Discussion on “Bayesian meta-analysis of penetrance for cancer risk” by Thanthirige Lakshika M. Ruberu, Danielle Braun, Giovanni Parmigiani, and Swati Biswas
Bibliographic record
Abstract
I read the paper by Ruberu et al. (henceforth “the authors”) with great interest, finding myself impressed by the novelty and relevance of the work. In the context of inferring cancer penetrance among those with a particular genetic mutation, the authors accept the thorny challenge of meta-analyzing a set of studies which have basic disparities in what is being estimated. Through a thoughtful, first-principles development of a Bayesian hierarchical model, the authors demonstrate a successful solution to the problem. The crux of the proposed method is to draw together inferences from multiple studies (and their chosen analyses), acknowledging that some study and analytic method dyads have different estimands (targets of inference) than others. This work is particularly timely given the recent groundswell of reflection on, and calls for clarity about, estimands. To quantify this surge, Figure 1 shows the time trend of “estimand(s)” appearing in article titles among journals indexed by the PubMed database (pubmed.ncbi.nlm.nih.gov). While statisticians have always thought carefully about the target of an inference, there is a push to bring this more to the fore in applied work.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.169 | 0.327 |
| Meta-epidemiology (narrow) | 0.001 | 0.002 |
| Meta-epidemiology (broad) | 0.004 | 0.007 |
| Bibliometrics | 0.005 | 0.006 |
| Science and technology studies | 0.003 | 0.005 |
| Scholarly communication | 0.006 | 0.011 |
| Open science | 0.006 | 0.006 |
| Research integrity | 0.021 | 0.031 |
| Insufficient payload (model declined to judge) | 0.021 | 0.006 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".