How to use VectaBind

Enter your API key

On the target bar (row below the logo), use the combined API connected · API key control. Click the status side to reconnect; click API key to paste your key (vb_...).

Free tier includes 500 scores per month and access to all embedded targets on the demo server.

Pick a target

Select a protein from the target bar — the scrollable pill list under the header (e.g. EGFR, HER2, MET). Use All indications to filter by disease area and Search targets to find a gene. The 3D pocket updates when you switch targets.

Import a library (CSV / TSV)

1. Click Template CSV to download vectabind_library_template.csv (columns: id, name, series, smiles, notes).
2. Click Upload CSV and select your file. smiles is required; id, name, series, and notes are optional.
3. Review the validation summary — e.g. 42 valid · 3 invalid SMILES — before saving.
4. Optionally name the library, then click Save. Libraries persist in your browser (localStorage; IndexedDB for 500+ compounds).

Filter targets by indication

Use the Indication dropdown in the target bar (Lung cancer, Breast cancer, AML / Leukemia, etc.) to narrow the target list. This uses the same disease groupings as the API GET /targets catalog.

Batch score with progress

1. Select a saved library from the dropdown (or paste SMILES in the textarea below). Generated libraries auto-select their target in the bar above.
2. For CSV imports, pick your target in the target bar manually.
3. Click Run batch score · TARGET in the library panel (or Run batch score below the SMILES textarea).
4. A progress bar shows chunk progress — batch size follows your API tier (shown in the usage meter: 42 / 500 scores · batch 10).

Multi-target selectivity panel

For HTS triage and off-target checks, screen the same library against every target in an indication at once:

1. Set Indication to a disease panel (e.g. AML / Leukemia / Lymphoma → FLT3, JAK2, BTK, …).
2. Load or save your compound library.
3. Click Screen indication panel — you'll see an estimated score count before it runs.
4. Results appear as a heatmap (compounds × targets) with pKd color coding and a Δ column (top target minus second-best — higher = more selective).
5. Click any cell to open full compound analysis for that target. Use Export panel CSV for Slack / Excel.

Results table, filters, and SAR view

Library runs show a sortable table: ID, series, MPO score, pKd, Δ vs reference, bind %, confidence tier, clinical analog flag, QED, and Lipinski.

• MPO sliders — weight pKd, QED, and Lipinski; sort by composite MPO
• Series filter / group — filter or group by R-series from your CSV
• Reference compound — set a SMILES standard; table shows ΔpKd vs reference
• Compare libraries — overlap and unique compounds vs a saved library
• Drop invalid — hide rows that failed scoring (on by default)
• Strong only (pKd ≥7) — keep high-potency hits
• Drug-like — keep compounds passing QED / Lipinski heuristics

After scoring, ChEMBL lookup runs automatically (browser-side). Export CSV includes clinical_analog_gt70 (yes/no). See Methods for confidence tier definitions.

Export results

Click Export CSV to download library_results_{target}_{date}.csv with ranks, metadata, scores, and physicochemical properties — ready for Excel or your ELN.

Generate → Library bridge

After a Generate job finishes, click Add top hits to library → on the Generate tab. VectaBind saves the library with its target (e.g. Generated · JAK2), switches you to Score, and selects that target in the target bar so batch scoring runs against the correct protein.

Selecting a generated library from the dropdown also switches the target bar and 3D pocket automatically. The button reads Run batch score · TARGET for the library’s protein.

Navigation & styles

• Drag to rotate · scroll to zoom · Reset returns to the default framing
• Cartoon — protein ribbon (default) · Surface — electrostatic VDW surface · Stick — atom/bond detail for inspecting contacts
• Widescreen — hide the triage panel for a wider 3D stage · Fullscreen — immersive view (Esc to exit)
• PNG — export a snapshot of the current view for slides or reports

Keyboard shortcuts (when not typing in a text field): 1 cartoon · 2 surface · 3 stick · R reset · F fullscreen

Click residues

Click any protein atom to see a tooltip (residue name, number, chain). The viewer zooms to that residue. Double-click for a tighter zoom. Useful for asking the pocket assistant about specific residues — the chat input is pre-filled with your selection.

Scored compounds in 3D

1. Run batch score — the #1 hit is auto-selected and overlaid in orange stick representation.
2. Click any result row to switch compounds; the ligand re-centers in the pocket.
3. Dashed lines show predicted polar contacts (≤3.6 Å) between ligand and pocket — orange for close contacts, blue for others. Labels show residue and distance.
4. The Pocket briefing dock (bottom right) lists the same contacts under the compound SMILES.

Pro users may get GNINA-docked poses; free tier uses a 3D conformer centered in the pocket for visualization (ranking still comes from the affinity model).

Chat-driven viewer commands

Use the pocket assistant (bottom left) or suggestion chips:

• "Zoom to active site" — frames catalytic / polar residues
• "Highlight binding pocket" — dims the protein and sticks key pocket residues in blue
• "Explain the top scored compound" — uses your latest results as context