Bibliographic record
Abstract
It is common to sex-stratify analyses of omics data and to report effects as 'sex-specific' when they are significant in only one sex. However, when analysing hundreds or thousands of molecules, this approach will yield many spurious 'sex-specific' effects if not supported by significant interactions. I illustrate this problem using an RNA sequencing dataset showing almost no significant sex by treatment interactions, but where sex-stratified analyses yield hundreds of 'sex-specific' effects of treatment. These 'sex-specific' effects could be spurious or could be real but not show interactions due to low statistical power. To distinguish these possibilities, I describe permutation tests, which provide an intuitive way to determine if a pattern of observations differs from what would be expected due to chance. For this dataset, assigning sex at random often generates more 'sex-specific' effects than the real data, demonstrating that there is little evidence of sex differences. Next, I simulate an RNA sequencing dataset that includes genes modelled to have sex-specific effects of a condition. As expected, analysis of this simulated dataset yields both significant interactions and sex-specific effects in sex-stratified analyses. While stratified analyses detect a higher number of sex-specific effects than the analysis of interactions, they erroneously identify genes not modelled to show sex-specific effects more often than interactions. A permutation test confirms that the number of sex-specific effects observed in the simulated dataset is greater than expected due to chance. Permutation tests can be applied to omics studies of sex differences, simultaneously providing (i) a clear and simple demonstration of the problems of sex-stratified analyses, and (ii) additional evidence of sex-specific effects where these are present. R code is provided for permutations, simulations, and plots to visualize potential sex-specific effects, which can be adapted to other types of data.
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 | 0.031 | 0.156 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.003 |
| Bibliometrics | 0.004 | 0.004 |
| Science and technology studies | 0.002 | 0.003 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.002 | 0.006 |
| Insufficient payload (model declined to judge) | 0.013 | 0.002 |
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".