goscripts - Python script and package for Gene Ontology enrichment analysis

The full documentation for this project is available at: https://pmoris.github.io/goscripts/

---

A ready to use python script to perform GO enrichment tests by inputting a list of uniprot_kb accession numbers, an ontology (.obo) file and a gene association (.gaf) file. No install required!

The goscripts package provides further functionality to parse and manipulate .obo and .gaf files; e.g.

• Parsing .obo gene ontology files in order to retrieve child/parent terms: goscripts.obo_tools
• Remapping gene ontology terms to a specified depth: goscripts.obo_tools
• Parse .gaf gene association files: goscripts.gaf_parser
• Performing an enrichment test using various multiple testing correction procedures (provided by statsmodels): goscripts.enrichment_stats

How to get started

Without installation

1. Download or clone the repository: git clone git@github.com:pmoris/goscripts.git

2. Run the script:

   python go_enrichment_script.py
   

The script requires functionality stored inside the goscripts directory and expects to find this directory. Consequently, if you wish to move the script to a different location, be sure to also copy this directory with it. Moreover, all :ref:`dependencies <dep-label>` should be installed.

Installing the goscripts package

1. Optionally: create a new virtual or conda environment: conda create -n goenv

2. Install the goscripts package:

   • Directly:

     pip install git+https://github.com/pmoris/goscripts.git
     

   • Manually: Download or clone the repository and from within the main directory (where setup.py resides), let pip install the package:

     git clone git@github.com:pmoris/goscripts.git
     cd goscripts/
     pip install . # don't forget the dot
     

3. You can now use the go_enrichment_script.py from any location:

   python go_enrichment_script.py
   

4. All additional goscripts functionality can be used in your own python scripts via an import statement:

   import goscripts
   # or
   from goscripts import gaf_parser
   

Downloading required ontologies and annotations

Example for human data:

# Ontology file
wget http://purl.obolibrary.org/obo/go.obo destination/directory
# Annotation file
wget ftp://ftp.ebi.ac.uk/pub/databases/GO/goa/HUMAN/goa_human.gaf.gz destination/directory
gunzip destination/directory/goa_human.gaf.gz

Using the GO enrichment test script

usage: go_enrichment_script.py [-h] [-b BACKGROUND] -s SUBSET -o OBO -g GAF
                            [-O OUTPUTFILE]
                            [-n {all,biological_process,molecular_function,cellular_component}]
                            [-m MINGENES] [-l TESTING_LIMIT] [-p THRESHOLD]
                            [--mult-test MULT_TEST] [-v] [--no-propagation]
                            [--no-part-of]

Script to perform GO enrichment analysis

optional arguments:
-h, --help            show this help message and exit
-b BACKGROUND, --background BACKGROUND
                        File containing a list of Uniprot accession numbers
                        for the background set of genes. If omitted, the full
                        list of genes in the .gaf file will be used. (default:
                        set())
-s SUBSET, --subset SUBSET
                        File containing a list of Uniprot accession numbers
                        for the subset of genes of interest (default: None)
-o OBO, --obo OBO     .obo file containing gene ontology (default: None)
-g GAF, --gaf GAF     .gaf file containing GO associations (default: None)
-O OUTPUTFILE, --output OUTPUTFILE
                        Output file or path (default: enrichment_results.csv)
-n {all,biological_process,molecular_function,cellular_component}, --namespace {all,biological_process,molecular_function,cellular_component}
                        Select the GO namespace to limit the enrichment test
                        to (default: all)
-m MINGENES, --min MINGENES
                        Minimum number of genes before considering a GO
                        category (default: 3)
-l TESTING_LIMIT, --limit-tests TESTING_LIMIT
                        P-value cut-off to use to stop GO tree propagation
                        during enchrichment tests (default: 0.05)
-p THRESHOLD, --pval-thresh THRESHOLD
                        Significant p-value threshold to use for significance
                        testing (default: 0.1)
--mult-test MULT_TEST
                        The type of multiple testing correction to use. Either
                        "fdr_bh" (default), "bonferroni" or any other method
                        offered by
                        statsmodels.stats.multitest.multipletests(). (default:
                        fdr_bh)
-v, --verbose         Verbose output. (default: False)
--no-propagation      Disables propagation during testing. Use if only
                        strictly associated terms should be tested. (default:
                        True)
--no-part-of          Ignore part_of relations between GO terms during
                        traversal. (default: False)

See the statsmodels documentation for an overview of all available multiple testing correction procedures: http://www.statsmodels.org/dev/_modules/statsmodels/stats/multitest.html.

Input files

• Ontology .obo files are described and available at the Gene Ontology Consortium.

• The gene association file format is described at the Gene Ontology Consortium and made available by EBI at the GOA ftp site.

• The background and subset files should be plain text files containing a single Uniprot accession number per line.

  P00750 A2BC19 P12345 A0A022YWF9

Details

Performs one-sided hypergeometric tests, starting from the most specific (child) GO terms associated with the genes in the set of interest. If the p-value of the test does not fall below the specified significance level alpha, the test will be carried out for all of the term's parent terms, otherwise the process will terminate. This method attempts to limit the total number of tests that need to be carried out, since a term that is enriched will likely also have enriched parent terms. Furthermore, GO terms associated with a small number of genes are skipped. Next, the Benjamini-Hochberg FDR or Bonferroni multiple testing correction are applied to the test results. Finally, a .csv file containing all the GO terms that were evaluated and their p-values are returned. More information is available in the docstrings.

---

Dependencies

numpy
pandas
scipy.stats
statsmodels.stats.multitest

---

Similar projects

• https://github.com/tanghaibao/goatools
• https://github.com/jdrudolph/goenrich
• https://www.psb.ugent.be/cbd/papers/BiNGO/Home.html

---

Copyright (c) 2018 Pieter Moris Adrem Data Lab - biomina - UAntwerpen