proteorift 1.1.0

End-to-end machine learning pipeline for peptide database search

ProteoRift Python Package

End-to-end machine learning pipeline for peptide database search in mass spectrometry proteomics.

Installation

pip install proteorift

Quick Start

Using Sample Data

from proteorift import ProteoRiftSearch

# Initialize and run with sample data
searcher = ProteoRiftSearch()
results = searcher.search_with_sample_data()

print(f"Results saved to: {results['output_dir']}")

Using Your Own Data

from proteorift import ProteoRiftSearch

# Initialize search
searcher = ProteoRiftSearch()

# Run peptide database search
results = searcher.search(
    mgf_dir="path/to/your/spectra",      # Directory with MGF files
    peptide_db="path/to/your/database",  # Directory with FASTA files
    output_dir="./results"
)

Custom Parameters

searcher = ProteoRiftSearch(
    precursor_tolerance=10,
    precursor_tolerance_type="ppm",
    charge=3,
    length_filter=True,
    device="cuda"  # or "cpu", "auto"
)

results = searcher.search(mgf_dir="...", peptide_db="...")

Command Line Interface

# Run search with sample data
proteorift search-sample --output-dir ./results

# Run search with your data
proteorift search \
    --mgf-dir path/to/spectra \
    --peptide-db path/to/database \
    --output-dir ./results \
    --tolerance 10 \
    --charge 3

# Download models only
proteorift download-models

Output

ProteoRift generates Percolator-compatible PIN files:

• target.pin - Target peptide-spectrum matches
• decoy.pin - Decoy peptide-spectrum matches

System Requirements

• Python: 3.8 or higher
• GPU: 12GB+ VRAM recommended (CPU also supported)
• OS: Linux, macOS, or Windows