Manolis Kellis on Dissecting Disease Circuitry at Single-Cell Resolution

From genomics to therapeutics: Uncovering and manipulating human disease circuitry at single-cell resolution

Talk at UC Irvine MultiScale Cell Fate Symposium

Outline:

1. Map and characterize the circuitry of non-coding elements

- Epigenomic maps of non-coding elements across many cell types

- Link SNPs to enhancers, target genes, regulators, cell types

2. Identify disease-relevant tissues, variants and regulators

- Cell type: Immune causal basis of Alzheimer’s. New loci: heart disease

- Causal SNP, target gene, TFs. Bayesian methods: RiVIERA, CONVERGE

3. Combine genetic + epigenetic + expression variation in disease

- AD mediation: GWAS, mQTLs, MWAS, iMWAS. Causal, new loci and targets

- Multi-tissue mediators: AD factorized QTL analysis (fQTL) in brain regions

4. Single-cell dissection of epigenomic + transcriptional variation

- scRNA-seq in AD: brain cell types, AD-specific sub-populations; sex diffs

- Deconvolution: Cell fraction changes; cfQTLs; cell-type-specific GWAS effects

5. Uncover + manipulate disease mechanism and circuitry

- FTO: cell type, target genes, causal SNP, regulator, browning

- Reverse disease phenotypes by DNA editing and gene KD/OE

6. Heterogeneity: massively-parallel assays, multi-trait modeling

- Health records, classify diseases/patients/faces/epig, reprogram cells

- Massively parallel assays: MPRA tiling, HiDRA, iPSC diff, 384-plex KO/KD

Abstract:

Perhaps the greatest surprise of human genome-wide association studies (GWAS) is that 90% of disease-associated regions do not affect proteins directly, but instead lie in non-coding regions with putative gene-regulatory roles. This has increased the urgency of understanding the non-coding genome, as a key com-ponent of understanding the mechanistic basis of human disease. To address this challenge, we generate transcriptional and epigenomic maps of cellular circuitry across 100s of reference human tissues and cell types, 1000s of individuals in disease-relevant tissues, and 10,000s of individual cells in complex tissues, across patients and control individuals. We use the resulting datasets to infer regulatory networks linking genetic variants to their target genes, their upstream regulators, the cell types where they act, and the pathways they perturb, thus providing unbiased views of disease mechanisms, and sometimes re-shaping our understanding of common disorders. For example, we found that genetic variants contributing to Alz-heimer’s disease act primarily through immune processes, rather than neuronal processes. We also found that the strongest genetic association with obesity acts via a master switch controlling energy storage vs. energy dissipation in our adipocytes, rather than through the control of appetite or exercise in the brain. Combining genetic, epigenomic, and transcriptional variation across patients and healthy controls, we infer causal genes and regions that mediate the effect of genetic variants on disease phenotypes, pinpointing driver genes and regions in Alzheimer’s disease, heart disease, and cancer. We combine single-cell profiles, tissue-level variation, and genetic variation across healthy and diseased individuals to deconvolve bulk profiles into single-cell profiles, to recognize changes in cell type proportion associated with disease and aging, and to partition genetic effects into the individual cell types where they act. We expand these methods to electronic health records to recognize meta-phenotypes associated with combinations of clinical notes, prescriptions, lab tests, and billing codes, to impute missing phenotypes in sparse medical records, and to recognize the molecular pathways underlying complex meta-phenotypes in genotyped individuals by integration of molecular phenotypes imputed in disease-relevant cell types. Lastly, we develop programmable and modular technologies for manipulating these pathways by high-throughput reporter assays, genome editing, and gene targeting in human cells and mice, demonstrating tissue-autonomous therapeutic avenues in Alzheimer’s, obesity, and cancer. These results provide a roadmap for translating genetic findings into mechanistic insights and ultimately new therapeutic avenues for complex disease and cancer.