Mutational Scanning

Proper diagnosis of genetic diseases is very important for patients and their families. Thanks to advances in genome sequencing, doctors are now able to list all the mutations present in a patient’s genome - but the interpretation of these mutations is difficult. In our lab, we use experiments in yeast and human cells to study the consequences of mutations found in human patients.
Our experiments are focused on selected genes including PAX6, mutations in which cause eye malformations, and TP53, frequently mutated in cancer. In our previous experiments, we found a surprisingly good agreement between mutations in PAX6 that cause slow growth of yeast, and mutations in the same gene that damage eye development in patients. This allows us to predict the severity for thousands of mutations that were never previously seen in patients.
We now aim to move beyond disease prediction and understand why certain mutations are more harmful than others. To study this, we will measure how each mutation influences the collection of all RNA molecules found in cells, known as the transcriptome. We will also analyse thousands of mutated cells under the microscope. Applying artificial intelligence methods to these measurements will help us understand why mutations are harmful, and help find appropriate treatments. Finally we will measure the performance of candidate mRNA therapeutics in human cells, in order to develop protocols and software for the design of new mRNA therapies.

With more than a million human genomes sequenced worldwide, we urgently need methods that will allow the interpretation of human genetic variation at scale. Important unsolved questions in this area include the classification of variants as benign or pathogenic; identification of molecular mechanisms through which variants influence phenotype; understanding the dependence of variant effects on genetic and environmental factors; and prediction of the effects of medical interventions for specific variants. All these questions can be addressed through an experimental strategy which has emerged over the past decade, and which recently became widely known as deep mutational scanning (DMS).
In the past five years, our group has focussed on the development of DMS and other high-throughput techniques to understand relationships between sequence, structure and function of RNA. We applied these methods to study gene regulation, noncoding RNA biology, virus genome structures, and RNA engineering. In the next quinquennium, we will translate these technologies for clinical benefit by applying them to protein-coding genes involved in human disease and by building data-driven tools for the design of mRNA therapeutics. Our aims are to:
1. Comprehensively study the effects of mutations in selected human disease genes, focusing on disease associations and molecular mechanisms;
2. Adapt multidimensional phenotypic assays: single-cell RNA sequencing and high-content microscopy, for high-throughput analysis of genetic variants;
3. Leverage our studies of codon usage to build widely accessible tools for coding sequence optimization of mRNA therapeutics.
Our research will facilitate the diagnosis and treatment of genetic diseases, clarify the molecular mechanisms of disease mutations, and facilitate the development of new RNA therapeutics.