David A Knowles - US grants

With ageing populations world-wide, neurodegenerative diseases are placing an ever increasing burden on long- term well-being, healthcare costs and family life. Despite decades of research and enormous investment, no disease-modifying treatment is available for the most common of these diseases: Alzheimer?s (AD). The majority of these, to-date unsuccessful, efforts have focused on one potential cause of AD: amyloid-? aggregation. Combining population-scale data collection, human genetics and machine learning provides a way forward to uncover and characterize new causal cellular processes involved in AD. This would provide an array of potential therapeutic targets, increasing the chance that one will be more easily modulated than the amyloid-? pathway. AD-specific genomic datasets of unprecedented scale are being actively collected: whole genome sequencing (WGS) from ~20k individuals, gene expression (RNA-seq) and epigenomics (ATAC-seq, histone ChIP-seq) from >1000 post-mortem AD brains, single-cell transcriptomes and similar modalities in peripheral and brain-resident innate immune cells (which we and others have shown to be AD-relevant). Effectively integrating these diverse data to better understand AD represents a substantial computational challenge, both in terms of data scale and analysis complexity. This proposal leverages state-of-the-art deep learning (DL) and machine learning (ML), combined with human genetic analyses, to address this challenge. We will train DL models to predict epigenomic signals and RNA splicing from genomic sequence, enabling in silico mutagenesis to estimate the functional impact (a ?delta score?) of any genetic variant. The delta scores will be used in genetic analyses that distinguish causal associations: cellular changes that drive AD pathogenesis rather than downstream/side effects of disease. Delta scores will aid in associating both rare and common variants to AD. To achieve sufficient power, rare variants must be aggregated (e.g. for a gene): delta scores will allow filtering out many likely non-functional (particularly non-coding) variants. Most common variants from AD Genome Wide Association Studies (GWAS) are simply correlated with the causal variant due to linkage disequilibrium (LD). Delta scores, combined with trans-ethnic GWAS, will enable estimation of the likely causal variant(s). These analyses will highlight variants and genes involved in AD. However, genes do not operate in a vacuum so robust probabilistic ML will be used to learn cell-type and disease-specific gene regulatory networks from sorted bulk and single-cell RNA-seq. The detected networks will be integrated with our genetic findings to discover network neighborhoods/pathways especially enriched in AD variants. Such pathways will be prime candidates for future functional and therapeutic studies of AD.

PROJECT SUMMARY Neuropsychiatric disorders (NPD) such as schizophrenia (SZ), autism spectrum disorders (ASD) and bipolar disorders (BD) are remarkably common, with SZ alone affecting nearly three million Americans. Despite more than fifty years of research, no cures exist for these conditions and the standard of treatment remains unsatisfactory. Genome-wide association studies (GWAS) indicate that, in addition to highly penetrant rare mutations, NPD risk also reflects the impact of hundreds of common single nucleotide polymorphisms with small effect sizes. A major challenge in the field has been illuminating the pathways connecting these genetic variants (the vast majority of which fall in non-coding sequences) to target genes and causal cellular phenotypes. To understand how these myriad risk loci causally contribute to disease risk, it is essential to screen for putatively causal variant(s) and determine how they influence gene expression, which has been shown to be cell-type specific, as well as cellular function. Recent evidence has emerged indicating a substantial contribution of RNA splicing variation to heritability across many complex genetic diseases, including SZ. Based on our preliminary analyses and the work of others, we hypothesize that a substantial proportion of NPD GWAS loci exert their pathogenic effects on neuronal function by impacting RNA: its structure, modifications, protein interactions and splicing. To test this, we will apply novel tools and machine learning methods to predict and quantify RNA splicing in the largest SZ, ASD and BD GWAS, in order to predict splicing quantitative trait loci (sQTLs, Aim 1). To confirm true effects on exon inclusion independently in glutamatergic and GABAergic neurons (i.e., the major cell-types impacted in NPD), up to several thousand of the predicted splice variants will be tested by a massively parallel reporter assay, MaPSy (Aim 2). Finally, in order to evaluate the cell-type-specific impact of putative causal sQTLs identified in Aims 1 and 2 on neuronal maturation and synaptic function, we will use CRISPR gene editing to engineer these mutations within human induced pluripotent stem cell (hiPSC)-based models of both neural cell types (Aim 3). Our overarching goal is to map and functionally evaluate the NPD-GWAS loci that impact alternative splicing and neuronal function. Our work may impact the field by delivering new insights into the role of common variants in NPD pathophysiology, which could inform ways of improving diagnostics, predicting clinical trajectories, and developing novel therapeutic interventions.