Overview

by Kozo Nishida and Alexander Pico

pywikipathways 0.0.2

WikiPathways is a well-known repository for biological pathways that provides unique tools to the research community for content creation, editing and utilization [1].

Python is a powerful programming language and environment for statistical and exploratory data analysis.

pywikipathways leverages the WikiPathways API to communicate between Python and WikiPathways, allowing any pathway to be queried, interrogated and downloaded in both data and image formats. Queries are typically performed based on “Xrefs”, standardized identifiers for genes, proteins and metabolites. Once you can identified a pathway, you can use the WPID (WikiPathways identifier) to make additional queries.

Prerequisites

All you need is this pywikipathways package! To install pywikipathways, run

!pip install pywikipathways

import pywikipathways as pwpw

Getting started

Lets first get oriented with what WikiPathways contains. For example, here’s how you check to see which species are currently supported by WikiPathways:

pwpw.list_organisms()

['Unspecified',
 'Acetobacterium woodii',
 'Anopheles gambiae',
 'Arabidopsis thaliana',
 'Bacillus subtilis',
 'Beta vulgaris',
 'Bos taurus',
 'Caenorhabditis elegans',
 'Canis familiaris',
 'Clostridium thermocellum',
 'Danio rerio',
 'Daphnia magna',
 'Daphnia pulex',
 'Drosophila melanogaster',
 'Escherichia coli',
 'Equus caballus',
 'Gallus gallus',
 'Glycine max',
 'Gibberella zeae',
 'Homo sapiens',
 'Hordeum vulgare',
 'Mus musculus',
 'Mycobacterium tuberculosis',
 'Oryza sativa',
 'Pan troglodytes',
 'Populus trichocarpa',
 'Rattus norvegicus',
 'Saccharomyces cerevisiae',
 'Solanum lycopersicum',
 'Sus scrofa',
 'Vitis vinifera',
 'Xenopus tropicalis',
 'Zea mays',
 'Plasmodium falciparum',
 'Brassica napus']

You should see 30 or more species listed. This list is useful for subsequent queries that take an organism argument, to avoid misspelling.

Next, let’s see how many pathways are available for Human:

hs_pathways = pwpw.list_pathways('Homo sapiens')

hs_pathways

	id	url	name	species	revision
0	WP100	https://www.wikipathways.org/index.php/Pathway...	Glutathione metabolism	Homo sapiens	107114
1	WP106	https://www.wikipathways.org/index.php/Pathway...	Alanine and aspartate metabolism	Homo sapiens	114258
2	WP107	https://www.wikipathways.org/index.php/Pathway...	Translation factors	Homo sapiens	117851
3	WP111	https://www.wikipathways.org/index.php/Pathway...	Electron transport chain: OXPHOS system in mit...	Homo sapiens	117097
4	WP117	https://www.wikipathways.org/index.php/Pathway...	GPCRs, other	Homo sapiens	117743
...	...	...	...	...	...
1327	WP734	https://www.wikipathways.org/index.php/Pathway...	Serotonin receptor 4/6/7 and NR3C signaling	Homo sapiens	117826
1328	WP75	https://www.wikipathways.org/index.php/Pathway...	Toll-like receptor signaling pathway	Homo sapiens	119233
1329	WP78	https://www.wikipathways.org/index.php/Pathway...	TCA cycle (aka Krebs or citric acid cycle)	Homo sapiens	119082
1330	WP80	https://www.wikipathways.org/index.php/Pathway...	Nucleotide GPCRs	Homo sapiens	111167
1331	WP98	https://www.wikipathways.org/index.php/Pathway...	Prostaglandin synthesis and regulation	Homo sapiens	117172

1332 rows × 5 columns

Yikes! That is a lot of information. Let’s break that down a bit:

help(pwpw.list_pathways)

Help on function list_pathways in module pywikipathways.list_pathways:

list_pathways(organism='')
    List Pathways

    Retrieve list of pathways per species, including WPID, name,
    species, URL and latest revision number.

    Args:
        organism (str): A particular species.

    Returns:
        pandas.DataFrame: A dataframe of pathway information.

    Examples:
        >>> list_pathways('Mus musculus')
            id  url     name    species revision
        0       WP1     https://www.wikipathways.org/index.php/Pathway...       Statin pathway  Mus musculus    117947
        1       WP10    https://www.wikipathways.org/index.php/Pathway...       IL-9 signaling pathway  Mus musculus    117067
        2       WP103   https://www.wikipathways.org/index.php/Pathway...       Cholesterol biosynthesis        Mus musculus    116834
        3       WP108   https://www.wikipathways.org/index.php/Pathway...       Selenium metabolism / selenoproteins    Mus musculus    117940
        4       WP113   https://www.wikipathways.org/index.php/Pathway...       TGF-beta signaling pathway      Mus musculus    116497
        ...     ...     ...     ...     ...     ...
        230     WP79    https://www.wikipathways.org/index.php/Pathway...       Tryptophan metabolism   Mus musculus    104913
        231     WP85    https://www.wikipathways.org/index.php/Pathway...       Focal adhesion  Mus musculus    116710
        232     WP87    https://www.wikipathways.org/index.php/Pathway...       Nucleotide metabolism   Mus musculus    116529
        233     WP88    https://www.wikipathways.org/index.php/Pathway...       Toll-like receptor signaling    Mus musculus    116521
        234     WP93    https://www.wikipathways.org/index.php/Pathway...       IL-4 signaling pathway  Mus musculus    117991
        235 rows × 5 columns

hs_pathways.shape

(1332, 5)

Ok. The help docs tell us that for each Human pathway we are getting a lot of information. A pandas.DataFrame.shape might be all you really want to know. Or if you’re interested in just one particular piece of information, check out these functions:

help(pwpw.list_pathway_ids)

Help on function list_pathway_ids in module pywikipathways.list_pathways:

list_pathway_ids(organism='')
    List Pathway WPIDs

    Retrieve list of pathway WPIDs per species.
    Basically returns a subset of list_pathways result.

    Args:
        organism (str): A particular species.

    Returns:
        pandas.Series: A series of WPIDs.

    Examples:
        >>> list_pathway_ids('Mus musculus')
        0        WP1
        1       WP10
        2      WP103
        3      WP108
        4      WP113
               ...
        230     WP79
        231     WP85
        232     WP87
        233     WP88
        234     WP93
        Name: id, Length: 235, dtype: object

help(pwpw.list_pathway_names)

Help on function list_pathway_names in module pywikipathways.list_pathways:

list_pathway_names(organism='')
    List Pathway Names

    Retrieve list of pathway names per species.
    Basically returns a subset of list_pathways result.

    Args:
        organism (str): A particular species.

    Returns:
        pandas.Series: A series of names.

    Examples:
        >>> list_pathway_names('Mus musculus')
        0                            Statin pathway
        1                    IL-9 signaling pathway
        2                  Cholesterol biosynthesis
        3      Selenium metabolism / selenoproteins
        4                TGF-beta signaling pathway
                               ...
        230                   Tryptophan metabolism
        231                          Focal adhesion
        232                   Nucleotide metabolism
        233            Toll-like receptor signaling
        234                  IL-4 signaling pathway
        Name: name, Length: 235, dtype: object

help(pwpw.list_pathway_urls)

Help on function list_pathway_urls in module pywikipathways.list_pathways:

list_pathway_urls(organism='')
    List Pathway URLs

    Retrieve list of pathway URLs per species.
    Basically returns a subset of list_pathways result.

    Args:
        organism (str): A particular species.

    Returns:
        pandas.Series: A series of URLs.

    Examples:
        >>> list_pathway_urls('Mus musculus')
        0      https://www.wikipathways.org/index.php/Pathway...
        1      https://www.wikipathways.org/index.php/Pathway...
        2      https://www.wikipathways.org/index.php/Pathway...
        3      https://www.wikipathways.org/index.php/Pathway...
        4      https://www.wikipathways.org/index.php/Pathway...
                                     ...
        230    https://www.wikipathways.org/index.php/Pathway...
        231    https://www.wikipathways.org/index.php/Pathway...
        232    https://www.wikipathways.org/index.php/Pathway...
        233    https://www.wikipathways.org/index.php/Pathway...
        234    https://www.wikipathways.org/index.php/Pathway...
        Name: url, Length: 235, dtype: object

These return simple lists containing just a particular piece of information for each pathway result.

Finally, there’s another way to find pathways of interest: by Xref. An Xref is simply a standardized identifier form an official source. WikiPathways relies on BridgeDb [2] to provide dozens of Xref sources for genes, proteins and metabolites. See the full list at https://github.com/bridgedb/datasources/blob/main/datasources.tsv

With pywikipathways, the approach is simple. Take a supported identifier for a molecule of interest, e.g., an official gene symbol from HGNC, “TNF” and check the system code for the datasource, e.g., HGNC = H (this comes from the second column in the datasources.txt table linked to above), and then form your query:

tnf_pathways = pwpw.find_pathways_by_xref('TNF','H')

tnf_pathways

	score	id	url	name	species	revision
0	0.38982576	WP5071	https://www.wikipathways.org/index.php/Pathway...	PPAR-gamma pathway	Homo sapiens	116510
1	0.38982576	WP5073	https://www.wikipathways.org/index.php/Pathway...	PPAR Beta/Delta pathway	Homo sapiens	115726
2	0.38982576	WP5115	https://www.wikipathways.org/index.php/Pathway...	Network map of SARS-CoV-2 signaling pathway	Homo sapiens	119638
3	0.27564844	WP176	https://www.wikipathways.org/index.php/Pathway...	Folate metabolism	Homo sapiens	118404
4	0.27564844	WP2328	https://www.wikipathways.org/index.php/Pathway...	Allograft Rejection	Homo sapiens	106557
...	...	...	...	...	...	...
80	0.27564844	WP5088	https://www.wikipathways.org/index.php/Pathway...	Prostaglandin signaling	Homo sapiens	119639
81	0.27564844	WP5093	https://www.wikipathways.org/index.php/Pathway...	Opioid receptor pathway annotation	Homo sapiens	119684
82	0.27564844	WP5094	https://www.wikipathways.org/index.php/Pathway...	Orexin receptor pathway	Homo sapiens	119637
83	0.27564844	WP5098	https://www.wikipathways.org/index.php/Pathway...	T-cell activation SARS-CoV-2	Homo sapiens	119538
84	0.27564844	WP2513	https://www.wikipathways.org/index.php/Pathway...	Nanoparticle triggered regulated necrosis	Homo sapiens	119820

85 rows × 6 columns

Ack! That’s a lot of information. We provide not only the pathway information, but also the search result score in case you want to rank results, etc. Again, if all you’re interested in is WPIDs, names or URLs, then there are these handy alternatives that will just return simple lists:

help(pwpw.find_pathway_ids_by_xref)

Help on function find_pathway_ids_by_xref in module pywikipathways.find_pathways_by_xref:

find_pathway_ids_by_xref(identifier, system_code)
    Find Pathway WPIDs By Xref

    Retrieve list of pathway WPIDs containing the query Xref by
    identifier and system code.

    Note:
        there will be multiple listings of the same pathway if the
        Xref is present mutiple times.

    Args:
        identifier (str): The official ID specified by a data source or system
        system_code (str): The BridgeDb code associated with the data source or system,
            e.g., En (Ensembl), L (Entrez), Ch (HMDB), etc. See column two of
            https://github.com/bridgedb/datasources/blob/main/datasources.tsv.

    Returns:
        pandas.Series: A series of WPIDs.

    Examples:
        >>> find_pathway_ids_by_xref('ENSG00000232810','En')
        0     WP2813
        1     WP4341
        2     WP4673
        3     WP1584
        4     WP2571
               ...
        82    WP5055
        83    WP5093
        84    WP5115
        85    WP5094
        86    WP5098
        Name: id, Length: 87, dtype: object

help(pwpw.find_pathway_names_by_xref)

Help on function find_pathway_names_by_xref in module pywikipathways.find_pathways_by_xref:

find_pathway_names_by_xref(identifier, system_code)
    Find Pathway Names By Xref

    Retrieve list of pathway names containing the query Xref by
    identifier and system code.

    Note:
        there will be multiple listings of the same pathway if the
        Xref is present mutiple times.

    Args:
        identifier (str): The official ID specified by a data source or system
        system_code (str): The BridgeDb code associated with the data source or system,
            e.g., En (Ensembl), L (Entrez), Ch (HMDB), etc. See column two of
            https://github.com/bridgedb/datasources/blob/main/datasources.tsv.

    Returns:
        pandas.Series: A series of names.

    Examples:
        >>> find_pathway_names_by_xref('ENSG00000232810','En')
        0     Mammary gland development pathway - Embryonic ...
        1       Non-genomic actions of 1,25 dihydroxyvitamin D3
        2                                      Male infertility
        3                             Type II diabetes mellitus
        4                     Polycystic kidney disease pathway
                                    ...
        82                                   Burn wound healing
        83                   Opioid receptor pathway annotation
        84          Network map of SARS-CoV-2 signaling pathway
        85                              Orexin receptor pathway
        86                         T-cell activation SARS-CoV-2
        Name: name, Length: 87, dtype: object

help(pwpw.find_pathway_urls_by_xref)

Help on function find_pathway_urls_by_xref in module pywikipathways.find_pathways_by_xref:

find_pathway_urls_by_xref(identifier, system_code)
    Find Pathway URLs By Xref

    Retrieve list of pathway URLs containing the query Xref by
    identifier and system code.

    Note:
        there will be multiple listings of the same pathway if the
        Xref is present mutiple times.

    Args:
        identifier (str): The official ID specified by a data source or system
        system_code (str): The BridgeDb code associated with the data source or system,
            e.g., En (Ensembl), L (Entrez), Ch (HMDB), etc. See column two of
            https://github.com/bridgedb/datasources/blob/main/datasources.tsv.

    Returns:
        pandas.Series: A series of URLs.

    Examples:
        >>> find_pathway_urls_by_xref('ENSG00000232810','En')
        0     https://www.wikipathways.org/index.php/Pathway...
        1     https://www.wikipathways.org/index.php/Pathway...
        2     https://www.wikipathways.org/index.php/Pathway...
        3     https://www.wikipathways.org/index.php/Pathway...
        4     https://www.wikipathways.org/index.php/Pathway...
                                    ...
        82    https://www.wikipathways.org/index.php/Pathway...
        83    https://www.wikipathways.org/index.php/Pathway...
        84    https://www.wikipathways.org/index.php/Pathway...
        85    https://www.wikipathways.org/index.php/Pathway...
        86    https://www.wikipathways.org/index.php/Pathway...
        Name: url, Length: 87, dtype: object

Be aware: a simple len function may be misleading here since a given pathway will be listed multiple times if the Xref is present mutiple times.

My favorite pathways

At this point, we should have one or more pathways identified from the queries above. Let’s assume we identified ‘WP554’, the Ace Inhibitor Pathway (https://wikipathways.org/instance/WP554). We will use its WPID (WP554) in subsequent queries.

First off, we can get information about the pathway (if we didn’t already collect it above):

pwpw.get_pathway_info('WP554')

{'id': 'WP554',
 'url': 'https://www.wikipathways.org/index.php/Pathway:WP554',
 'name': 'ACE inhibitor pathway',
 'species': 'Homo sapiens',
 'revision': '118788'}

Next, we can get all the Xrefs contained in the pathway, mapped to a datasource of our choice. How convenient! We use the same system codes as described above. So, for example, if we want all the genes listed as Entrez Genes from this pathway:

pwpw.get_xref_list('WP554','L')

['10159',
 '1215',
 '1511',
 '1585',
 '1636',
 '183',
 '185',
 '186',
 '3827',
 '4142',
 '4306',
 '4846',
 '59272',
 '5972',
 '623',
 '624',
 '7040']

Alternatively, if we want them listed as Ensembl IDs instead, then…

pwpw.get_xref_list('WP554', 'En')

['ENSG00000092009',
 'ENSG00000100448',
 'ENSG00000100739',
 'ENSG00000105329',
 'ENSG00000113889',
 'ENSG00000130234',
 'ENSG00000130368',
 'ENSG00000135744',
 'ENSG00000143839',
 'ENSG00000144891',
 'ENSG00000151623',
 'ENSG00000159640',
 'ENSG00000164867',
 'ENSG00000168398',
 'ENSG00000179142',
 'ENSG00000180772',
 'ENSG00000182220']

And, if we want the metabolites, drugs and other small molecules associated with the pathways, then we’d simply provide the system code of a chemical database, e.g., Ch (HMBD), Ce (ChEBI) or Cs (Chemspider):

pwpw.get_xref_list('WP554', 'Ch')

['HMDB0000016',
 'HMDB0000037',
 'HMDB0000464',
 'HMDB0001035',
 'HMDB00016',
 'HMDB00037',
 'HMDB0004246',
 'HMDB00464',
 'HMDB0061196',
 'HMDB01035',
 'HMDB04246',
 'HMDB61196']

pwpw.get_xref_list('WP554', 'Ce')

['16973',
 '2718',
 '2719',
 '27584',
 '29108',
 '3165',
 '35457',
 '55438',
 '80128',
 '80129',
 'CHEBI:16973',
 'CHEBI:2718',
 'CHEBI:2719',
 'CHEBI:27584',
 'CHEBI:29108',
 'CHEBI:3165',
 'CHEBI:35457',
 'CHEBI:55438',
 'CHEBI:80128',
 'CHEBI:80129']

pwpw.get_xref_list('WP554', 'Cs')

['102770', '110354', '150504', '23150106', '24774738', '266', '388341', '5932']

It’s that easy!

Give me more

We also provide methods for retrieving pathways as data files and as images. The native file format for WikiPathways is GPML, a custom XML specification. You can retrieve this format by…

gpml = pwpw.get_pathway('WP554')

gpml[:1000]

'<?xml version="1.0" encoding="UTF-8"?>\n<Pathway xmlns="http://pathvisio.org/GPML/2013a" Name="ACE inhibitor pathway" Organism="Homo sapiens">\n  <Comment Source="WikiPathways-description">The core of this pathway was elucidated over a century ago and involves the conversion of angiotensinogen to angiotensin I (Ang I) by renin, its subsequent conversion to angiotensin II (Ang II) by angiotensin converting enzyme. Ang II activates the angiotensin II receptor type 1 to induce aldosterone synthesis, increasing water and salt resorption and potassium excretion in the kidney and increasing blood pressure.\n\nSource: PharmGKB (https://www.pharmgkb.org/pathway/PA2023)\n\nProteins on this pathway have targeted assays available via the [https://assays.cancer.gov/available_assays?wp_id=WP554 CPTAC Assay Portal]</Comment>\n  <BiopaxRef>b93</BiopaxRef>\n  <Graphics BoardWidth="991.0" BoardHeight="651.0"/>\n  <DataNode TextLabel="NOS3" GraphId="ab119" Type="GeneProduct">\n    <Attribute Key="org.pathvisio.mo'

WikiPathways also provides a monthly data release archived at http://data.wikipathways.org. The archive includes GPML, GMT and SVG collections by organism and timestamped. There’s a Python function for grabbing files from the archive…

pwpw.download_pathway_archive()

organism argument is not specified. Open http://data.wikipathways.org/current/gpml with your web browser and specify the organism.

This will simply print the archive URL so you can look around (in case you don’t know what you are looking for). By default, it prints the latest collection of GPML files. However, if you provide an organism, then it will download that file to your current working directory or specified destpath. For example, here’s how you’d get the latest GMT file for mouse:

pwpw.download_pathway_archive(organism="Mus musculus", format="gmt")

'wikipathways-20210810-gmt-Mus_musculus.gmt'

And if you might want to specify an archive date so that you can easily share and reproduce your script at any time in the future and get the same result. Remember, new pathways are being added to WikiPathways every month and existing pathways are improved continuously!

pwpw.download_pathway_archive(date="20171010", organism="Mus musculus", format="gmt")

'wikipathways-20171010-gmt-Mus_musculus.gmt'

References

1. Pico AR, Kelder T, Iersel MP van, Hanspers K, Conklin BR, Evelo C: WikiPathways: Pathway editing for the people. PLoS Biol 2008, 6:e184+.

2. Iersel M van, Pico A, Kelder T, Gao J, Ho I, Hanspers K, Conklin B, Evelo C: The BridgeDb framework: Standardized access to gene, protein and metabolite identifier mapping services. BMC Bioinformatics 2010, 11:5+.