PROBETHEUS

A Python package to detect overrepresented sequences in a fastq file and design probes against them. Designed for single read sequencing from immunoprecipitation experiments, riboSeq, and other single read sequencing experiments.

Installation

pip install probetheus

Features

• Process single-end FASTQ files to find top represented sequences
• Generate probes from top sequences with customizable lengths
• Cluster sequences based on edit distance
• Detect probe binding sites against reference sequences
• Generate cumulative percentage plots with elbow point detection
• Reanalyze results with custom elbow points
• Subsample input files for faster analysis or testing

Usage

Processing FASTQ Files and Generating Probes

# Basic usage
probetheus process input.fastq.gz -o results.tsv

# With core sequence analysis
probetheus process input.fastq.gz -o results.tsv --find_core --core_length 25

# Process without length filtering
probetheus process input.fastq.gz -o results.tsv --skip_length_filter

# Check probe binding against reference
probetheus process input.fastq.gz -o results.tsv -r ref.fasta --max_binding_dist 2

# Process with subsampling (e.g., use 20% of reads)
probetheus process *.fastq.gz -o results.tsv --subsample 20

Reanalyzing Results

After initial processing, you can reanalyze the results with either a different elbow point or by specifying the number of probes:

# Reanalyze with a new elbow point
probetheus reanalyze --input results.tsv --elbow 5 --output-prefix new_results

# Reanalyze by specifying number of probes
probetheus reanalyze --input results.tsv --probes 10 --output-prefix new_results

This will create:

• new_results_reanalyzed.tsv: New results file with selected sequences
• new_results_reanalyzed_cumulative.png: Updated cumulative plot
• new_results_reanalyzed_probes.fasta: New probe sequences

Arguments

Process Command

• --output, -o: Output table file
• --min-length: Minimum sequence length (default: 20)
• --max-length: Maximum sequence length (default: 50)
• --top-n: Number of top sequences to use for probe generation (default: 20)
• --probe-length: Length of generated probes (default: 25)
• --min-probe-length: Minimum acceptable probe length (default: 20)
• --edit-distance: Maximum edit distance for clustering (default: 1)
• --find-core: Find core sequences
• --core-length: Length for core sequence analysis (default: 25)
• --min-core-occurrence: Minimum fraction of sequences a core must appear in (default: 0.1)
• --reference, -r: Reference FASTA file to check probe binding
• --max-binding-dist: Maximum edit distance allowed for probe binding (default: 2)
• --subsample: Subsample percentage (1-100) of reads from each file
• --reads: Number of reads to take from each file (overrides --subsample if provided)
• --cpus: Number of CPU cores to use (default: 8, max: number of cores - 1)

Reanalyze Command

• --input, -i: Input results.tsv file from previous analysis
• --elbow, -e: New elbow point (number of sequences to keep)
• --probes, -p: Number of probes to generate (alternative to --elbow)
• --output-prefix, -o: Prefix for output files (optional)

License

This project is licensed under the MIT License - see the LICENSE file for details.