Our long-term goal is to understand the mechanisms perturbed by disease risk variants and how these may be targeted for therapy tailored to individual patients’ pathogenic profile. We are particularly interested in diseases of the immune system with special emphasis on the role of immune function in the brain. To achieve this goal we pursue analytical projects grouped into three thematic categories:

  1. Genetic discovery: we lead efforts to identify genetic variants underlying disease risk in GWAS, sequencing projects in common and rare diseases and cross-disease analyses.
  2. Identifying pathogenic mechanisms: we integrate genetic and genomic data to propose and test molecular hypotheses about biological processes perturbed by risk variants.
  3. Identifying mechanistic defects in patients: we identify specific molecular defects in patients driven by genetic liability and test whether these predict treatment response.

These goals are often best achieved in collaborative frameworks, so we participate in a number of collaborative efforts, including several international consortia.

Genetic discovery efforts

The first step in our strategy is to discover genetic associations to phenotype. We lead a number of such projects in various disease areas, either within the context of international consortia or as discrete activities within our group. Specific projects include:

Shared genetic effects across immune-mediated diseases. We have previously reported that many associations are shared across groups of immune diseases and that these correspond to distinct sets of interacting genes. We are now leading a meta-analysis across diseases of almost 100,000 cases and controls genotyped on the Immunochip by several international disease consortia (collaboration with Mark Daly). We have developed new methods to detect truly shared (vs. co-localizing) associations at both the summary data and genotype level in large cohorts, which we are currently deploying (collaboration with Ben Voight).

Genetic mapping of multiple sclerosis. We are active participants in the International MS Genetics Consortium: we are key members of a group pursuing a GWAS meta-analysis (led by Philip De Jager) across 40,000 cases and 60,000 controls, and are currently replicating suggestive associations in 15,000 new cases and 15,000 controls using a custom Illumina chip querying 100,000 SNPs. We also lead the IMSGC’s efforts to assess the contribution of rare non-synonymous variation to MS risk (with ExomeChip) in 35,000 cases and 40,000 controls. This work is supported by NIAID and NMSS funding.

Genetic mapping of vaccine response. We are leading GWAS of molecular responses to influenza vaccination (750 samples from the US from 5 collaborating centers) and cholera vaccination (1000 samples from India; collaboration with Indian Statistical Institute), in the context of the Human Immunology Project Consortium.

Detecting trans-acting eQTLs in humans and correlating them to disease associations. We have recently extended our cross-phenotype meta-analysis to detect trans eQTLs. We aim to find replicable, robust evidence of regulating groups of genes in trans and begin dissecting these regulatory circuits. Their potential overlap with GWAS hits is also of interest. We are currently applying these approaches to the GTEx project.

Exome sequencing in families with Rasmussen’s Encephalitis (RE). We co-lead an international consortium investigating the pathogenesis of RE, a rare immune-mediated epilepsy. We are currently exome sequencing 30 affected offspring trios from around the world to identify mutations underlying the disease. This work is supported by the RE Children’s Project.

Identifying pathogenic mechanisms

To detect pathways perturbed by risk variants, and thus underlying susceptibility, we are developing approaches to interpret genetic associations using genomic data. Our framework includes RNA sequencing, methylation and DNase I hypersensitivity readouts across multiple cell populations and we are currently developing robust analytical approaches to integrate multiple datatypes into sophisticated molecular portraits of pathogenic processes. Specific projects include:

Identification of disease-relevant tissues and cell populations. We have begun to address the question of tissue relevance, i.e. in what context risk variants act. We have previously reported enrichment of associated SNPs in tissue-restricted regulatory DNA regions (collaboration with John Stamatoyannopoulos) and over-expression of nearby genes in specific immune cell populations (collaboration with David Hafler). We are now developing new methods to detect which genes are being perturbed in specific cell subsets and what the molecular perturbations are, in order to describe pathogenic processes in their native context.

Finding pathogenic T cell pathways in MS and other immune diseases. We are developing approaches to use multiple genomic datasets (DNase I hypersensitivity, methylation data, RNA sequencing data etc) simultaneously to identify pathways mediating susceptibility to MS and other immune diseases in carefully characterized subpopulations of T cells. (collaboration with John Stamatoyannopoulos).

Methods to identify genes targeted by risk variants. We have previously shown that genes near immune disease risk variants encode proteins which interact in dense networks. We are now extending our methods to identify networks driving genetic risk in protein-protein and gene regulatory interaction maps, using graph theoretical approaches. These methods form the basis of a toolkit to identify candidate genes and networks driving disease susceptibility.

Identifying mechanistic defects in patients

The key question for pathogenic pathway defects is whether they describe distinct patient subsets and predict patient response to therapy. We are thus pursuing projects to identify such defects in patient populations and describe distinct subsets, and to investigate if patients in these subsets have better outcomes with treatments targeting these defective pathways. Our efforts include the following projects:

Identifying T and B cell defects in patients with autoimmune disease. We have just begun collecting 280 new-onset, treatment naïve patients with one of five immune-mediated diseases. We will generate T and B cell subpopulation profiles, expression profiles in total white blood cells and signaling and expression responses to IL23 activation in flow-sorted Th17 cells from each patient in order to characterize compartmental and functional differences unique to or shared across diseases. This work is supported by the NIAID via the Autoimmune Centers of Excellence, six of which are assisting us with patient collection.

Finding pathway defects that alter treatment response in MS. To determine whether perturbation of the MS susceptibility pathways we are discovering drives patient heterogeneity and divergent treatment outcomes, we are correlating specific molecular perturbation signatures with therapeutic outcomes in MS patients from previously collected clinical trial data.

Several other projects are currently in development.