gaussf-pipeline 0.1.0

Pipeline for transcript abundance estimation using Gaussian smoothing of k-mer frequencies.

GaussF Pipeline

A collection of scripts to perform k-mer counting, normalization, and Gaussian fitting for transcript abundance estimation.

Description

This pipeline consists of four main steps:

1. Index Generation: Creates k-mer reference files from a transcript FASTA file (kmer_freq_dist.py).
2. K-mer Counting: Counts k-mers from FASTQ sequencing data based on the reference index (kmer_counter.py).
3. Merging & Normalization: Merges counts back to the reference and calculates normalized counts (merge_normalize.py).
4. Abundance Estimation: Applies Gaussian fitting to estimate transcript abundance (gaussf_tpm.py).

Data sample

Test Data (test_data/)

To facilitate testing and provide concrete examples, the test_data/ directory includes a small reference and sequencing dataset:

• Reference Genome Subset:

  • Homo_sapiens.GRCh38_first_200_transcripts.fa: This FASTA file provides a small slice of the human transcriptome (first 200 transcripts, GRCh38 assembly). It serves as the reference sequence set for operations like k-mer generation or indexing within test routines.

• Sample Sequencing Data:

  • 99272-N_extracted_100_raw_2.fastq.gz: This compressed FASTQ file contains a small subset (100 reads) of raw sequencing data. It represents typical input for sequence processing workflows tested by this project.

These files allow developers and users to run basic tests or demonstrations without requiring large external data downloads. They are primarily used by [mention specific test scripts or functionalities if applicable, e.g., "the unit tests in the tests/ directory"].

Installation

# Clone the repository (or download)
git clone https://github.com/QiangSu/gaussf_pipeline.git
cd gaussf_pipeline

# Install using pip (editable mode recommended for development)
pip install -e .

# 1. Generate reference index
# (Command will likely change to something like 'gaussf_index_gen --input ...')
python src/gaussf_pipeline/kmer_freq_dist.py --input_fasta /path/to/transcriptome.fa --output_dir /path/to/index_output --kmer_length 50 --threshold 3000

# 2. Count k-mers from FASTQ
# (Command will likely change to something like 'gaussf_kmer_count --fastq_path ...')
python src/gaussf_pipeline/kmer_counter.py --fastq_path /path/to/reads.fastq.gz --num_threads 30 --chunk_size 100000 --csv_input_dir /path/to/index_output --csv_output_dir /path/to/counts_output

# 3. Merge and Normalize
# (Command will likely change to something like 'gaussf_merge --kmer_reference_directory ...')
python src/gaussf_pipeline/merge_normalize.py --kmer_reference_directory /path/to/index_output --kmer_counts_directory /path/to/counts_output --output_directory /path/to/merged_output --read_length 150 --k 50

# 4. Estimate Abundance
# (Command will likely change to something like 'gaussf_abundance --input ...')
python src/gaussf_pipeline/gaussf_tpm.py --threshold 10 --input /path/to/merged_output --output /path/to/results.csv