Exploring the alternative conformation of a known protein structure based on contact map prediction

Abstract

The rapid development of deep learning-based methods has considerably advanced the field of protein structure prediction. The accuracy of predicting the 3D structures of simple proteins is comparable to that of experimentally determined structures, providing broad possibilities for structure-based biological studies. Another critical question is whether and how multistate structures can be predicted from a given protein sequence. In this study, analysis of multiple two-state proteins demonstrated that deep learning-based contact map predictions contain structural information on both states, which suggests that it is probably appropriate to change the target of deep learningbased protein structure prediction from one specific structure to multiple likely structures. Furthermore, by combining deep learning- and physics-based computational methods, we developed a protocol for exploring alternative conformations from a known structure of a given protein, by which we successfully approached the holo-state conformation of a leucine-binding protein from its apo-state structure.

Recent Advances and Challenges in Protein Structure Prediction

“…Many proteins exist in ensembles of different conformational states. , For example, transporters, G protein-coupled receptors (GPCRs) and enzymes undergo subtle or significant changes in their structures during their functional transitions from one stable state to another. , Elucidating the structures of these states is essential for unraveling its functional mechanism and biological process . Transporters, for example, mediate the vectorial translocation of substrates through movements that open and close extra- and intracellular gates. − Additionally, in drug discovery, certain drug molecules may only bind to a specific state of a protein. − However, most current protein structure prediction algorithms focus on predicting only a specific protein structure, ignoring the structures of different states of the protein.…”

“…In recent years, experimental and computational methods have advanced, significantly increasing in the number of protein structures in PDB. ,, However, the modeling of protein multiple conformational state structures remains challenging for current experimental and computational methods . Some recent attempts have been made to predict alternative structures of proteins, showing some promising results.…”

“…Many proteins exist in ensembles of different conformational states. 88,151 For example, transporters, 152 G protein-coupled receptors (GPCRs) 153 and enzymes 154 undergo subtle or significant changes in their structures during their functional transitions from one stable state to another. 108,151 Elucidating the structures of these states is essential for unraveling its functional mechanism and biological process.…”

“…88,151 For example, transporters, 152 G protein-coupled receptors (GPCRs) 153 and enzymes 154 undergo subtle or significant changes in their structures during their functional transitions from one stable state to another. 108,151 Elucidating the structures of these states is essential for unraveling its functional mechanism and biological process. 108 Transporters, for example, mediate the vectorial translocation of substrates through movements that open and close extra-and intracellular gates.…”

“…Recently, Chen Song et al proposed a method to explore the alternative conformation of a known protein structure based on predicted contact map. 151 In this method, the true contact map of a known structure was subtracted from the predicted contact map to form a difference contact map, and the key contacts from the difference contact map are identified as constraints to model another alternative structure from the known structure. 151 Our group proposed MultiSFold, 166 a method for folding multiple conformational states, which used a distance-based multiobjective evolutionary algorithm to generate multiple conformational states.…”