Dock it solves protein4/20/2023 ![]() While the strength of FFT docking is that it allows for an exhaustive search of the docking space, the problem, as we will show, is that the energy function is a limited approximation of the binding affinity, and thus even though the method samples many near-native decoys it often fails to separate them from poor decoys. Classically, this requires a close to correct rigid receptor and ligand, but a set of poses derived from protein fragments with a sequence similar to the peptide can consistently produce conformations near the native bound conformation ( Alam et al., 2017 Zhou et al., 2018). However, since the peptide ligand is a smaller molecule, it is possible to exhaustively sample the binding space by Fast-Fourier Transform docking (FFT docking). Template-based methods utilizing similarity to previously experimentally determined complexes, such as SPOT-Peptide ( Litfin et al., 2019), GalaxyPepDock ( Lee et al., 2015), and InterPep2 ( Johansson-Åkhe et al., 2020a), have consistently shown high performance in previous benchmarks but are limited by available templates. Several methods for predicting the structure of peptide-protein complexes exist, such as pepATTRACT ( Schindler et al., 2015), CABSDOCK ( Kurcinski et al., 2015), HPEPDOCK ( Zhou et al., 2018), and PIPER-FlexPepDock ( Alam et al., 2017). Because of the inherent flexibility of the peptide fragments, computational prediction of the structural details of peptide-protein interaction complexes is challenging. ![]() However, knowledge of structural details such as interactions is crucial to understanding the molecular mechanisms underlying the interactions and to guide further experiments. These short peptides have a high degree of conformational freedom and can be part of larger disordered regions ( Neduva, Victor et al., 2005 Petsalaki and Russell, 2008), making them difficult to study experimentally. Interactions between a short stretch of amino acid residues and a larger protein receptor, referred to as peptide-protein interactions, make up approximately 15–40% of all inter-protein interactions ( Petsalaki and Russell, 2008), and are involved in regulating vital biological processes ( Midic et al., 2009 Tu et al., 2015). The InterPepRank program as well as all scripts for reproducing and retraining it are available from. When included as a selection-method for selecting decoys for refinement in a previously established peptide docking pipeline, InterPepRank improves the number of medium and high quality models produced by 80% and 40%, respectively. This is an improvement compared to other state-of-the-art ranking methods that have a median AUC between 0.65 and 0.79. On this set, InterPepRank has a median AUC of 0.86 for finding coarse peptide-protein complexes with LRMSD < 4Å. InterPepRank is tested on a massive independent test set with no targets sharing CATH annotation nor 30% sequence identity with any target in training or validation data. The graph network is trained to predict the LRMSD of decoys by using edge-conditioned graph convolutions on a large set of peptide-protein complex decoys. ![]() InterPepRank is a machine learning-based method which encodes the structure of the complex as a graph with physical pairwise interactions as edges and evolutionary and sequence features as nodes. We present InterPepRank for peptide-protein complex scoring and ranking. Commonly, methods capable of ranking the decoys for selection fast enough for larger scale studies rely on first-principle energy terms such as electrostatics, Van der Waals forces, or on pre-calculated statistical potentials. A common approach for modeling such interactions is to exhaustively sample the conformational space by fast-Fourier-transform docking, and then refine a top percentage of decoys. Peptide-protein interactions between a smaller or disordered peptide stretch and a folded receptor make up a large part of all protein-protein interactions.
0 Comments
Leave a Reply.AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |