metrics
πŸ›‘οΈ

Robustness

Stability assessment across different experimental conditions including varying peptide lengths and spectral quality

β€’ Peptide Length
β€’ Noise Peaks
β€’ Missing Fragmentation Ions
View Metrics
πŸ“Š

Coverage Depth

Sequence coverage analysis evaluating the depth and completeness of peptide identification

β€’ 7-mer
β€’ Peptide
View Metrics
πŸ”

Error Types

Detailed error analysis categorizing different types of prediction mistakes and their patterns

β€’ Inversion, AA Replace
β€’ More than 6 AAs Wrong
β€’ Others
View Metrics
🧩

Assembly

Full-length sequence reconstruction evaluation including assembly quality and structural integrity

β€’ Assembly Score
β€’ Coverage, Accuracy
β€’ Gap, Insertion
View Metrics

Research Applications

De Novo Sequencing

Advanced algorithms for peptide sequence prediction from tandem mass spectrometry data

Benchmarking

Comprehensive evaluation metrics and standardized datasets for model comparison

Ready to Benchmark Your Model?

Join researchers worldwide in advancing antibody de novo sequencing