Multiple minima problem in protein folding solved appying multi criterial optimisation
Roterman, Leszek Konieczny · Computational and Structural Biotechnology Reports · 2025
Despite the progress brought by AlphaFold, the mechanism of protein folding process remains unresolved. Correct structure prediction based on artificial intelligence (AI) does not explain the protein folding mechanism. The fundamental unsolved issue is multiple minima problem (MMP). The original assumption of seeking a global minimum of the energy function expressing all interactions within the protein molecule turns out to be not sufficient. In fact, from the many states representing local energy minima, those that ensure a record of biological activity are selected in nature. The proposed model interprets the protein folding process as a multi-criterial optimisation by considering the dependence of the energy state of the protein on two functions: the internal force field and the external force field. The internal force field means all inter-atoms interactions within the polypeptide chain itself. The external force field expresses the interference of external factors (environment) in the protein folding process. The standard method used for multi-criterial optimisation is a model based on the Pareto front. Its application to the protein folding problem is discussed in this paper.