Peripheral Blood DNA Methylation–Based Machine Learning Models for Prediction of Knee Osteoarthritis Progression: Biologic Specimens and Data From the Osteoarthritis Initiative and Johnston County Osteoarthritis Project
Authors' conclusion
Our data suggest that pain and structural progression share similar early systemic immune epigenotypes.
Study context
- Population
- Individuals from the Osteoarthritis Biomarkers Consortium (OABC); validation in Johnston County Osteoarthritis Project and Osteoarthritis InitiativeOABC
- Participants
- 554 individuals
- Comparison
- Full models vs Parsimonious models
- Outcome
- Classification of future knee OA progression (pain, radiographic, pain plus radiographic, any progression)
Methodological detailsmachine learning, genome-wide DNA methylation (Illumina Infinium MethylationEPIC 850K arrays), development and validation sets, external validation in independent cohorts
- Machine learning
- Genome-wide DNA methylation (Illumina Infinium MethylationEPIC 850K arrays)
- Development and validation sets
- External validation in independent cohorts
Performance
As reported in the abstract
Classification of future knee OA progression (OABC)
- Full models, radiographic-only progression: mean ± SEM accuracy; AUC
- 87 ± 0.8%; 0.94 ± 0.004
- Full models, pain-only progression: accuracy; AUC
- 89 ± 0.9%; 0.97 ± 0.004
- Full models, pain plus radiographic progression: accuracy; AUC
- 72 ± 0.7%; 0.79 ± 0.006
- Full models, any progression: accuracy; AUC
- 78 ± 0.4%; 0.86 ± 0.004
- Pain-only vs radiographic-only progressors not distinguishable: mean ± SEM accuracy; AUC
- 58 ± 1%; 0.62 ± 0.001
Further findings
Parsimonious models, radiographic progression in validation cohorts
- JoCo osteoarthritis (OA) Project: mean ± SEM accuracy; AUC
- 80 ± 0.3%; 0.88 ± 0.003
- Previous Osteoarthritis Initiative (OAI) data: accuracy; AUC
- 80 ± 0.8%; 0.89 ± 0.002
- Parsimonious models showed similar performance to full models
Original publication
Peripheral Blood DNA Methylation–Based Machine Learning Models for Prediction of Knee Osteoarthritis Progression: Biologic Specimens and Data From the Osteoarthritis Initiative and Johnston County Osteoarthritis Project. Arthritis Rheumatol. Published 21 November 2022. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →