Which findings support this research question, and where do the sources disagree?
A source-linked evidence matrix and gaps for review.
A useful collaboration
starts with a real unknown.
Bring a bounded biomedical research question, relevant evidence and a possible way to test progress. We will use that starting point to scope a computational research or forecasting pilot.
Early-stage collaboration exploration. Scope, capacity and any delivery commitments are agreed for each project.
WHY COLLABORATEA question becomes useful
A question becomes useful
when it can meet evidence.
Research often pauses at an unclear assumption: the wrong population, an unsupported mechanism, a missing comparison or a prediction without an answer rule.
A first pilot should make that uncertainty specific. The aim is an inspectable piece of work: evidence gathered, assumptions declared, calculations checked and next steps that follow from what was actually observed.
QUESTIONS WE CAN SCOPEStart with an uncertainty
Start with an uncertainty
that has a workable test.
Can a defined observation be reproduced from the available public or authorized data?
A scoped calculation, checks and a reproducible output record.
Which population, comparison or endpoint assumption changes the proposed study?
A research-design draft and explicit sensitivity questions.
What exact future outcome would test this judgment?
A candidate forecast question, evidence cutoff and proposed decision rule.
Forecast submission and full model comparison are planned capabilities. A pilot scope must reflect the tools and review resources actually available.
WHO THIS IS FOR
People close to the question.
Computational researchers
With public datasets, reproducible code or a well-defined analysis bottleneck.
Biotech and translational teams
With a candidate asset, target population or clinical-development question to examine.
Scientific reviewers
Who can assess evidence quality, assumptions and what would count as a useful result.
Experimental collaborators
Who may later test an agreed hypothesis in their own authorized setting. Such experiments are a separate scope.
WHAT EACH SIDE BRINGSDefine the work
Define the work
before promising an outcome.
Predome’s starting point
- A structured question and evidence workflow.
- Public-source collection, numeric checks and descriptive analysis.
- Traceable records, explicit limits and a reviewable deliverable.
- Engineering effort scoped to the available tools and capacity.
A collaborator’s contribution
- Domain context and why the unknown matters.
- Public sources or authorized access to relevant data and tools.
- A feasible verification path and appropriate expert review.
- Feedback on what the analysis supports and what it cannot establish.
Before sharing non-public material, agree the data scope, confidentiality, responsibilities and any publication or intellectual-property terms. No funding, authorship or scientific outcome is guaranteed by exploring a collaboration.
A BOUNDED FIRST PILOTAgree the question.
Agree the question.
Inspect the work. Review the result.
Frame the question
Describe the uncertainty, existing evidence and a possible way to judge progress.
Check the fit
Review data access, tool availability, expertise and the actual limits of the prototype.
Define the pilot
Agree inputs, responsibilities, deliverable, review criteria and a realistic scope.
Run and record
Keep the sources, parameters, outputs, failures and any revisions together.
Review what follows
Assess the result against the agreed criteria. Continue, revise or stop with the reason on record.
SEE THE WORK BEFORE SCOPING A PILOTInspect a question.
Inspect a question.
Follow its evidence.
The public case records show the kind of bounded computational work available today. The existing research workspace is separate and requires authorization.
FREQUENT QUESTIONS
Before we begin.
Do I need a complete proposal?
No. A concise question, relevant context and a possible source of evidence can provide a starting point. The first discussion should establish what progress would look like and whether a feasible pilot can be scoped.
Does Predome carry out wet-lab experiments?
The current project is computational. Any future laboratory work would require an identified collaborator, an agreed protocol, suitable authorization and a separate account of what was actually performed.
How would non-public research data be handled?
This public website does not collect research files. Any non-public data work would need an appropriate access arrangement, an agreed data scope and suitable review before material is shared.
Are model predictions already validated?
No. The public prototype includes registry references, question drafts and a scoring explainer. Independent model submissions and prospective outcome adjudication remain in development; a validated prediction track record has not been established.
Is a funded collaboration program open?
No funded program is announced here. Any collaboration, available capacity, fees, support or schedule would need to be discussed and agreed for a specific project.