Additional file 1 of Quantitative estimate of cognitive resilience and its medical and genetic associations

Additional file 1: Supplementary Figure 1. QQ plot for score analysis. Supplementary Figure 2. The trajectories of CR score-related components in different theoretical scenarios. Supplementary Figure 3. The correlation between predicted neuropsychological test scores and the test scores. Supplementary Figure 4. Characteristics of the damage and CR scores. Supplementary Figure 5. Comparison of CR scores stratified by any APOE ε4 vs. no APOE ε4 in the main cohort with comprehensive cognitive assessments and neuropathologic features in the expanded cohort with imputed data to be used in GWAS (n=6,518). Supplementary Figure 6. The protective effect of APOE ε2 could come from reducing damage. Supplementary Figure 7. The sex-specific model performance and model reduction. Supplementary Figure 8. Feature importance from two ML models combined, one trained on female-only data and the other male-only data, to explore sex-specific differences. Supplementary Figure 9. The evidence for MYOM2<>LOC101927815 association to CR-ANIMALS. Supplementary Figure 10. The evidence for APOE association to CR score from LOGIMEM. Supplementary Figure 11. The evidence for multiple associations to CR score from NACCUDSD. Supplementary Figure 12. The correlation between CR scores and the model’s residual. Supplementary Table 1. Cohort description and neuropathologic features. Supplementary Table 2. Cohort description with cognitive assessment scores. Supplementary Table 3. List of demographic and medical features used for CR score prediction. Supplementary Table 4. Demographics and CR scores for each cohort used in the GWAS. Supplementary Table 5. Summary statistics from linear regression models associating all primary factors of interest and CR scores from different types of cognitive assessment. Supplementary Table 6. Genome-wide association of damage estimates derived from different cognitive assessments.