genemapper 1.0.4

A command-line tool for retrieving GenBank metadata and mapping coding sequences between query and subject FASTA files to identify their most similar counterparts.

A command-line tool for retrieving GenBank metadata and mapping coding sequences between query and subject FASTA files to identify their most similar counterparts.

• Free software: MIT license

• Documentation: https://genemapper.readthedocs.io

Overview

GeneMapper provides two main functionalities:

1. Metadata Retrieval: - Extracts an accession number from a FASTA header and fetches the corresponding GenBank record from NCBI. - Prints out relevant details such as organism name, record description, and publication date.

2. Mapping Sequences: - Reads a query FASTA and a subject FASTA, each containing coding sequences (CDS). - Uses pairwise alignment to score each query sequence against possible subject matches. - Outputs an Excel file listing best matches, tie scenarios, unmatched sequences, and more.

Features

• Automated retrieval of GenBank metadata for query and reference sequences files in FASTA format.

• Parallelized mapping of query CDS against subject CDS with scoring and threshold-based similarity.

• Detailed log and Excel output of mapping results

• Adjustable similarity thresholds and multi-processor usage.

Installation

pip install genemapper

If installing from source, navigate to the package directory and run:

pip install .

Usage

Below are the two primary subcommands: metadata and map.

Usage - Metadata Retrieval

Fetch GenBank metadata for query/reference FASTA files. The FASTA headers must contain an accession in the form “>lcl|ACCESSION_cds_…” for automatic extraction.

genemapper metadata --query my_query_gene_cds.fasta \
                    --reference my_subject_gene_cds.fasta \
                    --email [e-mail]

What this does:

• Parses your query FASTA to extract an accession.

• Fetches the corresponding GenBank record from NCBI, printing details like: organism name, record date, and a brief description.

• Repeats for the reference FASTA.

Usage - Gene Mapping

Compare a query FASTA to a subject FASTA, finding the best-matching CDS:

genemapper map --query my_query_gene_cds.fasta \
               --subject my_subject_gene_cds.fasta \
               --output results.xlsx \
               --similarityThreshold 0.99 \
               --numProcesses 20

What this does:

• Parses both FASTA files, extracting locus tags and sequences.

• Aligns each query sequence to all subject sequences using a global aligner.

• Filters results by the similarity threshold.

• Saves everything in an Excel file (“results.xlsx”).

How It Works

1. parse_accession_from_fasta Looks at the first record in a FASTA, uses a regular expression to detect the accession in the header.

2. fetch_genbank_record Contacts NCBI using Biopython’s “Entrez.efetch” to download the GenBank record. Requires a valid email to respect NCBI’s policy.

3. parseFastaFile Reads all CDS in a given FASTA, extracting a “locus_tag” (and optional “[location=…]”) to store each sequence in a dictionary.

4. alignAndScore Performs a global alignment with Biopython’s PairwiseAligner. Normalizes the score by the length of the longer sequence, returning a floating-point similarity value.

5. findBestMatchesWithLocation Loops over all subject sequences, keeps track of the highest-scoring match, and flags ties if multiple subjects share the same top score.

6. parallelGenomeMappingWithLocation Uses Python’s multiprocessing to handle each query locus tag in parallel, which speeds up large genome comparisons. Writes results to Excel when finished.

Logging

GeneMapper writes runtime events and warnings to a log file called “genome_mapping.log”. This includes:

• Metadata retrieval failures (e.g., no valid accession).

• Mapping details, warnings about no matches above threshold, ties, or I/O errors.

• Summary of how many query genes matched, remained unmatched, etc.

Check this log file if something goes wrong or for deeper insight into the script’s operations.

Troubleshooting & Tips

1. No Accession Found Make sure your FASTA headers follow the format “>lcl|ACCESSION_cds_…”.

2. No Matches Above Threshold Lower the “–similarityThreshold” or check for sequence divergence. Review the log to see actual scores.

3. Parallel Performance If you have many CPU cores, increase “–numProcesses”. NOTE: by deafault all cores are used. If you hit resource limits, please reduce it by setting a suitable value.

4. Spreadsheet Issues The script writes an Excel file via pandas and openpyxl. If needed, you can adapt the code to write CSV by modifying “mappingDf.to_csv(…)”.

5. Versions - Biopython ≥ 1.78 is required for “PairwiseAligner”. - Python 3.7+ is recommended. - Openpyxl 3.1.5 (pinned) is required for the creation of the excel file - Pandas > 2.0 is recommended.

License

MIT License: See the licence file for full details.