motif-scanner

A command-line tool for scanning DNA sequences and predicting transcription factor binding sites.

Features

• 🧬 Batch processing of sequence files
• 📊 PWM/EWM-based binding site analysis
• 🔍 Configurable occupancy threshold filtering
• 📈 Multiple output formats (CSV, Parquet)
• ⚡ Parallel processing for large datasets

Installation

From crates.io

cargo install motif-scanner

From Source

git clone https://github.com/peter6866/tf-binding-rs
cd tf-binding-rs
cargo install --path motif-scanner

Usage

Basic usage:

motif-scanner input.csv motifs.meme output.csv

With options:

motif-scanner input.csv motifs.meme output.parquet --cutoff 0.3 --mu 12

Arguments

• DATA_FILE: Input CSV file containing sequences (must have a 'sequence' column)
• PWM_FILE: MEME format file containing Position Weight Matrices
• OUTPUT_FILE: Path for output file (.csv or .parquet format)
• --cutoff: Minimum occupancy threshold (default: 0.2)
• --mu: Chemical potential parameter (default: 9)

Input Format

The input CSV file must contain a column named 'sequence' with DNA sequences:

id,sequence
seq1,ATCGATCGTGCTAGCTA
seq2,GCTAGCTAGCTAGCTAG

Output Format

The tool generates a table with the following columns:

• label: Sequence index from input file
• position: Position of the binding site
• motif: Name of the transcription factor
• strand: Binding strand (F/R)
• length: Length of the motif
• occupancy: Predicted occupancy score

Example

# Scan sequences with default parameters
motif-scanner sequences.csv pwm.meme results.csv

# Use stricter threshold and higher chemical potential
motif-scanner sequences.csv pwm.meme results.parquet --cutoff 0.4 --mu 15

# Process and save as Parquet format
motif-scanner data.csv motifs.meme output.parquet

Performance

The tool uses parallel processing for efficient scanning of large sequence datasets. Memory usage scales with the number of input sequences and motifs being scanned.