Keynote: A renaissance for the point mutation: from legacy data to semantic web service
Bibliographic record
Abstract
Abstract Experiments that construct or discover protein point-mutations and investigate their functional consequences represent one of the cornerstones of biomedical investigation. And yet, despite the importance of these annotations, the interpretation and reuse of knowledge about their impacts remains a formidable task. This is due to a number of reasons including (i) the publishing of salient mutation impact descriptions in unstructured text, (ii) the existence of numerous boutique databases of mutation information which can have many years of latency, and (iii) errors within manually populated mutation databases. In recent years the mining of mutations from scientific documents has emerged as a promising research theme resulting in automated methods for mutation extraction and denovo database creation. While such methods have shown good performance, this alone does not suffice in seamlessly integrating mutation annotations to other biological datatypes.In this talk I demonstrate the challenges and innovations that have resulted in the deployment of semantic services that supply text-extracted mutation impact annotations on demand. These include; mutation grounding algorithms linking extracted mutations to the correct position on wild type protein sequences, the grounding of mutated protein properties to GO Molecular Function, rule based extraction of mutation impacts and impact direction, conceptualization and population of mutation impact ontology, graphical composition of SPARQL queries using mutation specific metadata, visualization of mutation impact annotations on protein structures, dynamic annotation of PubMed abstracts containing mutations mentions using the Semantic Assistant framework , and lastly SADI web service deployment facilitating queries that leverage mutation impact annotations integrated with multiple semantic web services, namely queries that select for mutations that impact specific protein properties, mutations that impact specific metabolic or signaling pathways, drugs that target mutated proteins, and literature describing mutations on proteins with a given nsSNP.
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 | 0.016 | 0.021 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.002 | 0.004 |
| Scholarly communication | 0.012 | 0.021 |
| Open science | 0.003 | 0.008 |
| Research integrity | 0.005 | 0.011 |
| Insufficient payload (model declined to judge) | 0.026 | 0.015 |
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".