Troubleshooting & FAQ

Login & Access

Authentication and session questions — sign-in, two-step verification, social login, and HIPAA session lock.

I can't log in.

Ensure you're using your corporate email address (not personal). If you just signed up, check your inbox for the activation email and click the link. If your account is already activated, try "Forgot Password" to reset.

My session locked — did I lose my work?

No. Session lock is normal HIPAA security behavior. Re-enter your password to resume exactly where you left off. All data is preserved.

I see a verification code screen after entering my password.

This is two-step verification (2FA) for security. Check your email for a 6-digit code and enter it. The code expires after a few minutes. If you don't receive it, check your spam folder.

Social login popup was blocked.

Your browser may block the Google/Microsoft/Apple popup window. Allow popups for this site in your browser settings, then try again.

Session lock requires re-authentication on every page reload by design. To avoid losing context, finish a workflow before refreshing.

Running BVCTs

Questions about generating virtual clinical trials — stratification timing, finding past runs, missions, modality detection, and the ENHANCEMENT workflow.

Stratification is taking a long time.

AI-powered stratification can take up to 60 seconds depending on disease complexity. The loading spinner is normal. If it exceeds 2 minutes, try refreshing and re-entering your parameters.

I don't see my previous BVCTs.

Check the left sidebar panel filters — you may have a status filter active (Pending/Validated/Failed). Click "All" to see everything. Also ensure you're logged in with the same account.

Protocol PDF won't download.

Check your browser's popup blocker — the download may have been blocked. Also ensure you have BVCT credits remaining (visitors cannot download protocols).

Which modality should I pick for my drug?

The platform auto-detects modality from your input: SMILES strings = small molecule, consecutive amino acid letters = peptide, VH/VL separator = antibody, ACGTU bases = oligonucleotide. If auto-detection is wrong, manually select from the dropdown.

What is BVCT Purpose (VALIDATION vs DISCOVERY)?

VALIDATION means you're testing a specific drug-disease hypothesis against known clinical data. DISCOVERY means you're exploring new possibilities without a pre-existing hypothesis. VALIDATION is the default.

What happens when I switch BVCT Type between ONGOING and PLANNING?

ONGOING evaluates a real trial currently running (requires an NCT ID). PLANNING designs a new trial from scratch (NCT ID is optional). The input fields change based on your selection.

Where are my BVCT Optimization / Remediation / Expansion records?

Each mission type has its own sidebar panel. Click BVCT Optimization (emerald), BVCT Remediation (amber), or BVCT Expansion (sky) in the left sidebar to see records for that mission.

Why is the "Predefine BVCT Protocols" button greyed out on my ENHANCEMENT mission?

ENHANCEMENT requires at least one Developability Goal selected (oral bioavailability, CNS penetration, half-life, immunogenicity, effect sizes, or adverse effects). Each selected goal becomes a GO/NO-GO gate the BVCT verdict must clear. Hover the disabled button to see the in-app explainer; check at least one goal in the Seed Drug Input panel to enable Predefine.

Where do my modality-specific design lever values go?

Lever values are persisted backend-primary, encrypted at rest with AES-256-GCM in enhancement_briefs.design_levers_encrypted. They populate the PDF appendix in your generated protocol, and they are restored when you reopen the same enhancement on another device. Switching modalities preserves your prior modality's lever values per-modality, so flipping between, say, mAb and ADC lets you compare designs without losing inputs.

Graph & Generator

3D knowledge-graph rendering and the hypothesis Generator — blank states, the Load Example(s) popup, and authoring a hypothesis from scratch.

The 3D knowledge graph is blank or very slow.

The graph requires WebGL support. Use a modern browser (Chrome, Edge, Firefox, Safari 15+). If your computer has integrated graphics, try reducing the layer count in the toolbar.

Why is my TARGETS tab empty after picking a disease?

Selecting a disease via Select Disease(s) no longer pre-regulates any default nodes — the TARGETS tab starts empty so you can author your own hypothesis. To start with a pre-populated set of regulated targets, use the Load Example(s) button to the left of Select Disease(s), or right-click nodes on the 3D graph to regulate them manually.

What does Randomize do in the Examples popup?

Clicking Randomize in the Load Example(s) popup header picks one card uniformly at random from the currently-filtered list. If you have a therapeutic-area chip active (e.g., "Oncology"), Randomize will only choose from examples in that area. Click the "All" chip to randomize across all 13 hypotheses.

Can I start a hypothesis without picking an example?

Yes. Close the Load Example(s) popup (Escape or X) and either click Select Disease(s) to pick a disease freely, or let the default obesity graph load silently in the background. Either way you get a disease-centered radiating layout with a blank TARGETS tab — build your hypothesis by right-clicking nodes to up- or down-regulate them.

Credits, Account & Team

Credits, sharing, team invitations, and the investigator fields that pre-fill your protocol PDFs.

My credits ran out.

Contact your organization administrator. If you're on an individual plan, credits can be purchased through the Profile settings.

Can I share BVCT results with my team?

Yes. Download the protocol PDFs and infographic reports. If your organization has "Shared History" enabled, team members can see each other's BVCT records directly in the sidebar.

How do I invite team members?

Open the Organization panel in the left sidebar (admin accounts only). Enter the team member's corporate email, assign a role (Admin or Member), and click Invite.

How do I pre-fill my name on protocol PDFs?

Open the Profile panel in the left sidebar. Fill in the Investigator fields (name, title, affiliation, sponsor, contact). These auto-populate every future protocol PDF.

Platform, Language & Positioning

Everything else — the DECIDE values, cookie preferences, languages, in-app feedback, and how BVCT differs from a data-trained AI model.

What are the DECIDE values on the welcome page?

DECIDE stands for: Drug-design-based, Effect-size-quantified, Causal-biology-driven, Independent of prior data, Decision-ready, Exact SoC-controlled. These are the six principles that make BioinvestGPT's virtual clinical trials unique.

How do I manage cookie preferences?

Scroll to the footer on the Hub or Login page and click "Manage Cookies." This re-opens the consent banner so you can change your choice.

How do I switch the interface to Chinese / Danish / German?

Use the language dropdown in the page footer (or the top-right on Login). Currently translated: Login, Hub, Landing page (FIC and BIC sections), AdvisorPanel welcome, and the Language Switcher itself. The Generator, Validator workflows, and this user guide remain English-only — broader translation is on the roadmap.

How do I report a bug or request a feature from inside the app?

Click the floating Feedback button at the bottom-right of any screen. It captures the current view, lets you draw on the screenshot with marker pens (rose / amber / emerald / slate), and sends the annotated image with your message to the support team. Works on the 3D knowledge graph too.

Is BVCT a data-driven AI model trained on patient datasets?

No. BVCT is a data-free causal-biology AI: predictions derive from drug structure, mechanism of action, and disease biology — not from training on real patient datasets. Virtual patient cohorts are de novo synthesized from disease physiology, not extrapolated from prior trial data. Your proprietary drug structures are never used to fine-tune any model. See the DECIDE acronym (D — Drug-design-based, I — Independent of prior data) on the welcome page for the full positioning.

Prospective track record: data.bioinvestgpt.com. BVCT outputs are model-based decision-support analyses, not investment advice.