MétaCan
Menu
Back to cohort
Record W3198806124 · doi:10.1021/acs.est.1c01418

Novel Bayesian Method to Derive Final Adjusted Values of Physicochemical Properties: Application to 74 Compounds

2021· article· en· W3198806124 on OpenAlexafffund
Timothy F. M. Rodgers, Joseph O. Okeme, J. Mark Parnis, Kyle Girdhari, Terry F. Bidleman, Yuchao Wan, Liisa M. Jantunen, Miriam L. Diamond

Bibliographic record

VenueEnvironmental Science & Technology · 2021
Typearticle
Languageen
FieldComputer Science
TopicComputational Drug Discovery Methods
Canadian institutionsThe Scarborough HospitalTrent UniversityOccupational Cancer Research CentreCancer Care OntarioEnvironment and Climate Change CanadaUniversity of Toronto
FundersNatural Sciences and Engineering Research Council of CanadaSveriges RegeringUniversity of Toronto
KeywordsIn silicoWorkflowBayesian probabilityComputer scienceConsistency (knowledge bases)Benchmark (surveying)HeteroscedasticityData miningBayes' theoremChemistryMachine learningArtificial intelligenceDatabase

Abstract

fetched live from OpenAlex

Accurate values of physicochemical properties are essential for screening semivolatile organic compounds for human and environmental hazard and risk. In silico approaches for estimation are widely used, but the accuracy of these and measured values can be difficult to ascertain. Final adjusted values (FAVs) harmonize literature-reported measurements to ensure consistency and minimize uncertainty. We propose a workflow, including a novel Bayesian approach, for estimating FAVs that combines measurements using direct and indirect methods and in silico values. The workflow was applied to 74 compounds across nine classes to generate recommended FAVs (FAV R s). Estimates generated by in silico methods (OPERA, COSMOtherm, EPI Suite, SPARC, and polyparameter linear free energy relationships (pp-LFER) models) differed by orders of magnitude for some properties and compounds and performed systematically worse for larger, more polar compounds. COSMOtherm and OPERA generally performed well with low bias although no single in silico method performed best across all compound classes and properties. Indirect measurement methods produced highly accurate and precise estimates compared with direct measurement methods. Our Bayesian method harmonized measured and in silico estimated physicochemical properties without introducing observable biases. We thus recommend use of the FAV R s presented here and that the proposed Bayesian workflow be used to generate FAV R s for SVOCs beyond those in this study.

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 imitation

Not 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.

metaresearch head score (Codex)0.005
metaresearch head score (Gemma)0.015
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Bench or experimental · Consensus signal: none
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.007
Threshold uncertainty score0.026

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0050.015
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0020.001
Science and technology studies0.0010.000
Scholarly communication0.0010.001
Open science0.0010.001
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0030.001

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.

Opus teacher head0.021
GPT teacher head0.289
Teacher spread0.268 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designBench or experimental
Domainnot available
GenreEmpirical

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".

Quick stats

Citations19
Published2021
Admission routes2
Has abstractyes

Explore more

Same venueEnvironmental Science & TechnologySame topicComputational Drug Discovery MethodsFrench-language works237,207