Additional file 4 of Refining epigenetic prediction of chronological and biological age

Additional file 4. Table S1. Top 10,000 epigenome-wide significant associations between CpG and age. Table S2. Top 10,000 epigenome-wide significant associations between CpG^2 and age. Table S3. Epigenome-wide significant CpG associations with all-cause mortality. Table S4. Epigenome-wide significant CpG associations with all-cause mortality, replicated using the coxme R package to account for relatedness. Table S5. Gene-set enrichment results for mortality EWAS hits, as returned by FUMA. Table S6. Model performance as a function of CpG subset size input into elastic net. Table S7. Model performance as a function of CpG2 subset size input into elastic net, in addition to top 10K linear CpG-age associated sites. Table S8. Components of cAge predictor, along with their respective weights. Table S9. Components of log(cAge) predictor, along with their respective weights. Table S10. bAge predictor components, along with their weights. Table S11. bAgeAccel and GrimAgeAccel associations to mortality across 6 datasets. Table S12. bAgeAccel, GrimAgeAccel, PhenoAgeAccel, and DunedinPACEAccel associations to mortality across 2 datasets. Table S13. bAgeAccel associations to mortality across 2 datasets, including another biological age clock as a covariate (GrimAgeAccel, PhenoAgeAccel, or DunedinPACEAccel). Table S14. Schoenfeld residuals and HR as a function of TTE cut-off in bAge prediction. (XLSX 5.3Mb).