Protein Structure and Function: Applications of Bioinformatics Methods - John Rigden 2014

Ab Initio Protein Structure Prediction
Conformational Search Methods

The success of ab initio Structure/41.html">Cell/13.html">Protein Structure Prediction depends heavily on an effective conformational search method capable of efficiently locating the global energy minimum on a complex potential energy surface for a given energy function. Historically, the most popular Methods for exploring the conformational space of macromolecules such as Proteins are Monte Carlo (MC) simulation and Molecular Dynamics (MD). Simulating complex systems like proteins using canonical MD/MC methods with the requirement of a thorough conformational space exploration typically demands immense computational resources. Direct Applications of MD to predict native protein structures have yielded few, if any, genuinely successful Examples. One likely reason for these failures is that the simulation time required to fold a small protein is on the order of milliseconds—which is 10¹² times larger than a typical integration step of femtoseconds (10¹⁵). The primary technical difficulty associated with the MC method stems from the fact that the Conformational Energy landscape of a protein is typically rugged, featuring numerous energy barriers that can readily trap and stall the MC simulation Procedure.

This section discusses recent developments in Conformational Search Methods aimed at overcoming the challenges outlined above. We illustrate the Key Concepts of conformational search algorithms utilized in various ab initio and related protein modeling Procedures. For further details, the reader is referred to the cited literature. Unlike the diverse Energy Functions employed in ab initio modeling, search methodologies are, in principle, transferable from one protein modeling approach to another, much like practices in science and technology as a whole. Currently, no single method exhibits exceptional performance across all types of research tasks. Investigating and systematically enhancing the efficiency of various search strategies remain ongoing challenges yet to be fully resolved.