Genomic restricted maximum likelihood (GREML) analysis to estimate the heritability of response/resistance in major depressive disorder (MDD)

Introduction: MDD is the leading cause of disability worldwide and lack of response to treatment is reported in ~30% of subjects. Pharmacogenomics studies performed so far failed to identify single SNPs with a replicable effect on anti-depressants response (AD). Instead of focusing on single SNPs, our study aims to evaluate the aggregated contribution on AD response for genome-wide variants and currently known genetic variants from pharmGKB, the main catalog of known pharmacogenetic variations. Methods: We used GREML-LDMS (LD- and MAF-Stratified) analysis to estimate the heritability of Citalopram response and treatment-resistant status (TRD) in MDD patients from the STAR*D anti-depressants pharmacogenetic trial. We estimated the overall genetic contribution for genome-wide SNPs present in the dataset, as well as for different groups of candidate SNPs from pharmGKB. Results: We detected a small, but significant, proportion of variance explained by SNPs associated with Citalopram in pharmGKB when including those with the lower level of evidence (V(G)/Vp 0.04; SE 0.02; p 2.23E-05). High heritability emerged when considering genome-wide SNPs for Citalopram response (V(G)/Vp 0.76; SE 0.23; p 0.026), Citalopram symptoms improvement (V(G)/Vp 0.82; SE 0.22; p 0.046) and TRD phenotype (V(G)/Vp 0.81; SE 0.24; p 0.022). Conclusions: Our preliminary results confirm a little effect of single SNPs and suggest that taking into account the overall genetic variability of pharmGKB AD-related SNPs and, even more, of genome-wide SNPs could improve antidepressant response/resistance prediction