What is the relationship between the number of transcription factor binding motifs a SNP associated with Alzheimer's Disease alters and the extent to which the SNP changes gene expression (measured by the sum of AlphaGenome RNA-seq variant scores)?

This study uses computation to examine whether HaploReg annotated motif-alteration counts correlate with AlphaGenome RNA-seq expression predictions for Alzheimer associated single nucleotide polymorphisms (SNPs). In total 188 SNPs, 28 SNPs with zero motif alterations and 160 motif altering SNPs were analysed. The mean of the absolute RNA-seq scores predicted by AlphaGenome was taken across 371 tracks each within all SNP gene context combinations with the sum across 285 cell contexts and average across all scored genes being taken. The resulting gene expression index showcased an extremely weak positive association as the number of motif alterations increased ( with Spearman rho of 0.103891 and nominal two-sided p value of 0.155950). Therefore, no statistically significant monotonic trend was observed. The results concern model predictions with specific criteria and do not underline an absence of biological effects. Interpretation was limited due to confounding variables including: a small dataset, linkage dependence, lack of annotation regarding the direction of a transcription factor’s effects, and etc.