Getting started
A short, practical walkthrough of the African DSI DataBank as an actual researcher experiences it — browsing, searching, requesting access to a gated record, running a compute job across the Hub and its connected nodes, and downloading a result. If you're building against the API directly instead, see the API guide; unfamiliar terms are explained in the glossary.
1. Browse the species catalog
The homepageis a live, searchable catalog of every species this federated Hub currently holds data for — aggregated from every connected Sub-regional, National, and Spoke node across the continent, not just the Hub's own local holdings. Search by scientific name, or filter by genome type (whole assembly, organellar, virus/prokaryote, or individual sequence) using the form above the results. Each result card links to that species' own detail page.
2. Explore a species page
A species page (/species/[id]) is the single place everything this platform knows about that species comes together, in the order the data actually flows: local/ indigenous names, an embedded genome browser (when a real assembly file is available), sequence records (genomes, transcriptomes, proteins, proteomes), geographic range, chemical and metabolic profiles, predicted protein structures, and field/specimen and cellular/ molecular imaging — each section only appears when that species actually has that kind of data, so a page never shows an empty placeholder. Every citable record carries a real, durable citation id (AFDSI-...) with a one-click copy/BibTeX export — look for the small "Cite" chip next to a record.
3. Browse by data type
Beyond a single species' own page, every major data type also has its own dedicated, filterable listing: Genomes (with country/license/assembly- quality filters and a bulk CSV/Darwin Core Archive export), Protein Structures (an interactive 3D viewer per structure), and the rest of the Biomolecular Data and Imaging menus in the top navigation. The Data Dashboardgives a per-species matrix of exactly which data types the platform actually holds — useful for spotting real gaps at a glance before searching for something that isn't there yet.
4. Create an account
Every page above is fully browsable without an account — search, read, and explore freely. An account is only needed to request gated data, submit a compute job, or (if you operate a node) manage that node's own settings. Two separate applications exist, and it matters which one you want: Apply for an accountis for someone who already knows which node they'll be operating (or wants access to gated records); Register a new nodeis for standing up a brand-new node in the federation in the first place. If you're unsure which applies to you, start with Apply — its own form explains the distinction and cross-links to the other one.
5. Request access to a gated record
Most records are fully public. A record tagged regional_researcher or restrictedneeds an approved access request first — from that record's own page, or via Access requests, submit a request naming the specific record and your purpose. regional_researcher-tier requests may involve an automated identity check against a third-party ABS (Access and Benefit-Sharing) service; restricted-tier requests are always reviewed by hand and are never auto-approved. You can track the status of your own requests from the same page at any time.
6. Run a compute search
From Compute requests, you can run a real query — a similarity search, a phylogenetic tree build, or (for nodes that offer it) a specific containerized bioinformatics tool — against the Hub's own aggregated holdings and/or any number of real connected nodes at once, in a single submission. The form is fully explorable while logged out; only clicking Submit actually requires an account. If a target node has published its own participation terms (e.g. an attribution request), they appear right below that node once you select it — worth a read before submitting.
7. Download your results
Once every target in a submission finishes, you'll see each one's own outcome inline on the same page — matched hits, a real Newick tree, or a tool run's exit code and output. Where a real output file was produced, a "Download output" button fetches it directly. If you're logged in with a real email on file, you'll also get a short email the moment the whole submission finishes running, so there's no need to keep the tab open.
8. Get help
Something not working, or found a bug? Use the Contact Us form on the homepageand choose "Testing feedback / bug report" (or the category that best matches what you're reaching out about) — most submissions are automatically tracked as a real, referenceable ticket, and the confirmation email includes a direct link back to the same conversation thread, with no account needed to reply on it.