Protein complexes play deterministic roles in live entities in sensing, compiling, controlling, and responding to external and internal stimuli. Thermodynamic stability is an important property of protein complexes; having knowledge about complex stability helps us to understand the basics of protein assembly-related diseases and the mechanism of protein assembly clearly. Enormous protein–protein interactions, detected by high-throughput methods, necessitate finding fast methods for predicting the stability of protein assemblies in a quantitative and qualitative manner. The existing methods of predicting complex stability need knowledge about the three-dimensional (3D) structure of the intended protein complex. Here, we introduce a new method for predicting dissociation free energy of subunits by analyzing the structural and topological properties of a protein binding patch on a single subunit of the desired protein complex. The method needs the 3D structure of just one subunit and the information about the position of the intended binding site on the surface of that subunit to predict dimer stability in a classwise manner. The patterns of structural and topological properties of a protein binding patch are decoded by recurrence quantification analysis. Nonparametric discrimination is then utilized to predict the stability class of the intended dimer with accuracy greater than 85%.

中文翻译：

---

从单体的未结合结构对蛋白质-二聚体稳定性的伴侣特异性预测。

蛋白质复合物在活体中在确定，编译，控制和响应外部和内部刺激中起决定性作用。热力学稳定性是蛋白质复合物的重要特性。具有复杂稳定性的知识有助于我们清楚地了解蛋白质组装相关疾病的基础知识以及蛋白质组装的机理。通过高通量方法检测到巨大的蛋白质间相互作用，必须找到快速的方法以定量和定性的方式预测蛋白质装配的稳定性。预测复合物稳定性的现有方法需要有关预期蛋白质复合物的三维（3D）结构的知识。这里，我们介绍了一种通过分析所需蛋白复合物的单个亚基上的蛋白结合补丁的结构和拓扑特性来预测亚基解离自由能的新方法。该方法仅需要一个亚基的3D结构，以及有关该亚基表面上预期结合位点位置的信息，就可以以分类方式预测二聚体的稳定性。蛋白质结合斑块的结构和拓扑特性的模式通过重复定量分析进行解码。然后，利用非参数判别法来预测目标二聚体的稳定性等级，其准确度大于85％。该方法仅需要一个亚基的3D结构，以及有关该亚基表面上预期结合位点位置的信息，就可以以分类方式预测二聚体的稳定性。蛋白质结合斑块的结构和拓扑特性的模式通过重复定量分析进行解码。然后，利用非参数判别法来预测目标二聚体的稳定性等级，其准确度大于85％。该方法仅需要一个亚基的3D结构，以及有关该亚基表面上预期结合位点位置的信息，就可以以分类方式预测二聚体的稳定性。蛋白质结合斑块的结构和拓扑特性的模式通过重复定量分析进行解码。然后，利用非参数判别法来预测目标二聚体的稳定性等级，其准确度大于85％。