PASSer: fast and accurate prediction of protein allosteric sites

Abstract

Allostery refers to the biological process by which an effector modulator binds to a protein at a site distant from the active site, known as allosteric site. Identifying allosteric sites is essential for discovering allosteric process and is considered a critical factor in allosteric drug development. To facilitate related research, we developed PASSer (Protein Allosteric Sites Server) at https://passer.smu.edu, a web application for fast and accurate allosteric site prediction and visualization. The website hosts three trained and published machine learning models: (i) an ensemble learning model with extreme gradient boosting and graph convolutional neural network, (ii) an automated machine learning model with AutoGluon and (iii) a learning-to-rank model with LambdaMART. PASSer accepts protein entries directly from the Protein Data Bank (PDB) or user-uploaded PDB files, and can conduct predictions within seconds. The results are presented in an interactive window that displays protein and pockets’ structures, as well as a table that summarizes predictions of the top three pockets with the highest probabilities/scores. To date, PASSer has been visited over 49 000 times in over 70 countries and has executed over 6 200 jobs.

Markov State Models and Perturbation-Based Approaches Reveal Distinct Dynamic Signatures and Hidden Allosteric Pockets in the Emerging SARS-Cov-2 Spike Omicron Variant Complexes with the Host Receptor: The Interplay of Dynamics and Convergent Evolution Modulates Allostery and Functional Mechanisms

“…The reversed allosteric communication approach is based on the premise that allosteric signaling in proteins is bidirectional and can propagate from an allosteric to orthosteric site and vice versa. ,− A more integrated computational and experimental strategy exploited the reversed allosteric communication concepts to combine MD simulations and MSM analysis to monitor shifts in the protein conformational ensembles and detect cryptic allosteric sites. , We have recently developed a fast and accurate allosteric site prediction method PASSer, and here, this approach was employed for detection of allosteric RBD pockets using information about conformational ensembles and major functional macrostates of the Omicron RBD-ACE2 complexes. To simplify the presentation of the results, we focused on the predicted RBD pockets for the optimized structural models of XBB and BQ variants (Supporting Materials, Figures S9,S10) as well as on a detailed analysis of the determined cryptic pockets that are specific for each of the determined macrostates (Figure ).…”

“…86 The PASSer web server has been widely used for the validation of known functional pockets and the discovery of new allosteric sites. 87 Here, we used the LTR model capable of ranking binding pockets and identifying the most probable allosteric binding sites. 86 The LTR model in PASSer was trained and validated on two widely used data sets, the Allosteric Database (ASD) 126,127 which is a comprehensive database of allosteric proteins and modulators and CASBench which is a benchmarking set that includes annotated catalytic and allosteric sites.…”

“…Machine learning can help detect and predict allosteric sites in proteins since it can deal with numerous input features, including local or static features of pockets and delocalized or dynamic features of proteins . In the current study, we employed the recently developed PASSer approach that provides fast and accurate predictions of allosteric sites in proteins. − PASSer offers three pretrained machine learning-based models: (a) an ensemble learning model consisting of extreme gradient boosting (XGBoost) and a graph convolutional neural network (GCNN), (b) an automated machine learning model powered by AutoGluon from Amazon Web Services (AWS), and (c) a learning-to-rank (LTR) model . The PASSer web server has been widely used for the validation of known functional pockets and the discovery of new allosteric sites …”

“…Our findings suggest a mechanism in which convergent Omicron mutations can promote high transmissibility and antigenicity of the virus by controlling the interplay between the RBD stability and conformational adaptability, allowing for optimal fitness tradeoffs between binding to the host receptor and robust immune evasion profile. We also explore our method for fast and accurate allosteric site prediction PASSer (Protein Allosteric Sites Server) () − to characterize the distribution allosteric sites in the Omicron complexes and detect previously unknown hidden allosteric pockets present in the stable RBD conformations. Through integrative computational approaches, this study provides a systematic analysis and comparison of the effects of Omicron subvariants on conformational dynamics and allosteric signaling in the complexes with the ACE2 receptor.…”

PASSerRank: Prediction of allosteric sites with learning to rank

“…Another ML‐based approach, the three‐way random forest (RF) model developed by Chen et al, 14 is capable of predicting allosteric, regular, or orthosteric sites. PASSer 15‐17 is a recently developed method that combines extreme gradient boosting (XGBoost) 18 with a graph convolutional neural network 19 to learn physical and topological properties without any prior information. In addition to ML, traditional methods such as normal mode analysis 20 and molecular dynamics 21 are widely used to investigate the communication between regulatory and functional sites, including SPACER 22 and PARS 23 …”