Commons

Commons is a catalogue of public research data — over a thousand datasets from archives such as DANDI, IDR and the Allen Institute — that open straight into a Viewer. Nothing is copied into your Drive: the data streams from the archive that holds it, and the licence and citation stay attached.

Browsing

Commons has five tabs along the top: Datasets, Papers, Atlases (reference volumes), Connectomes (EM reconstructions) and Assets (illustrations), next to a count of the datasets indexed. It opens on Datasets. All five open in place under the same header with the same search box, and what you typed carries over when you switch tab.

Beside the search box, Filters holds the current tab’s own filters, and each tab keeps its own choices. Datasets filter by Data type, Assay, Organism, Archive and Licence. Papers filter by Field, Journal or server, Paper type, Linked data, Access and Your library. Atlases filter by Species and Availability, connectomes by Species, and assets by Category, Source and Licence.

Datasets also has three ways to look at the results: List, Globe and Map. Switching view never changes what was asked for.

Assay appears once a data type is chosen. Filters you have set are listed under the search box, each with a cross to clear it, along with any year range or lab you have applied elsewhere. Long lists have a box to type into. The counts next to each option are live, so you can see what a filter would leave before you click it.

You can also just describe what you want. Type something like mouse ephys from dandi and, after a moment, the search box offers the filters those words describe (Electrophysiology, DANDI, Mus musculus). Select one to apply it, or Apply all and clear search to swap your words for the filters. Nothing is applied until you choose it.

On the Globe, a timeline curves under the sphere: a ribbon that is thicker in the years more data was released. Drag either end along the curve, or focus it and use the arrow keys, to narrow everything (globe, map and list) to those years; all years puts it back. Narrowing leaves out datasets with no release date, and the timeline says how many. The Map has the same ribbon, running straight.

Sources indexed today: DANDI, IDR, Allen Institute, ENCODE, GDC, RCSB PDB, AlphaFold, Cell Painting Gallery, BioImage Archive, Human Cell Atlas and E11 Bio.

A dataset

Open a dataset and you get its licence, version and access level at the top, a Cite block with the DOI and a one-click copy of the full citation, the abstract, the authors, and the organism, modality and assay, with tissue, disease and keywords one click away.

When Relay knows the paper a dataset was released with, the page names it under From the paper; select the title to open the paper. Many datasets have no paper of their own, and those pages simply leave the line out.

Below that is Assets: the files. Filter them with Opens in Relay, At source or All. Each row shows its size, an Open button, and an Ask Ray button. Files the archive does not serve in a form Relay can stream are marked At source with a link to the original.

An asset’s own page adds a Provenance card: dataset, path, size, checksum, whether the archive supports range requests, and a reminder that the data is streamed from the source, not mirrored by Relay. Archives that charge the reader for egress are badged Requester pays.

What you can do with an asset

Open — a Viewer window, exactly as for a file of your own. Annotate it, run a processing tool on it, capture it into Studio.
Ask Ray — attaches the asset to your next question, so you can ask summarise the recording metadata or plot the mean trace without opening it first. While Ray is already waiting for an attachment the button reads Add to Ray.
Original asset — the archive’s own page for the file.

Papers and labs

Papers is a feed of newly indexed papers and the datasets behind them. It has four views:

For you: papers that match your feed preferences. This is the one Papers opens on.
Following: papers from the labs and journals you follow.
My library: the papers in your own reference library, with the ones that released data in Commons first.
Everything: every paper, newest first.

Pick a paper to read its abstract and see its linked datasets; Open dataset takes you to that dataset, and the paper’s title opens the paper itself. Follow sits next to the paper’s lab, and the lab’s name opens its page.

Feed preferences

The first time you open Commons it asks what you work on: your fields and organisms, the journals and preprint servers you read, and the data types and archives you care about. Skip, show everything leaves the feed untuned.

Change your answers at any time from the gear on the Papers tab, or from Feed preferences in the ⋯ menu at the top of Commons. There you can also follow or mute keywords, and switch on Email me a weekly digest: one email a week with the new papers and datasets that match, and none when there is nothing new.

If For you comes up empty, add a field or a journal to your preferences, or select Show everything.

Journals

Select a journal’s name on any paper to open its page: its papers, newest first, beside the datasets released with them. Follow puts the journal in your Following view, Add to my journals adds it to your feed preferences, and All papers takes you back.

Your paper library

Sync library brings in the papers you already have: from Zotero, from a BibTeX or RIS file exported by any reference manager, or your own papers by ORCID iD. Commons then marks those papers in the feed, shows which of them released data you can open, and brings papers that cite them into For you. With Zotero connected, Save to Zotero on a paper sends it to your Zotero library.

Not every dataset has a paper. Some archives require one, many do not, and predicted structures such as AlphaFold’s never have their own, so a dataset without a linked paper is normal and not a gap.

Illustrations

Assets is a library of open-licence scientific drawings, such as a mitochondrion, a pipette, a mouse or a neuron, for figures, slides and posters. Search by name or narrow by Category. Drawings under Relay style share one look, can be recoloured, and are free for any use. Drawings under From open libraries each show the credit line to use.

Pick a drawing and you can:

Copy it, then paste it into a Studio slide or poster, or into another app.
Copy credit to put its credit line in your caption, when it needs one.
Open in SVG editor, which saves a copy to the Assets folder in your Drive and opens that copy for editing.
Download it as an SVG file.

Publishing your own dataset

If your account can publish, the ⋯ menu at the top of Commons has Publish a dataset. It takes three steps:

Files: drop in files, a folder or a zip. Folders are kept as they are and a zip is unpacked. Relay checks each file for anything that might identify a person, such as certain column names or DICOM headers, and asks you to Leave it out or confirm It is de-identified.
Describe: a title, a description, organism, data type, authors in citation order with their institutions, and optionally the DOI of the paper the data supports. Pick a licence: CC BY 4.0 (reuse with credit), CC0 (no conditions) or CC BY-NC 4.0 (credit, no commercial use).
Review: choose who can find it (Everyone, listed in Commons, Anyone with the link, or Listed after an embargo), confirm that the files identify no one and that you have the right to publish them, then select Publish to Commons.

Files are scanned for malware before a dataset goes public. If the scan is still running, publishing asks you to try again in a minute.

A published dataset gets its own page with a citation and is marked Community, which means it was published by its authors and not indexed from an archive. Any reader can use Report a problem on that page.

Your datasets, in the same menu, lists your drafts and what you have published. Withdraw hides a dataset’s files and takes it out of Commons; its page stays, marked withdrawn, with your reason.