Distance-guided protein folding based on generalized descent direction

Abstract

Advances in the prediction of the inter-residue distance for a protein sequence have increased the accuracy to predict the correct folds of proteins with distance information. Here, we propose a distance-guided protein folding algorithm based on generalized descent direction, named GDDfold, which achieves effective structural perturbation and potential minimization in two stages. In the global stage, random-based direction is designed using evolutionary knowledge, which guides conformation population to cross potential barriers and explore conformational space rapidly in a large range. In the local stage, locally rugged potential landscape can be explored with the aid of conjugate-based direction integrated into a specific search strategy, which can improve exploitation ability. GDDfold is tested on 347 proteins of a benchmark set, 24 FM targets of CASP13 and 20 FM targets of CASP14. Results show that GDDfold correctly folds (TM-score ≥ 0.5) 316 out of 347 proteins, where 65 proteins have TM-scores that are greater than 0.8, and significantly outperforms Rosetta-dist (distance-assisted fragment assembly method) and L-BFGSfold (distance geometry optimization method). On CASP FM targets, GDDfold is comparable with five state-of-the-art full-version methods, namely, Quark, RaptorX, Rosetta, MULTICOM and trRosetta in the CASP 13 and 14 server groups.

Construct a variable-length fragment library for de novo protein structure prediction

“…First, the master structure database is constructed by using MMseqs2 [23,24], HHsuite [25], and DSSP [26,27] on the basis of the latest Protein Data Bank (PDB, 2021-08). Then, the HMM-profile of query sequence is constructed and chopped into a nested set of segments (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21) including the clustering of sequences in PDB, the generation of the HMM-profile, and the calculation of secondary structure distribution. The generation of the variable-length fragment library is shown by the orange arrow.…”

“…The generation of the variable-length fragment library is shown by the orange arrow. First, the HMM-profile of the query sequence is generated by HHsuite and chopped into a nested set of segments (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21). Then, the fragments are obtained by profile-profile comparison.…”

“…Previous studies have demonstrated that fragments are essential for fragment assembly-based methods. However, current fragment libraries remain inadequate, and the accuracy of prediction models generated by fragment assembly will be difficult to reach expectations when fragment libraries lack fragments with nearnative fragment structures [21,22]. Designing a variable-length fragment library with increased completeness is crucial for improving the accuracy of prediction models.…”

Improved model quality assessment using sequence and structural information by enhanced deep neural networks

“…Proteins are essential to life, and predicting their structures from sequences can facilitate the understanding of their functions. Various methods have been proposed to predict the three-dimensional structure of proteins [1-6]. The successful application of deep learning in recent years has significantly improved the accuracy of structure prediction [7-14], especially the end-to-end deep learning methods AlphaFold2 [15], whose prediction accuracy on most of the Critical Assessment of techniques for protein Structure Prediction (CASP14) targets is close to experimental results [16].…”

Construct a variable-length fragment library forde novoprotein structure prediction

“…Previous studies have demonstrated that fragments are essential for fragment assembly-based methods. However, current fragment libraries remain inadequate, and the accuracy of prediction models generated by fragment assembly will be difficult to reach expectations when fragment libraries lack fragments with near-native fragment structures [21, 22]. Designing a variable-length fragment library with increased completeness is crucial for improving the accuracy of prediction models.…”