Modeling of Protein Structure and Function
In this subproject, differentially expressed AS events in the coding sequences will be mapped to protein sequences and three-dimensional (3D) structures. We will provide characterization of the functional and structural impact of the AS events on proteins and protein interactions. Thus, this subproject will constitute the crucial step in the pipeline leading from the raw Omics data to comprehensive functional characterization of disease-causing mechanisms behind aberrant AS.
We will model protein 3D structure and protein complexes affected by the AS events and describe their potential functional impact in terms of affecting functional sites and protein-protein interaction (PPI) interfaces. We will split all AS events into those introducing in-frame insertions and deletions into the corresponding mRNA, and those that introduce not-in-frame events. The latter class is expected to be rare and will be deemed immediately deleterious for protein function. To assess the functional effect of the insertions and deletions caused by the in-frame AS events, we will train a machine-learning classifier based on annotated genetic indels.
We will then predict the functions of the affected proteins using publicly available databases. We will predict specific amino acid residues crucial for protein function and substrate specificity and will further characterize their spatial positioning and their change in alteration upon the AS events and study their effect on PPI interfaces . Finally, we will develop a new method to assess the functional impact of AS events on a systems level.
This will provide the basis for functional hypotheses for disease-causing mechanisms of AS transcripts, and will allow to prioritize AS transcripts for experimental validation. The results of this subproject will also feed the analysis of PPI networks in SP3 and hence contribute to the reconstruction of the complete picture of the disease. We will apply the developed tools to differentially expressed isoforms in DCM, HN, and healthy controls. Together with SP1 and 3, this will allow to propose hypotheses for disease-causing mechanisms of AS, which can be subsequently tested in animal models in SP4.