Database search

Databases searches scientific archives from ProteinIQ and opens records with their identifiers, available properties, and downloads. Search one provider at a time, then check the source record before using its data in an analysis.

Choose a database

Select a provider that matches the information you need. The provider selector and introductory cards offer:

• RCSB PDB and PDBe: Macromolecular structure records.
• AlphaFold DB: Predicted protein structure records.
• UniProt: Protein sequences and annotations.
• PubChem and ChEBI: Small molecules, identifiers, and chemical properties.
• ChEMBL and BindingDB: Compound and bioactivity information.
• InterPro and QuickGO: Protein families, domains, and functional annotations.
• Rhea: Biochemical reactions.

The search field's placeholder and the Try examples change with the selected provider. Use these to see the kinds of names or identifiers the provider accepts.

Search for a record

To look up an example compound:

1. Open Databases and select PubChem.
2. Enter aspirin in the search field, or select that example under Try.
3. Select Search or press Enter.
4. Open a result to inspect its record.

Selecting a Try example fills the field; you still need to run the search. The result count describes the records returned by that search, not the size of the source database.

Use a specific accession or identifier when you already know the record you need. A name search can return several related records. Confirm the identity in the detail view instead of assuming the first match is the intended molecule or protein.

Inspect the source and contents

The record detail shows the database, record type, identifiers, and available properties. Depending on the source, it can also show sequences, structure links, or an interactive structure preview.

• Provider button: Open the original database record in another tab.
• Copy controls: Copy an identifier or other supported value.
• Download buttons: Open available representations, such as FASTA, PDB, CIF, SDF, or JSON. The choices depend on the record.

A missing preview does not necessarily mean the record has no downloadable structure. Check its download and source links. For a compound, distinguish a 2D representation from a 3D structure; for a protein, distinguish a predicted model from an experimental structure.

Use downloaded data in a tool

Searching or downloading a record does not automatically save it to your workspace library or start an analysis.

Download the representation accepted by the destination tool, then upload it to the tool's input field. To keep it for reuse, upload it to Files and use Send to tool from the file's actions menu. Review the assigned input and settings before submitting.

Some tools also offer a database identifier input directly. Use that method only when the tool provides it. A saved JSON record is not a substitute for a required sequence or coordinate file.

Keep the source database, accession, and downloaded file with your research records. The input preparation guide describes checks to make before analysis.

Search and download problems

If a search returns no records, check the selected provider, identifier spelling, and example format. A failed request is different from a successful search with no matches; read the displayed error before changing your query.

If a download is unavailable, open the original provider record and check its current files. Database availability and available representations can vary independently of the ProteinIQ search page.

Browse datasets opens the dataset catalog, a separate collection of datasets rather than another provider search.