Loading data
On the homepage, pick single cell or single molecule, then choose how to load: in your browser, on our servers, or straight from a public bucket.
Which way should I load my data?
The three ways, side by side
| Preview in browser | Upload & process on server | Load from S3 | |
|---|---|---|---|
| Where it is processed | In your browser; nothing leaves your computer | On our servers, by a background worker | Not at all: read from the bucket as you look |
| Sign in | Not to look; yes to save it | Yes | Not to look; yes to add it to your account |
| Size | What your browser's memory allows (already-processed data: any size) | No limit from your browser | Any size |
| You end up with | A view now; Upload & Save gives a link that lasts | An email with a link when processing is done | A link that reopens it |
| Best for | A quick look; data that is already processed | Large raw datasets | Data you already host publicly |
Preview in browser
The file is parsed entirely in your browser and opens straight away. Nothing leaves your computer. It suits quick looks and datasets that fit in your browser's memory — roughly under 500,000 cells or tens of millions of molecules.
To keep it, press Upload & Save in the viewer (you need to be signed in). It is then in your account with a link that lasts.
Upload & process on server
For large datasets, upload the raw files and let a background worker process them. You need to be signed in, so we know where to email you.
- 1
Drop your file or folder
Switch the homepage to Upload & process on server and drop it as you would for a preview, then confirm the dataset name. Single cell data can also take a cell type CSV you made yourself, e.g. with MapMyCells.
- 2
Confirm the columns (molecules only)
A preview of the first rows, with a dropdown for each of gene, x, y, z and cell id, filled in from the detected platform. This is what makes sure your columns are read correctly.
- 3
Wait for the upload
A progress bar follows the upload. You can leave once it says Uploaded — we'll email you.
- 4
Get the email
The worker processes the data and emails you a link to the viewer. The dataset is already in your account.
Already-processed data opens as it is
If your data is already in a format the viewer reads, nothing is converted. Drop the folder on the single cell Folder card in Preview in browser and it opens at once, whatever its size: each part is read from disk a piece at a time, as you look at it. Cells, molecules and an image in the same folder are overlaid. Folder names don't matter — each part is recognised by what is inside it.
- 1CellsA SpatialData store, an AnnData table in Zarr (v2 or v3), or a merfisheyes cell folder
- 2+ MoleculesA merfisheyes molecule folder (manifest.json.gz + genes/)
- 3+ ImageAn OME-Zarr image, or the one inside the SpatialData store
- 4One overlaid viewMolecules over the cells, the image under both
Any part may be missing. Upload & Save from this preview uploads every part exactly as it is — nothing is processed — each as its own dataset in one project, and the saved link opens with the same overlays. Raw cells (H5AD, Xenium, MERSCOPE) beside processed parts are shown on their own here; use Upload & process on server to combine those.
Load from S3
Data already on a public bucket opens straight from it. Click Load from S3 on the homepage and paste one of:
- a merfisheyes cell or molecule folder;
- a SpatialData store (the
.zarrfolder); - an OME-Zarr image;
- the folder holding a store with its molecules beside it — it opens with everything overlaid.
Nothing is copied. A bucket's folders can't be listed from a browser, so a folder names its store in its mapping.json ("cells": "name.zarr"); without that, a store named after the folder (sample/sample.zarr) is tried.
The bucket must allow public reads and allow this site in its CORS settings (including the Range header, for stores written by recent anndata). A dataset opened this way that nobody owns yet shows Add to my account.
Very large datasets on your own machine
You can also process data yourself with the Python scripts in the GitHub repository and drop the output folder on the homepage, where it opens as it is:
python scripts/process_spatial_data.py data.h5ad output/ # cells python scripts/process_single_molecule.py data.parquet output/ # molecules
For an AnnData table, a molecule folder and an image you made yourself, see Preparing data: one script makes the folder ready.