F32. dissecting the relationship between major depressive disorder and cognition using genetics

Advancements in methodology have enabled the combination of innovative statistical approaches for analyzing summary statistics in genome-wide association studies (GWAS), and to investigate the shared genetic component of two disorders or traits. Genetic correlations give a genome-wide assessment of shared genetics, where the direction of effect is similar in both traits. New methods, such as LAVA, investigate genetic correlations within short genomic regions and enable more refined analysis of shared genetics highlighting concordant or discordant effects between traits. We conducted a comprehensive analysis of genome-wide local genetic correlation to identify genetic overlaps between Major Depressive Disorder (MDD) and the two traits of Intelligence (INT), and Educational Attainment (EA). Genome-wide summary statistics for each trait (MDD, EA, INT) were identified. We used LDSC regression to estimate the genome-wide heritability of each trait and derive global genetic correlation estimates. For the pairs of MDD-EA and MDD-INT, we used LAVA to obtain local genetic correlation estimates across approximately 2495 independent genomic blocks, defined by patterns of linkage disequilibrium (LD). Similarities and differences between LAVA results were assessed, to identify regions that might contribute differently to the role of EA and INT in MDD. To further analyze associations between pairs of traits, we applied statistical fine-mapping and colocalization techniques in regions where the False Discovery Rate (FDR) adjusted p-value of the local genetic correlation analysis was below 0.05, accounting for all bivariate comparisons. For MDD-EA, LAVA identified 402 local genetic correlations significant after a multiple testing correction for the number of regions, with all regions having a negative genetic correlation (i.e. MDD was associated with lower educational attainment). For intelligence, 62 regions with positive genetic correlations and 41 with negative correlations with MDD were identified. Comparisons of MDD-EA and MDD-INT regional correlations showed diverse patterns of correlations across the trait pairs. For example, regions on chromosome 3 were negatively correlated in both pairs, while regions on chromosome 9 had opposite effects. Some regions were associated in only one pair. Regions of interest within these clusters will undergo further exploration using colocalization techniques, potentially uncovering the presence of shared causal variants that play a role in the relationship between Major Depressive Disorder and cognitive traits. This study has identified multiple genetic regions that contribute differently to the role of educational attainment and intelligence in MDD. Fine mapping of these regions may lead to identification of causal variants shared between the traits, potentially elucidating the nature of the relationship between cognition and Major Depression Disorder.