References, not copies: every entry links to its original source and keeps where its metadata came from. Reading states organise the library; they are not a judgement of scientific quality.
Published sources behind the validated MVA · BUB1B evidence snapshot. Each reference is cited by at least one evidence record. Open the research case.
Supplementary audit documentation · October 24, 2026 research revision · 15 files at commit bc39224
Computational audit documents of this collection's research case, linked at a fixed commit of the research repository. They are not references of the library, experiments or an official resubmission, and they are not counted above. They are read in the “Robustness and limits” step of the case trail and on the case page.
Decision matrixCandidate decision matrix
One page: verified product, exact versus other-model evidence, exposure, function, current gate and decisive experiment. Not an efficacy ranking.
Audit guideBounded exploratory robustness audit
What is stable, what changes and what remains insufficient; reproduction commands and execution provenance.
Protocol amendmentProtocol amendment 01
Recorded before execution: omits a shared-denominator genome-wide correlation, adds an equal-context-weighted PTC summary.
ResultsSensitivity results (machine-readable)
Source inventory and hashes, PTC rows and contexts, every deletion, GSE65145 features and contrasts, stable / changes / insufficient.
Elena Torban, Lucie Canaff, Sima Babayeva and 5 moreScientific Reports
Autosomal-Dominant Polycystic Kidney Disease, ADPKD, is the most common genetic kidney disease affecting 1:1000 people worldwide. It is caused by mutations in the PKD1 (~ 80%) or PKD2 gene (~ 15%). Although the germline … abstract via Europe PMC
Evidence snapshot note: Torban et al. — PKD1 readthrough and partial cellular rescue. Model: Human HEK293 PKD1 reporters and WT9-12 cells; other-gene evidence. Where: Figures 1H-I, 4, 5 and 6; published 2025-11-22 · cited by MVA-RT-TRANSFER-009
External results are not part of the library. They come live from the public sources named on each card; save one to keep it with its provenance.
Type a query above and search the public sources. Results stay separate from the library until you save them.
BioLab is the biomedical research area of Stackfire Labs: research cases built around open evidence, reproducible reanalyses and an explicit record of what remains unresolved. It organizes evidence and proposes experiments; it does not run patient experiments or recommend treatments.
Two germline BUB1B variants in mosaic variegated aneuploidy raise separate questions: L737* is an RNA and translation question, N1002K a protein stability and function question. The case records reanalyses and evidence audits, an October 24, 2026 robustness audit and a candidate decision matrix. No exact-genotype pharmacological rescue is established; wet-lab validation remains future work.
A shared research library of publications kept as references, not copies. Every entry links to its original source and records which public bibliographic services supplied its metadata and when. Reading states and collections are bibliographic organization, never scientific validation.
The graph connects a case to its variants, processes, evidence records, publications and hypotheses. Solid lines are documented connections; dashed lines are hypotheses under review. A connection says that a source or record links two items: it is not evidence of causation or efficacy. Every view also has a list alternative.
A research trail is an ordered reading of the evidence, step by step. Hypotheses are statements under research with the evidence that informs or limits them; none is marked validated. Evidence gaps name what has not been measured or located. All of them link to existing records and publications instead of copying them.
The auditable record of the MVA reanalyses with their execution records and downloads, and a Protein Structure Explorer of the predicted BUBR1 model. Tools marked Planned are not built yet. Reanalyses of public processed data and predicted structures are not biological experiments.
No. BioLab is research context, not a clinical diagnostic report or a treatment recommendation. Drug candidates are research hypotheses with explicit gates, not proven treatments. Questions about a person’s care belong with their clinical team.