Identification of functional regulatory variants implicates distinct transcriptional networks in dementia

We used 2 massively parallel reporter assays to characterize 5,706 unique noncoding genetic variants derived from genome-wide association studies for two neurodegenerative disorders: Alzheimer's disease (AD) and Progressive Supranuclear Palsy (PSP). Specifically, we synthesized a library consisting of both alleles of each SNV embedded within their native genomic context, which was cloned upstream of a minimal promoter within an expression vector. We then measured the transcriptional drive of each library element using next-generation sequencing of barcoded transcripts uniquely associated with each regulatory element. This enabled the identification of variants with significant transcriptional skew between alleles representing likely functional regulatory variants underlying disease risk. Overall design: We first used next-generation sequencing to uniquely map barcodes to alleles to create a barcode-variant lookup table. We then assessed allelic expression by amplifying and performing next-generation sequencing of barcodes from library plasmid DNA and cDNA across multiple biological replicates. Please note that the variant_list_MPRA* (i.e. MPRA*_variant_list.tsv) files have no associated raw data files and these files are actually variant ID metadata files needed to process the raw data.