SAAFEC-SEQ: A Sequence-Based Method for Predicting the Effect of Single Point Mutations on Protein Thermodynamic Stability

International Journal of Molecular Sciences, Год журнала: 2021, Номер 22(2), С. 606 - 606

Опубликована: Янв. 9, 2021

Modeling the effect of mutations on protein thermodynamics stability is useful for engineering and understanding molecular mechanisms disease-causing variants. Here, we report a new development SAAFEC method, SAAFEC-SEQ, which gradient boosting decision tree machine learning method to predict change folding free energy caused by amino acid substitutions. The does not require 3D structure corresponding protein, but only its sequence and, thus, can be applied genome-scale investigations where structural information very sparse. SAAFEC-SEQ uses physicochemical properties, features, evolutionary features make predictions. It shown consistently outperform all existing state-of-the-art sequence-based methods in both Pearson correlation coefficient root-mean-squared-error parameters as benchmarked several independent datasets. has been implemented into web server available stand-alone code that downloaded embedded other researchers’ code.

Язык: Английский

---

Computational prediction of the effect of amino acid changes on the binding affinity between SARS-CoV-2 spike RBD and human ACE2

Proceedings of the National Academy of Sciences, Год журнала: 2021, Номер 118(42)

Опубликована: Сен. 29, 2021

The association of the receptor binding domain (RBD) severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein with human angiotensin-converting enzyme (hACE2) represents first required step for cellular entry. SARS-CoV-2 has continued to evolve emergence several novel variants, and amino acid changes in RBD have been implicated increased fitness potential immune evasion. Reliably predicting effect on ability interact more strongly hACE2 can help assess implications public health spillover adaptation into other animals. Here, we introduce a two-step framework that relies 48 independent 4-ns molecular dynamics (MD) trajectories RBD-hACE2 variants collect energy terms decomposed Coulombic, covalent, van der Waals, lipophilic, generalized Born solvation, hydrogen bonding, π-π packing, self-contact correction terms. second implements neural network classify quantitatively predict affinity using as descriptors. computational base achieves validation accuracy 82.8% classifying single-amino substitution worsening or improving correlation coefficient 0.73 between predicted experimentally calculated affinities. Both metrics are fivefold cross-validation test. Our method thus sets up screening caused by unknown single- multiple-amino offering valuable tool host toward tighter binding.

Язык: Английский

---

Cluster learning-assisted directed evolution

Nature Computational Science, Год журнала: 2021, Номер 1(12), С. 809 - 818

Опубликована: Дек. 9, 2021

Язык: Английский

---

Free Energy Perturbation Calculations of Mutation Effects on SARS-CoV-2 RBD::ACE2 Binding Affinity

Journal of Molecular Biology, Год журнала: 2023, Номер 435(15), С. 168187 - 168187

Опубликована: Июнь 23, 2023

The strength of binding between human angiotensin converting enzyme 2 (ACE2) and the receptor domain (RBD) viral spike protein plays a role in transmissibility SARS-CoV-2 virus. In this study we focus on subset RBD mutations that have been frequently observed infected individuals probe affinity changes to ACE2 using surface plasmon resonance (SPR) measurements free energy perturbation (FEP) calculations. Our SPR results are largely accord with previous studies but discrepancies do arise due differences experimental methods protocol even when single method is used. Overall, find FEP performance superior other computational approaches examined as determined by agreement experiment and, particular, its ability identify stabilizing mutations. Moreover, calculations successfully predict cooperative stabilization Q498R N501Y double mutant present Omicron variants offer physical explanation for underlying mechanism. our suggest despite significant cost, may an effective strategy understand effects interfacial protein-protein affinities hence, variety practical applications such optimization neutralizing antibodies.

Язык: Английский

---

Deciphering principles of nucleosome interactions and impact of cancer-associated mutations from comprehensive interaction network analysis

Briefings in Bioinformatics, Год журнала: 2024, Номер 25(2)

Опубликована: Янв. 22, 2024

Abstract Nucleosomes represent hubs in chromatin organization and gene regulation interact with a plethora of factors through different modes. In addition, alterations histone proteins such as cancer mutations post-translational modifications have profound effects on histone/nucleosome interactions. To elucidate the principles interactions those alterations, we developed interactomes for comprehensive mapping histone–histone (HHIs), histone–DNA (HDIs), histone–partner (HPIs) DNA–partner (DPIs) 37 organisms, which contains total 3808 HPIs from 2544 binding 339 HHIs, 100 HDIs 142 DPIs across 110 variants. With networks, explored at levels granularities (protein-, domain- residue-level) performed systematic analysis large scale. Our analyses characterized preferred hotspots both nucleosomal/linker DNA octamer unraveled diverse modes between nucleosome classes partners. Last, to understand impact cancer-associated interactions, complied one mutation dataset including 7940 further mapped onto 419,125 residue level. quantitative point mutations' strongly disruptive HPIs. We predicted 57 recurrent that may driver status oncogenesis.

Язык: Английский

---

Variant Impact Predictor database (VIPdb), version 2: trends from three decades of genetic variant impact predictors

Human Genomics, Год журнала: 2024, Номер 18(1)

Опубликована: Авг. 28, 2024

Abstract Background Variant interpretation is essential for identifying patients’ disease-causing genetic variants amongst the millions detected in their genomes. Hundreds of Impact Predictors (VIPs), also known as Effect (VEPs), have been developed this purpose, with a variety methodologies and goals. To facilitate exploration available VIP options, we created Predictor database (VIPdb). Results The (VIPdb) version 2 presents collection VIPs over past three decades, summarizing characteristics, ClinGen calibrated scores, CAGI assessment results, publication details, access information, citation patterns. We previously summarized 217 features VIPdb 2019. Building upon foundation, identified categorized an additional 190 VIPs, resulting total 407 2. majority capacity to predict impacts single nucleotide nonsynonymous variants. More tailored insertions deletions since 2010s. In contrast, relatively few are dedicated prediction splicing, structural, synonymous, regulatory increasing rate citations reflects ongoing growth use, evolving trends reveal development field individual methods. Conclusions summarizes features, potentially facilitating various variant applications. at https://genomeinterpretation.org/vipdb

Язык: Английский

---

Structure-function relationships of the disease-linked A218T oxytocin receptor variant

Molecular Psychiatry, Год журнала: 2022, Номер 27(2), С. 907 - 917

Опубликована: Янв. 4, 2022

Abstract Various single nucleotide polymorphisms (SNPs) in the oxytocin receptor (OXTR) gene have been associated with behavioral traits, autism spectrum disorder (ASD) and other diseases. The non-synonymous SNP rs4686302 results OXTR variant A218T has linked to core characteristics of ASD, trait empathy preterm birth. However, molecular intracellular mechanisms underlying those associations are still elusive. Here, we uncovered consequences this mutation that may affect psychological or outcome (OXT)-treatment regimens clinical studies, provide a mechanistic explanation for an altered function. We created two monoclonal HEK293 cell lines, stably expressing either wild-type OXTR. detected increased protein stability, accompanied by shift Ca 2+ dynamics reduced MAPK pathway activation cells. Combined whole-genome RNA sequencing analyses OXT-treated cells revealed 7823 differentially regulated genes compared cells, including 429 being ASD. Furthermore, computational modeling provided basis observed change stability suggesting affects downstream events altering signaling, agreement our vitro results. In summary, study provides cellular mechanism links genetic dysregulations aspects

Язык: Английский

---

Robust Prediction of Relative Binding Energies for Protein–Protein Complex Mutations Using Free Energy Perturbation Calculations

Journal of Molecular Biology, Год журнала: 2024, Номер 436(16), С. 168640 - 168640

Опубликована: Июнь 5, 2024

Язык: Английский

---

Comparative molecular dynamics study of the receptor-binding domains in SARS-CoV-2 and SARS-CoV and the effects of mutations on the binding affinity

Journal of Biomolecular Structure and Dynamics, Год журнала: 2020, Номер 40(10), С. 4662 - 4681

Опубликована: Дек. 17, 2020

Here, we report on a computational comparison of the receptor-binding domains (RBDs) spike proteins severe respiratory syndrome coronavirus‐2 (SARS-CoV-2) and SARS-CoV in free forms as complexes with angiotensin-converting enzyme 2 (ACE2) their receptor humans. The impact 42 mutations discovered so far structure thermodynamics SARS-CoV-2 RBD was also assessed. binding affinity for ACE2 is higher than that RBD. COVA2-04 antibody to more energetically favorable COVA2-39, but less formation RBD-ACE2 complex. net charge, dipole moment hydrophilicity are those RBD, producing lower solvation surface energies thus stability. flexible open, larger solvent-accessible area Single-point have dramatic effect distribution charges, most prominently at site substitution its immediate vicinity. These charge alterations alter energy landscape, while X→F exhibit stabilizing through π stacking. F456 W436 emerge two key residues governing stability protein receptor. analyses structural differences different viral strains members coronavirus genera an essential aid development effective therapeutic strategies.

Язык: Английский

---

PremPLI: a machine learning model for predicting the effects of missense mutations on protein-ligand interactions

Communications Biology, Год журнала: 2021, Номер 4(1)

Опубликована: Ноя. 19, 2021

Abstract Resistance to small-molecule drugs is the main cause of failure therapeutic in clinical practice. Missense mutations altering binding ligands proteins are one critical mechanisms that result genetic disease and drug resistance. Computational methods have made a lot progress for predicting affinity changes identifying resistance mutations, but their prediction accuracy speed still not satisfied need be further improved. To address these issues, we introduce structure-based machine learning method quantitatively estimating effects single on ligand (named as PremPLI). A comprehensive comparison predictive performance PremPLI with other available two benchmark datasets confirms our approach performs robustly presents similar or even higher than approaches relying first-principle statistical mechanics mixed physics- knowledge-based potentials while requires much less computational resources. can used guiding design ligand-binding proteins, understanding driver finding potential different drugs. freely at https://lilab.jysw.suda.edu.cn/research/PremPLI/ allows do large-scale mutational scanning.

Язык: Английский

---