Public catalogue
What the corpus needs, and what this platform can run.
The corpus reads research papers and records the analysis each one used. This catalogue joins that to the assay definitions this platform holds, so the gap between what science needs and what can be run is visible rather than asserted.
Corpus snapshot ba34b49490d9bfaf · run run-20261023T122449232913 · published 2026-10-23T12:25:29.022232+00:00
- 5,529
- Papers analysed
- 303
- Papers with the method pinned down
- 20
- Families of method
- 8
- Assay definitions held
- 2
- Families runnable end to end
- 24
- Analysis operations still to write
The corpus sample read for the method each paper used and for where its data lives.
Of those, the analysis was identified precisely enough to route without a person deciding.
The methods cluster into families — group comparison, incidence rates, survival analysis — and each family needs a known set of analysis operations.
Versioned programmes the platform can run today, from qPCR measurement to compound-response screens to statistical operations.
One-way analysis of variance and incidence-rate modelling. Both proved against synthetic data with known answers, both replayable from their evidence bundle.
The corpus names the operations it needs and the platform has not implemented yet. They are ranked by how many papers each one unlocks and written in that order.
The method families
Each family is a group of papers that share a kind of analysis. A family is runnable here when this platform holds an assay definition that covers every operation the family needs.
| Family | Papers | Analysis it needs | Status here |
|---|---|---|---|
| Regression modelling regression_modelling | 87 | glm_logit_fit ols_fit regression_modelling_mixed_effects regression_modelling_negative_binomial regression_modelling_poisson | Not built here |
| Group comparison group_comparison | 43 | group_comparison_anova group_comparison_mann_whitney group_comparison_wilcoxon student_t | Partly built group-comparison-analysis |
| Survival analysis survival_analysis | 40 | cox_ph logrank_test survival_km | Not built here |
| Categorical (contingency table) analysis contingency_analysis | 24 | chi_square_test contingency_analysis_fisher_exact | Not built here |
| Image analysis / quantification image_quantification | 22 | image_quantification_densitometry image_quantification_segmentation | Not built here |
| Machine learning machine_learning | 19 | machine_learning_decision_tree machine_learning_gradient_boosting machine_learning_random_forest machine_learning_support_vector_machine | Not built here |
| Epidemiological rate modelling epidemiological_modelling | 16 | epidemiological_modelling_incidence_rate | Runnable end to end incidence-rate-analysis |
| Principal component analysis principal_component_analysis | 16 | pca_fit | Not built here |
| Diagnostic test evaluation diagnostic_evaluation | 13 | diagnostic_evaluation_roc | Not built here |
| Meta-analysis / evidence synthesis meta_analysis | 12 | fixed_effects_pool random_effects_pool | Not built here |
| Correlation analysis correlation_analysis | 11 | pearson_r spearman_rho | Not built here |
| Genetic association (GWAS / linkage / MR) genetic_association | 9 | genetic_association_linkage gwas_locus_scan mr_estimate | Not built here |
| Sequence alignment / homology sequence_alignment | 8 | sequence_alignment_blast sequence_alignment_msa | Not built here |
| Multi-omics integration multi_omics_integration | 7 | multi_omics_integration_pathway_progeny | Not built here |
| Network / pathway analysis network_analysis | 7 | network_analysis_centrality network_analysis_ppi_network pathway_enrich | Not built here |
| Clustering clustering | 5 | hierarchical_cluster kmeans_fit | Not built here |
| Flow cytometry analysis flow_cytometry_analysis | 2 | flow_cytometry_analysis_mfi | Not built here |
| Differential expression analysis differential_expression | 1 | differential_expression_limma_voom | Not built here |
| Model calibration / clinical utility model_calibration | 1 | model_calibration_decision_curve | Not built here |
| Reliability and agreement analysis reliability_analysis | 1 | cronbach_alpha | Not built here |
The build list
Every analysis operation the corpus names, in the order this platform intends to write them: by the number of papers each one unlocks on its own. The two at the top are written and proved.
| Operation | Papers that need it | Papers it unlocks alone | Built here |
|---|---|---|---|
| group_comparison_anova | 27 | 23 | Yes |
| epidemiological_modelling_incidence_rate | 16 | 15 | Yes |
| image_quantification_segmentation | 17 | 14 | Not yet, and not in the corpus catalogue |
| diagnostic_evaluation_roc | 13 | 12 | Not yet, and not in the corpus catalogue |
| contingency_analysis_fisher_exact | 9 | 7 | Not yet, and not in the corpus catalogue |
| machine_learning_support_vector_machine | 11 | 6 | Not yet, and not in the corpus catalogue |
| multi_omics_integration_pathway_progeny | 7 | 6 | Not yet, and not in the corpus catalogue |
| sequence_alignment_blast | 6 | 6 | Not yet, and not in the corpus catalogue |
| group_comparison_mann_whitney | 9 | 5 | Not yet, and not in the corpus catalogue |
| student_t | 6 | 5 | Not yet, and not in the corpus catalogue |
| machine_learning_random_forest | 9 | 4 | Not yet, and not in the corpus catalogue |
| image_quantification_densitometry | 5 | 4 | Not yet, and not in the corpus catalogue |
| network_analysis_ppi_network | 4 | 4 | Not yet, and not in the corpus catalogue |
| group_comparison_wilcoxon | 5 | 3 | Not yet, and not in the corpus catalogue |
| regression_modelling_mixed_effects | 4 | 3 | Not yet, and not in the corpus catalogue |
| machine_learning_decision_tree | 4 | 2 | Not yet, and not in the corpus catalogue |
| fixed_effects_pool | 2 | 2 | Not yet, and not in the corpus catalogue |
| flow_cytometry_analysis_mfi | 2 | 2 | Not yet, and not in the corpus catalogue |
| regression_modelling_poisson | 2 | 2 | Not yet, and not in the corpus catalogue |
| sequence_alignment_msa | 2 | 2 | Not yet, and not in the corpus catalogue |
| genetic_association_linkage | 4 | 1 | Not yet, and not in the corpus catalogue |
| regression_modelling_negative_binomial | 2 | 1 | Not yet, and not in the corpus catalogue |
| differential_expression_limma_voom | 1 | 1 | Not yet, and not in the corpus catalogue |
| model_calibration_decision_curve | 1 | 1 | Not yet, and not in the corpus catalogue |
| network_analysis_centrality | 1 | 1 | Not yet, and not in the corpus catalogue |
| machine_learning_gradient_boosting | 2 | 0 | Not yet, and not in the corpus catalogue |
What this catalogue does not tell you
- It counts papers, not verified replications. No dataset has been downloaded, access-checked or licence-read, so a paper being in the corpus says nothing about whether its data can be used.
- "Runnable end to end" means the analysis operation is implemented, proved against synthetic data and replayable from its evidence bundle. It does not mean a physical assay has been run at a facility.
- The facilities named in the provider registry have not been contacted. Their access state stays "named, not verified" and any rate shown is a published rate, never a quote.
- The corpus is a sample of the published literature, not the literature. It covers the methods it could determine, and it says so on the process map for every paper it could not.
Caveats published with this snapshot
- Method extraction is rule-based and has no hand-judged accuracy sample. Evidence bands order textual support; method agreement intervals are null (n=0), never probabilities. Work-type validation does not validate methods.
- Operation coverage describes only explicitly named methods. It does not prove a complete protocol, compatible parameters, successful compilation or re-execution. Unrecognised or ambiguous methods require review.
- Data reachability is a textual accession or public-data statement. No dataset was downloaded or checked for access, completeness or licence.
- Capabilities are planning states seeded from the dated equipment register. No bench ownership, completed quote or partner booking is inferred. Procurement counts are dependencies, not guaranteed replications or financial returns.
- Tier and blocker counts are deterministic assessments of this corpus, not measured agreement. R3 is reserved for procurement; unresolved methods and data require human work (R4). The most restrictive axis wins, with capability, method, data as tie-breaks. Resolving one blocker can leave another at the same tier.