After my initial scientific training in synthetic biology and protein engineering, I became excited to apply this molecular perspective to understand natural genetic variation and the complexity of heritable traits.
My research program combines (I) retrospective approaches to assess the mechanisms of action of naturally occurring mutations through experiment and computation with (II) prospective, synthetic approaches to test precisely defined molecular hypotheses derived from retrospective analyses.
A central challenge of genetics is to understand the function of natural genetic variants. Yet despite the proliferation of genome sequencing, interpreting the importance of mutations remains difficult. Most variants are of uncertain significance, and heritable traits in health and disease often depend upon large numbers of mutations throughout the genome.
My strategy to bridge this gap in understanding is to chart natural genotype-to-molecule-to-phenotype maps by comprehensively connecting individual genetic changes to changes in molecular outcomes (abundance of mRNA, proteins, and metabolites) and cellular and organismal phenotypes. We will directly test predictive models derived from these maps using genome editing and saturating mutagenesis.
The genotype-to-phenotype relationship is not static: a mutation that is pathological in one patient may be harmless in another. The importance of ‘epigenetic’ factors has long been appreciated, but it remains challenging to predict how epigenetic perturbations will alter the impacts of mutations. For instance, inhibiting the protein chaperone Hsp90 (a hub for the proper folding of many regulatory proteins, including transcription factors and kinases) has profound developmental consequences and can alter the fitness effects of natural genetic variants, but the molecular mechanisms underlying these interactions remain largely obscure.
My colleagues and I recently found that the RNA chaperone Lhp1 (LARP) can have similarly widespread influences on how mutations impact molecular and cellular outcomes. Understanding the influence of RNA and protein quality control on complex traits is a promising avenue to predict the variable penetrance of genetic disease and understand how the environment influences evolution.
The newfound availability of predicted structures for nearly every known protein sequence has unlocked vast possibilities in using protein structure to understand the outcomes of mutations. Although structural data has typically been used to predict the effects of amino acid changes, there is mounting evidence that synonymous variants (which do not alter protein sequences) also control protein abundance, fold, and function. Indeed, synonymous mutations can cause diseases like cystic fibrosis and Huntington’s disease.
Combining my nucleotide-resolution molecular mapping approaches, which assign function to missense and synonymous variants alike, with comprehensive mRNA and protein structural information and databases of natural mutations from fungi to plants to humans will enable predictive models of these previously enigmatic effects.