omnilab/interpro

InterPro (EMBL-EBI), through its API (https://www.ebi.ac.uk/interpro/api/), for omnisense/omnilab: InterPro entries (IPR000971) and Pfam families (PF00042) - type, description, member signatures, GO terms, counts, clan, hierarchy -, and a UniProt protein's domains and families with their positions, as entities; a search. No key.

Installation

composer require omnilab/interpro

use Omnilab\InterPro\InterProSourceFactory;

$interpro = (new InterProSourceFactory($httpClient))->create([
    'mailto' => 'support@glitchr.io',  // optional
]);

Credentials

None: InterPro's API takes no key and no account.

Options

The factory's options, with the defaults it applies (InterProSourceFactory):

Capabilities

capabilities() answers entity, search; an operation it does not answer throws NotSupportedException before any call.

The schemes

An InterPro entry is named in the core's interpro scheme (IPR000971), a Pfam family in pfam (PF00042, its version dropped: PF00042.27); both are read from their shapes, their prefixes and their pages in InterPro (https://www.ebi.ac.uk/interpro/entry/InterPro/IPR000971/, .../entry/pfam/PF00042/). A protein is named by its UniProt accession (P68871, uniprot:P68871): an accession alone is read as one, none of the other schemes this source reads having its shape.

Which API

https://www.ebi.ac.uk/interpro/api/ (InterPro 110.0 on 2026-10-15, the InterPro-Version header):

Calls go over HTTP/1.1: InterPro answers "nothing found" with a 204 that carries a Content-Length, which curl refuses over HTTP/2 ("Invalid HTTP header field", met on 2026-10-15); over HTTP/1.1 it is read as nothing. An unknown entry, protein or search is that 204: null, an empty page. An identifier of a base InterPro does not read here (ChEBI, GO, NCBI Gene...) is refused before any call.

What an entity holds

An entry: id (interpro:IPR000971, pfam:PF00042), label (its name), synonyms (its short name when it differs), kind (its type: domain, family, homologous superfamily, repeat, conserved site, active site, binding site, ptm), description (InterPro's text, tags and reference marks [[cite:PUB...]] taken off; for a Pfam family without one, Wikipedia's extract InterPro gives), url, and:

related: member (the Pfam families an InterPro entry integrates), integrated (the InterPro entry a Pfam family is integrated into), go, structure (the representative PDB entry), overlap (the entries it overlaps), parent and child (InterPro's hierarchy: IPR002339, the pi hemoglobins, is a child of IPR002338, the alpha-type ones).

A protein: id (uniprot:P68871), label (its name), kind protein, description (UniProt's function, its evidence marks taken off), url (its page in InterPro), and entry_name (HBB_HUMAN), gene, organism, sequence_length (147 aa), reviewed, protein_existence, fragment, proteome, in_alphafold, structures - UniProt's keys and words where omnilab/uniprot has them, so that a merge sets both side by side -; one property per entry it matches, its key the entry's type and its value where it lies (domain: Globin (IPR000971): 3-147; family, homologous_superfamily...; pfam: Globin (PF00042): 27-142), InterPro's first; go_term (one each). related: by the entries' types - domain (interpro:IPR000971, pfam:PF00042), family, homologous_superfamily, repeat, conserved_site... -, go, organism. The other member databases (CDD, PANTHER, CATH-Gene3D, SUPERFAMILY, PRINTS, PROSITE) are left out: no scheme of the core names them.

The search

InterPro's text search over its entries, in InterPro's order: by accession, not ranked ("globin" gives Uteroglobin first, then Globin; 47 entries on 2026-10-15). An entity of a page holds its name, type, member signatures and GO terms: entity() reads it whole. A domain other than biology answers nothing.

The Collector: hemoglobin beta from UniProt

$collector->collect('uniprot', 'P68871', 1, ['interpro']);

Against the real services on 2026-10-15: three calls (UniProt's entry, InterPro's protein and its entries) in 3.3 to 4.1 s, sources = uniprot, interpro; both sequence lengths (147 aa [uniprot], 147 aa [interpro]), the domains, families and superfamilies related.

What the API does not give here

The member databases' own entries (CDD, PANTHER...), the proteins of an entry (protein/entry/...: tens of thousands), domain architectures, taxonomy and proteome breakdowns, structures and AlphaFold models of an entry, InterPro's sets (Pfam clans as entities), the HMMs and alignments, and the sequence's own features (InterPro's residue endpoints).

Example

<?php

require __DIR__.'/vendor/autoload.php';

use Omnilab\InterPro\InterProSourceFactory;
use Omnilab\Model\Scheme;
use Omnilab\Source\Query;
use Symfony\Component\HttpClient\HttpClient;

$interpro = (new InterProSourceFactory(HttpClient::create()))->create(['mailto' => 'you@example.org']);

$globin = $interpro->entity('IPR000971');
echo $globin->id->display(), ' ', $globin->label, ' (', $globin->kind, '): ', $globin->value('proteins'), ' proteins, member ', $globin->value('member'), "\n";

$hbb = $interpro->entity('uniprot:P68871');
echo $hbb->label, ': ', implode(', ', array_map('strval', $hbb->properties('domain'))), "\n";
echo '  families: ', implode(', ', $hbb->related('family')->values(Scheme::INTERPRO)), "\n";

$page = $interpro->search(new Query('globin', limit: 3));
printf("%d entries match \"globin\"; the first %d: %s\n", $page->total, count($page), implode(', ', array_map(static fn ($e) => $e->id->display(), $page->items)));

IPR000971 Globin (domain): 35483 proteins, member PS01033 Globin domain profile (PROSITE profiles)
Hemoglobin subunit beta: Globin (IPR000971): 3-147
  families: IPR002337, IPR050056
47 entries match "globin"; the first 3: IPR000329, IPR000971, IPR001032

(as answered on 2026-10-27: four calls; value('member') gives the first member, properties('member') every one)

# config/packages/omnilab.yaml
omnilab:
    sources:
        interpro: { factory: interpro, options: { mailto: '%env(OMNILAB_MAILTO)%' } }

Errors

Every call goes through HttpSource: a 404 or a 410 (and a 204) is null - an empty page for a list -; a 429 is a RateLimitedException (retryAfter from Retry-After); a 5xx, a time-out or a host that does not answer an UnavailableException; any other 4xx a ProviderException with its status and what the service said; an answer that is not JSON a ProviderException, a web page answered for JSON an UnavailableException.

InterPro's own: "nothing found" is a 204 (an unknown entry, protein or search): null, an empty page; the calls go over HTTP/1.1 so that curl reads it. Before any call: an identifier of a base InterPro does not read here (ChEBI, GO, NCBI Gene...) (NotSupportedException).

Verified, and not

Against the real service on 2026-10-15 (InterPro 110.0): the globin domain IPR000971 (35,483 proteins, 1,461 structures, its Pfam member, heme binding), the beta-type hemoglobins IPR002337, the pi hemoglobins IPR002339 under the alpha-type ones, Pfam's PF00042 (clan CL0090, integrated into IPR000971); hemoglobin beta P68871 and its six InterPro and Pfam matches; an unknown entry and an unknown protein (204); the search for "globin" in two pages of three; the Collector's walk above. The fixtures in Tests/Fixtures/ are its answers. Each call took 1 to 5 s.

Again on 2026-10-27: IPR000971, hemoglobin beta's matches, the search for "globin" (47 entries; the example above).

Not verified: a rate limit (none published, none met); a protein matching more than 200 entries (one page is read); the licence's text (its page is rendered by a script).

Responsibility

This package is an independent, unofficial client: it is not affiliated with, endorsed by or sponsored by InterPro or the EMBL-EBI. InterPro is free to use under the EMBL-EBI's terms of use, its licence stated on its site, and asks to be cited (Blum M. et al., Nucleic Acids Res. 2025, 10.1093/nar/gkae1082), as the README's Terms of use says. Whoever runs the package is responsible for complying with them and for any use made of what it returns: see the README's Responsibility section.