Compare an assay
What can a model reproduce under a specified experimental protocol—and where does it disagree?
Bring the protocol, endpoint definition and reference measurements you have rights to use.
Mechanistic modeling · Research services
Mechanistic simulation studies for focused research questions.
A useful model starts with a specific question. Work directly with us to scope a study, examine its assumptions, and compare results against experimental evidence.
Scoped research engagements. Not a clinical decision tool.
THE STUDY, NOT THE SALES PITCH
Start here
What do you need
to understand?
- 01Define the inputsQuestion · data · endpoint
- 02Scope the modelMechanisms · assumptions · limits
- 03Compare & challengeReference data · sensitivity · failures
- 04Deliver an inspectable reportResults · run record · next experiment
Question first.
Assumptions visible.
Evidence before claims.
01 / A useful starting point
Narrow the question.
Make the work testable.
We discuss defined modeling problems—not a promise to simulate everything. Feasibility depends on the mechanisms, available data and endpoint you need.
Explore a mechanism
Which modeled processes could explain a response, and which assumptions are doing the most work?
Bring the biological hypothesis, relevant conditions and competing explanations.
Examine sensitivity
How much does a result change when uncertain inputs or experimental conditions change?
Bring the endpoint, plausible parameter ranges and the decision the study should inform.
These are potential study types, not completed client studies or guaranteed capabilities for every system.
02 / Working together
From a research question
to a defined engagement.
Start with a nonconfidential outline. We agree on feasibility, deliverables, fees and the limits of the study before analysis begins.
Download the study brief template- 01
Discuss the question
Clarify the endpoint, available inputs, intended research use and what would make an answer useful.
- 02
Agree on a paid scope
Define the model boundaries, comparisons, acceptance criteria, timeline and price. If the question is not a fit, say so.
- 03
Analyze, review, document
Review the results against the agreed criteria—including mismatches, uncertainty and what remains untested.
03 / The deliverable
More than a plot.
A result you can interrogate.
The exact package is agreed in the scope. A research report should make clear what went into the model, what came out and what those results do not establish.
Study report / proposed structureRESEARCH USE
- 01 Inputs & assumptions
- Data provenance, experimental conditions, units, parameter choices and calibration inputs.
- 02 Reproducibility record
- Model and run versions, configurations, commands and agreed output artifacts.
- 03 Comparisons & uncertainty
- Endpoint-matched comparisons, held-out status, sensitivity, errors and excluded or failed runs.
- 04 Limits & next questions
- What the analysis supports, what it cannot establish and what experiment could test it next.
Where the evidence stands
Ambitious research.
Specific claims.
oNeura is under active development. Current cardiac assay reproduction work is ongoing; we are not claiming a completed full-paper reproduction or independent validation.
Previously displayed trial comparisons are under review. They should not be treated as established predictive accuracy. Engineering demonstrations, retrospective comparisons and held-out validation are different kinds of evidence.
Read the evidence status & publication standard04 / Explore the ideas
A window into modeling.
Not proof of prediction.
These simplified browser demonstrations are educational interfaces, separate from a scoped native research study. Do not use them for patient care, dosing or treatment decisions.
Educational demoICU simulation Explore how simplified physiological variables interact in a simulated scenario.
Educational demoDrug interactions Explore modeled interaction mechanics—not prescribing or safety guidance.
Educational demoMolecular properties Explore structure-derived descriptors and the assumptions behind simple estimates.
Before we begin
Good questions.
Clear expectations.
Is this a self-service subscription?
No. The offer here is a directly scoped research engagement. A browser demonstration is not access to a production research service, and a subscription or delivery portal is not being promised.
What does a study cost?
Fees and timing depend on the agreed question, data preparation, model work and deliverables. We discuss feasibility first and provide a defined paid scope before work begins. Existing written commitments are not changed by this website.
Can I share confidential research?
Start with a nonconfidential outline only. Before transferring proprietary data, discuss confidentiality, permitted use, data rights and a suitable transfer method with us. Do not put patient information, confidential compounds or unpublished data into the inquiry form.
Can results support clinical or regulatory decisions?
The website and demos are for research and education, not clinical decision-making. We do not claim clinical validation or regulatory approval. Any intended regulatory use requires its own context-specific evidence and review; a research engagement does not guarantee it.
What if a result disagrees with the experiment?
That disagreement belongs in the report. We agree on comparison criteria before analysis and distinguish calibration from held-out testing. Failed runs, exclusions and limitations should be visible—not hidden behind an aggregate score.
Start with the question
What are you trying
to understand?
Tell us the endpoint, what data you have and your timeline. Keep the first conversation nonconfidential.
bobby@blackweb.ai
The inquiry tool prepares an email draft. This website does not send a message or confirm delivery.