Developing an integrative bioinformatic approach for the selection of therapeutic targets

化学探针是合成的小分子配体，可选择性地与给定蛋白质结合以调节其在细胞中的功能。因此，它们是研究蛋白质功能最有用的工具之一。为了系统地探索人类生物学，理想情况下应该为人类蛋白质组中的每种蛋白质开发化学探针。事实上，这一目标是无法实现的，因为并非所有蛋白质都具有具有裂缝/口袋的结构，这些裂缝/口袋具有细胞渗透性小分子配体的有效和特异性结合所必需的几何和物理化学特性：并非所有蛋白质都是可配体的。此外，结合口袋在结构上必须是独特的，以便开发选择性结合目标蛋白质的小分子配体：化学探针只能针对具有足够结构多样性的蛋白质类别开发。 我们建议结合多种结构生物信息学方法来分析人类蛋白质组中发现的所有蛋白质结构域（即 3D 结构模块）的化学易处理性。 1) 我们将分析蛋白质结构域的原子 3D 结构，以预测其结合口袋的配体性。 2）我们将评估结合袋的结构多样性，这是开发选择性配体的重要参数。我们还将分析蛋白质结构域的构象动力学，这对基于结构的配体设计和变构抑制剂的发现具有影响。 3）我们将结合配体性、结构多样性和构象动力学特性来绘制蛋白质结构域的化学易处理性景观，并建立路线图来指导新兴的大规模努力，系统地开发人类蛋白质组的化学探针，加拿大在这一研究领域处于领先地位。长期目标是将分析扩展到非人类蛋白质组（包括原核生物）。 从事该项目的学生将获得结构生物信息学方面的一套罕见的专业知识，将体验采用开放科学原理的好处，并将在由结构生物学家、生物化学家、细胞生物学家组成的跨学科、开放概念的工作环境中进行培训。他们还将能够与多伦多市中心充满活力的当地生物信息学社区分享并向其学习。

Chemical probes are synthetic small-molecule ligands that bind selectively to a given protein to modulate its function in cell. As such, they are one of the most useful tools to investigate the function of proteins. To systematically explore human biology, a chemical probe should ideally be developed for each protein in the human proteome. In reality, this goal is not attainable as not all proteins have a structure that features crevices/pockets with geometrical and physicochemical properties necessary for the potent and specific binding of a cell-permeable small molecule ligand: not all proteins are ligandable. Additionally, binding pockets must be structurally unique in order to develop small molecule ligands that bind selectively to the protein of interest: chemical probes can only be developed for classes of proteins with sufficient structural diversity. We propose to combine a variety of structural bioinformatic approaches to analyze the chemical tractability of all protein domains (i.e. 3D structural modules) found in the human proteome. 1) We will analyze the atomic 3D structure of protein domains to predict the ligandability of their binding pockets. 2) We will evaluate the structural diversity of binding pockets, an important parameter for the development of selective ligands. We will also analyze the conformational dynamics of protein domains, which has implications for structure--based ligand design and the discovery of allosteric inhibitors. 3) We will combine ligandability, structural diversity, and conformational dynamics properties to draw a chemical tractability landscape of the protein domains, and establish a roadmap to guide emerging large-scale efforts to systematically develop chemical probes for the human proteome, an area of research where Canada is a leader. A longer-term goal will be to expand the analysis to non-human proteomes (including from prokaryotes). Students working on this project will acquire a rare set of expertise in structural bioinformatics, will experience the benefits of adopting open science principles, and will train in a cross-disciplinary, open-concept working environment composed of structural biologists, biochemists, cell biologists. They will also be able to share with and learn from the vibrant local bioinformatics community of downtown Toronto.