The PLOD2/succinate axis regulates the epithelial-mesenchymal plasticity and cancer cell stemness.

Aberrant accumulation of succinate has been detected in many cancers. However, the cellular function and regulation of succinate in cancer progression is not completely understood. Using stable isotope-resolved metabolomics analysis, we showed that the epithelial mesenchymal transition (EMT) was associated with profound changes in metabolites, including elevation of cytoplasmic succinate levels. The treatment with cell-permeable succinate induced mesenchymal phenotypes in mammary epithelial cells and enhanced cancer cell stemness. Chromatin immunoprecipitation and sequence analysis showed that elevated cytoplasmic succinate levels were sufficient to reduce global 5-hydroxymethylcytosinene (5hmC) accumulation and induce transcriptional repression of EMT-related genes. We showed that expression of procollagen-lysine,2-oxoglutarate 5-dioxygenase 2 (PLOD2) was associated with elevation of cytoplasmic succinate during the EMT process. Silencing of PLOD2 expression in breast cancer cells reduced succinate levels and inhibited cancer cell mesenchymal phenotypes and stemness, which was accompanied by elevated 5hmC levels in chromatin. Importantly, exogenous succinate rescued cancer cell stemness and 5hmC levels in PLOD2-silenced cells, suggesting that PLOD2 promotes cancer progression at least partially through succinate. These results reveal the previously unidentified function of succinate in enhancing cancer cell plasticity and stemness.

Evaluation of sequence-based predictors for phase-separating protein.

Liquid-liquid phase separation (LLPS) of proteins and nucleic acids underlies the formation of biomolecular condensates in cell. Dysregulation of protein LLPS is closely implicated in a range of intractable diseases. A variety of tools for predicting phase-separating proteins (PSPs) have been developed with the increasing experimental data accumulated and several related databases released. Comparing their performance directly can be challenging due to they were built on different algorithms and datasets. In this study, we evaluate eleven available PSPs predictors using negative testing datasets, including folded proteins, the human proteome, and non-PSPs under near physiological conditions, based on our recently updated LLPSDB v2.0 database. Our results show that the new generation predictors FuzDrop, DeePhase and PSPredictor perform better on folded proteins as a negative test set, while LLPhyScore outperforms other tools on the human proteome. However, none of the predictors could accurately identify experimentally verified non-PSPs. Furthermore, the correlation between predicted scores and experimentally measured saturation concentrations of protein A1-LCD and its mutants suggests that, these predictors could not consistently predict the protein LLPS propensity rationally. Further investigation with more diverse sequences for training, as well as considering features such as refined sequence pattern characterization that comprehensively reflects molecular physiochemical interactions, may improve the performance of PSPs prediction.

Peridroplet mitochondria are associated with the severity of MASLD and the prevention of MASLD by diethyldithiocarbamate.

Mitochondria can contact lipid droplets (LDs) to form peridroplet mitochondria (PDM) which trap fatty acids in LDs by providing ATP for triglyceride synthesis and prevent lipotoxicity. However, the role of PDM in metabolic dysfunction associated steatotic liver disease (MASLD) is not clear. Here, the features of PDM in dietary MASLD models with different severity in mice were explored. Electron microscope photographs show that LDs and mitochondria rarely come into contact with each other in normal liver. In mice fed with high-fat diet, PDM can be observed in the liver as early as the beginning of steatosis in hepatocytes. For the first time, we show that PDM in mouse liver varies with the severity of MASLD. PDM and cytosolic mitochondria were isolated from the liver tissue of MASLD and analyzed by quantitative proteomics. Compared with cytosolic mitochondria, PDM have enhanced mitochondrial respiration and ATP synthesis. Diethyldithiocarbamate (DDC) alleviates choline-deficient, L-amino acid-defined diet-induced MASLD, while increases PDM in the liver. Similarly, DDC promotes the contact of mitochondria-LDs in steatotic C3A cells in vitro. Meanwhile, DDC promotes triglyceride synthesis and improves mitochondrial dysfunction in MASLD. In addition, DDC upregulates perilipin 5 both in vivo and in vitro, which is considered as a key regulator in PDM formation. Knockout of perilipin 5 inhibits the contact of mitochondria-LDs induced by DDC in C3A cells. These results demonstrate that PDM might be associated with the progression of MASLD and the prevention of MASLD by DDC.

Integrative analysis of single-cell embryo data reveals transcriptome signatures for the human pre-implantation inner cell mass.

As the source of embryonic stem cells (ESCs), inner cell mass (ICM) can form all tissues of the embryo proper, however, its role in early human lineage specification remains controversial. Although a stepwise differentiation model has been proposed suggesting the existence of ICM as a distinct developmental stage, the underlying molecular mechanism remains unclear. In the present study, we perform an integrated analysis on the public human preimplantation embryonic single-cell transcriptomic data and apply a trajectory inference algorithm to measure the cell plasticity. In our results, ICM population can be clearly discriminated on the dimension-reduced graph and confirmed by compelling evidences, thus validating the two-step hypothesis of lineage commitment. According to the branch probabilities and differentiation potential, we determine the precise time points for two lineage segregations. Further analysis on gene expression dynamics and regulatory network indicates that transcription factors including GSC, PRDM1, and SPIC may underlie the decisions of ICM fate. In addition, new human ICM marker genes, such as EPHA4 and CCR8 are discovered and validated by immunofluorescence. Given the potential clinical applications of ESCs, our analysis provides a further understanding of human ICM cells and facilitates the exploration of more unique characteristics in early human development.

PLANET: A Multi-objective Graph Neural Network Model for Protein-Ligand Binding Affinity Prediction.

Predicting protein-ligand binding affinity is a central issue in drug design. Various deep learning models have been published in recent years, where many of them rely on 3D protein-ligand complex structures as input and tend to focus on the single task of reproducing binding affinity. In this study, we have developed a graph neural network model called PLANET (Protein-Ligand Affinity prediction NETwork). This model takes the graph-represented 3D structure of the binding pocket on the target protein and the 2D chemical structure of the ligand molecule as input. It was trained through a multi-objective process with three related tasks, including deriving the protein-ligand binding affinity, protein-ligand contact map, and ligand distance matrix. Besides the protein-ligand complexes with known binding affinity data retrieved from the PDBbind database, a large number of non-binder decoys were also added to the training data for deriving the final model of PLANET. When tested on the CASF-2016 benchmark, PLANET exhibited a scoring power comparable to the best result yielded by other deep learning models as well as a reasonable ranking power and docking power. In virtual screening trials conducted on the DUD-E benchmark, PLANET's performance was notably better than several deep learning and machine learning models. As on the LIT-PCBA benchmark, PLANET achieved comparable accuracy as the conventional docking program Glide, but it only spent less than 1% of Glide's computation time to finish the same job because PLANET did not need exhaustive conformational sampling. Considering the decent accuracy and efficiency of PLANET in binding affinity prediction, it may become a useful tool for conducting large-scale virtual screening.

In-depth analysis reveals complex molecular aetiology in a cohort of idiopathic cerebral palsy.

Cerebral palsy is the most prevalent physical disability in children; however, its inherent molecular mechanisms remain unclear. In the present study, we performed in-depth clinical and molecular analysis on 120 idiopathic cerebral palsy families, and identified underlying detrimental genetic variants in 45% of these patients. In addition to germline variants, we found disease-related postzygotic mutations in ∼6.7% of cerebral palsy patients. We found that patients with more severe motor impairments or a comorbidity of intellectual disability had a significantly higher chance of harbouring disease-related variants. By a compilation of 114 known cerebral-palsy-related genes, we identified characteristic features in terms of inheritance and function, from which we proposed a dichotomous classification system according to the expression patterns of these genes and associated cognitive impairments. In two patients with both cerebral palsy and intellectual disability, we revealed that the defective TYW1, a tRNA hypermodification enzyme, caused primary microcephaly and problems in motion and cognition by hindering neuronal proliferation and migration. Furthermore, we developed an algorithm and demonstrated in mouse brains that this malfunctioning hypermodification specifically perturbed the translation of a subset of proteins involved in cell cycling. This finding provided a novel and interesting mechanism for congenital microcephaly. In another cerebral palsy patient with normal intelligence, we identified a mitochondrial enzyme GPAM, the hypomorphic form of which led to hypomyelination of the corticospinal tract in both human and mouse models. In addition, we confirmed that the aberrant Gpam in mice perturbed the lipid metabolism in astrocytes, resulting in suppressed astrocytic proliferation and a shortage of lipid contents supplied for oligodendrocytic myelination. Taken together, our findings elucidate novel aspects of the aetiology of cerebral palsy and provide insights for future therapeutic strategies.

Binding of berberine derivates to G-quadruplex: insight from a computational study.

Human telomerase exhibits significant activity in cancer cells relative to normal cells, which contributes to the immortal proliferation of cancer cells. To counter this, the stabilization of G-quadruplexes formed in the guanine-rich sequence of the cancer cell chromosome has emerged as a promising avenue for anti-cancer therapy. Berberine (BER), an alkaloid that is derived from traditional Chinese medicines, has shown potential for stabilizing G-quadruplexes. To investigate the atomic interactions between G-quadruplexes and BER and its derivatives, molecular dynamics simulations were conducted. Modeling the interactions between G-quadruplexes and ligands accurately is challenging due to the strong negative charge of nucleic acids. Thus, various force fields and charge models for the G-quadruplex and ligands were tested to obtain precise simulation results. The binding energies were calculated by a combination of molecular mechanics/generalized Born surface area and interaction entropy methods, and the calculated results correlated well with experimental results. B-factor and hydrogen bond analyses demonstrated that the G-quadruplex was more stable in the presence of ligands than in the absence of ligands. Calculation of the binding free energy showed that the BER derivatives bind to a G-quadruplex with higher affinity than that of BER. The breakdown of the binding free energy to per-nucleotide energies suggested that the first G-tetrad played a primary role in binding. Additionally, energy and geometric properties analyses indicated that van der Waals interactions were the most favorable interactions between the derivatives and the G-quadruplexes. Overall, these findings provide crucial atomic-level insights into the binding of G-quadruplexes and their inhibitors.

Modular networks and genomic variation during progression from stable angina pectoris through ischemic cardiomyopathy to chronic heart failure.

BACKGROUND: Analyzing disease-disease relationships plays an important role for understanding etiology, disease classification, and drug repositioning. However, as cardiovascular diseases with causative links, the molecular relationship among stable angina pectoris (SAP), ischemic cardiomyopathy (ICM) and chronic heart failure (CHF) is not clear. METHODS: In this study, by integrating the multi-database data, we constructed paired disease progression modules (PDPMs) to identified relationship among SAP, ICM and CHF based on module reconstruction pairs (MRPs) of K-value calculation (a Euclidean distance optimization by integrating module topology parameters and their weights) methods. Finally, enrichment analysis, literature validation and structural variation (SV) were performed to verify the relationship between the three diseases in PDPMs. RESULTS: Total 16 PDPMs were found with K > 0.3777 among SAP, ICM and CHF, in which 6 pairs in SAP-ICM, 5 pairs for both ICM-CHF and SAP-CHF. SAP-ICM was the most closely related by having the smallest average K-value (K = 0.3899) while the maximum is SAP-CHF (K = 0.4006). According to the function of the validation gene, inflammatory response were through each stage of SAP-ICM-CHF, while SAP-ICM was uniquely involved in fibrosis, and genes were related in affecting the upstream of PI3K-Akt signaling pathway. 4 of the 11 genes (FLT1, KDR, ANGPT2 and PGF) in SAP-ICM-CHF related to angiogenesis in HIF-1 signaling pathway. Furthermore, we identified 62.96% SVs were protein deletion in SAP-ICM-CHF, and 53.85% SVs were defined as protein replication in SAP-ICM, while ICM-CHF genes were mainly affected by protein deletion. CONCLUSION: The PDPMs analysis approach combined with genomic structural variation provides a new avenue for determining target associations contributing to disease progression and reveals that inflammation and angiogenesis may be important links among SAP, ICM and CHF progression.

Impact of feedback bandwidth on Raman random fiber laser remote-sensing.

In the ultra-long distance sensing domain, recently Raman random fiber laser (RRFL) demonstrated advantages of ultrawide sensing-bandwidth in dynamic sensing, compared with pulse-probing cases. However, such a scheme is still in the preliminary stage, and the key parameters such as sensitivity have not been characterized. In this work, a time-dependent spectrum-balanced model is proposed, which can accurately and quickly describe the spectral shape of RRFL and the evolution of the power and the spectrum. Based on this model, the relationship between the sensitivity and the feedback bandwidth is studied. The calculated results show that the sensitivity is inversely proportional to the feedback bandwidth. Then in the proof-of-concept experiment, by changing the bandwidth of sensing FBG, the results of sensitivity are well coincident with the simulation. This work provides an effective platform for studying the evolution of RRFL spectrum, as well as a novel way for further enhancing the performance of the dynamic sensing system based on ultra-long RRFL.

HergSPred: Accurate Classification of hERG Blockers/Nonblockers with Machine-Learning Models.

The human ether-à-go-go-related gene (hERG) K+ channel plays an important role in cardiac action potentials. The inhibition of the hERG channel may lead to long QT syndrome (LQTS) and even sudden cardiac death. Due to severe hERG-related cardiotoxicity, many drugs have been withdrawn from the market. Therefore, it is necessary to estimate the chemical blockade of hERG in the early stage of drug discovery. In this study, we collected 12,850 compounds with hERG inhibition data from the literature and trained a series of hERG blocking classification models based on the MACCS and Morgan fingerprints. A consensus model named HergSPred was generated based on the individual models using voting principles. The accuracy of HergSPred is higher than previous models using identical training and test sets. Moreover, we analyzed the contribution of each input fingerprint to the prediction output to obtain intuitive chemical insights into the hERG inhibition, which allows visualization of warning substructures that may cause cardiotoxicity in the input compound. The model is available at http://www.icdrug.com/ICDrug/T.

A fast-slow method to treat solute dynamics in explicit solvent.

Aiming to reduce the computational cost in the current explicit solvent molecular dynamics (MD) simulation, this paper proposes a fast-slow method for the fast MD simulation of biomolecules in explicit solvent. This fast-slow method divides the entire system into two parts: a core layer (typically solute or biomolecule) and a peripheral layer (typically solvent molecules). The core layer is treated using standard MD method but the peripheral layer is treated by a slower dynamics method to reduce the computational cost. We compared four different simulation models in testing calculations for several small proteins. These include gas-phase, implicit solvent, fast-slow explicit solvent and standard explicit solvent MD simulations. Our study shows that gas-phase and implicit solvent models do not provide a realistic solvent environment and fail to correctly produce reliable dynamic structures of proteins. On the other hand, the fast-slow method can essentially reproduce the same solvent effect as the standard explicit solvent model while gaining an order of magnitude in efficiency. This fast-slow method thus provides an efficient approach for accelerating the MD simulation of biomolecules in explicit solvent.

Generation of non-iterative phase-only hologram based on a hybrid phase mask.

The random phase method and quadratic phase method are most widely used in the generation of non-iterative phase holograms. However, the former leads to the reconstruction being severely disturbed by speckle noise, with serious loss of detailed information, and the latter leads to the reconstruction being contaminated with ringing artifacts. To solve these problems, we present a novel, to the best of our knowledge, method capable of generating non-iterative phase holograms, hereafter referred to as hybrid-phase-only holograms (HPOHs). Our proposal is to use a weight factor to combine the random phase and quadratic phase to generate a hybrid phase mask. The hybrid phase mask is then superimposed on the target image to obtain a complex hologram by simple Fourier transform. Followed by retaining the phase of the complex hologram, we can generate the corresponding HPOH. The effects of different weight factors on the holographic reconstructions are discussed. Numerical simulations of reconstruction quality associated with the proposed method, random phase method, and quadratic phase method are presented for comparison purposes. Optical experiments based on liquid crystal on silicon also demonstrate the validity of the method.

LLPSDB: a database of proteins undergoing liquid-liquid phase separation in vitro.

Liquid-liquid phase separation (LLPS) leads to a conversion of homogeneous solution into a dense phase that often resembles liquid droplets, and a dilute phase. An increasing number of investigations have shown that biomolecular condensates formed by LLPS play important roles in both physiology and pathology. It has been suggested the phase behavior of proteins would be not only determined by sequences, but controlled by micro-environmental conditions. Here, we introduce LLPSDB (http://bio-comp.ucas.ac.cn/llpsdb or http://bio-comp.org.cn/llpsdb), a web-accessible database providing comprehensive, carefully curated collection of proteins involved in LLPS as well as corresponding experimental conditions in vitro from published literatures. The current release of LLPSDB incorporates 1182 entries with 273 independent proteins and 2394 specific conditions. The database provides a variety of data including biomolecular information (protein sequence, protein modification, nucleic acid, etc.), specific phase separation information (experimental conditions, phase behavior description, etc.) and comprehensive annotations. To our knowledge, LLPSDB is the first available database designed for LLPS related proteins specifically. It offers plenty of valuable resources for exploring the relationship between protein sequence and phase behavior, and will enhance the development of phase separation prediction methods, which may further provide more insights into a comprehensive understanding of LLPS in cellular function and related diseases.

CHARMM-GUI Supports Hydrogen Mass Repartitioning and Different Protonation States of Phosphates in Lipopolysaccharides.

Hydrogen mass repartitioning (HMR) that permits time steps of all-atom molecular dynamics simulation up to 4 fs by increasing the mass of hydrogen atoms has been used in protein and phospholipid bilayers simulations to improve conformational sampling. Molecular simulation input generation via CHARMM-GUI now supports HMR for diverse simulation programs. In addition, considering ambiguous pH at the bacterial outer membrane surface, different protonation states, either -2e or -1e, of phosphate groups in lipopolysaccharides (LPS) are also supported in CHARMM-GUI LPS Modeler. To examine the robustness of HMR and the influence of protonation states of phosphate groups on LPS bilayer properties, eight different LPS-type all-atom systems with two phosphate protonation states are modeled and simulated utilizing both OpenMM 2-fs (standard) and 4-fs (HMR) schemes. Consistency in the conformational space sampled by standard and HMR simulations shows the reliability of HMR even in LPS, one of the most complex biomolecules. For systems with different protonation states, similar conformations are sampled with a PO41- or PO42- group, but different phosphate protonation states make slight impacts on lipid packing and conformational properties of LPS acyl chains. Systems with PO41- have a slightly smaller area per lipid and thus slightly more ordered lipid A acyl chains compared to those with PO42-, due to more electrostatic repulsion between PO42- even with neutralizing Ca2+ ions. HMR and different protonation states of phosphates of LPS available in CHARMM-GUI are expected to be useful for further investigations of biological systems of diverse origin.

CHARMM-GUI DEER facilitator for spin-pair distance distribution calculations and preparation of restrained-ensemble molecular dynamics simulations.

The double electron-electron resonance (DEER) is a powerful structural biology technique to obtain distance information in the range of 18 to 80 å by measuring the dipolar coupling between two unpaired electron spins. The distance distributions obtained from the experiment provide valuable structural information about the protein in its native environment that can be exploited using restrained ensemble molecular dynamics (reMD) simulations. We present a new tool DEER Facilitator in CHARMM-GUI that consists of two modules Spin-Pair Distributor and reMD Prepper to setup simulations that utilize information from DEER experiments. Spin-Pair Distributor provides a web-based interface to calculate the spin-pair distance distribution of labeled sites in a protein using MD simulations. The calculated distribution can be used to guide the selection of the labeling sites in experiments as well as validate different protein structure models. reMD Prepper facilities the setup of reMD simulations using different types of spin labels in four different environments including vacuum, solution, micelle, and bilayer. The applications of these two modules are demonstrated with several test cases. Spin-Pair Distributor and reMD Prepper are available at http://www.charmm-gui.org/input/deer and http://www.charmm-gui.org/input/deerre. DEER Facilitator is expected to facilitate advanced biomolecular modeling and simulation, thereby leading to an improved understanding of the structure and dynamics of complex biomolecular systems based on experimental DEER data. © 2019 Wiley Periodicals, Inc.

Unfolding of a ClC chloride transporter retains memory of its evolutionary history.

ClC chloride channels and transporters are important for chloride homeostasis in species from bacteria to human. Mutations in ClC proteins cause genetically inherited diseases, some of which are likely to involve folding defects. The ClC proteins present a challenging and unusual biological folding problem because they are large membrane proteins possessing a complex architecture, with many reentrant helices that go only partway through membrane and loop back out. Here we were able to examine the unfolding of the Escherichia coli ClC transporter, ClC-ec1, using single-molecule forced unfolding methods. We found that the protein could be separated into two stable halves that unfolded independently. The independence of the two domains is consistent with an evolutionary model in which the two halves arose from independently folding subunits that later fused together. Maintaining smaller folding domains of lesser complexity within large membrane proteins may be an advantageous strategy to avoid misfolding traps.

Binding Modes of Small-Molecule Inhibitors to the EED Pocket of PRC2.

Polycomb Polycomb repressive complex 2 (PRC2) plays a key role in silencing epigenetic gene through trimethylation of lysine 27 on histone 3 (H3K27). Dysregulations of PRC2 caused by overexpression and mutations of the core subunits of PRC2 have been implicated in many cancers. The core subunits EZH1/2 are histone-lysine N-methyltransferases that function as the enzymatic component of PRC2. While the core subunit EED is a scaffolding protein to support EZH1/2 and binds JARID2K116me3/H3K27me3 to enhance the enzymatic activity of PRC2 through allosteric activation. Recently, several small molecules that compete with JARI2K116me3 and H3K27me3 have been reported. These molecules selectively bind to the JARID2K116me3/H3K27me3-binding pocket of EED, thereby preventing the allosteric regulation of PRC2. These first-in-class PRC2 inhibitors show robust suppression in DLBCL cell lines, demonstrating anticancer drugs that target the EED subunit of PRC2 are viable. In this study, we used the recently developed MM/GBSA_IE and the alanine scanning method to analyze the hot spots in EED/inhibitor interactions. The analysis of these hot and warm spots helps us to understand the fundamental differences between inhibitors. Our results give a quantitative explanation on why the binding affinities of EED/A-395 interactions are stronger than that of EED/EED226 while their binding modes are similar and provide valuable insights for rational design of novel EED inhibitors.

DenseCPD: Improving the Accuracy of Neural-Network-Based Computational Protein Sequence Design with DenseNet.

Computational protein design remains a challenging task despite its remarkable success in the past few decades. With the rapid progress of deep-learning techniques and the accumulation of three-dimensional protein structures, the use of deep neural networks to learn the relationship between protein sequences and structures and then automatically design a protein sequence for a given protein backbone structure is becoming increasingly feasible. In this study, we developed a deep neural network named DenseCPD that considers the three-dimensional density distribution of protein backbone atoms and predicts the probability of 20 natural amino acids for each residue in a protein. The accuracy of DenseCPD was 53.24 ± 0.17% in a 5-fold cross-validation on the training set and 55.53% and 50.71% on two independent test sets, which is more than 10% higher than those of previous state-of-the-art methods. Two approaches for using DenseCPD predictions in computational protein design were analyzed. The approach using the cutoff of accumulative probability had a smaller sequence search space compared with the approach that simply uses the top-k predictions and therefore enabled higher sequence identity in redesigning three proteins with Rosetta. The network and the datasets are available on a web server at http://protein.org.cn/densecpd.html. The results of this study may benefit the further development of computational protein design methods.

Pluripotent anti-inflammatory immunomodulatory effects of papaverine against cerebral ischemic-reperfusion injury.

The mechanism of the papaverine (PV) for the treatment of cerebral ischemia remains unclear. A total of 42 mice induced with focal cerebral ischemia were randomly divided into three groups: PV,baicalin (BA)and vehicle group. Both PV and BA could significantly reduce the ischemic infarct volume (P < 0.05). Pathway enrichment analysis was performed on MetaCore™ to search the molecular pathways associated with the gene expression profile of PV, compared with vehicle and BA. Compared with vehicle, we found that 60% of the top 10 pathways in PV group were related to immune response. The top 10 biological processes of the targeted pathways were mainly related to the multiple immunomodulatory process of neuro-vascular inflammation, including immune_Th17-deried cytokins, regulation of angiogenesis, cell adhesion_Leucocyte chemotaxis, antigen presentation, cell adhesion_synaptic contact, and inflammation related to Amphoterin signaling. Moreover, compared with BA, 17 pathways of PV were identified, and 58.82% (10/17) were also related to immune response, especially, half of the top 10 pathways with the lower p-value. In these top 10 pathways, 4 were the cytokine-mediated signaling pathways, which play key role in inflammation, like IL-17 signaling pathways with the activation of G-CSF,IL-23,RANKL, p38MAPK and NF-κB.These findings indicate that PV may be an efficacious pluripotent anti-inflammatory agent against cerebral ischemic-reperfusion injury by targeting on multiple immunomodulatory process of neuro-vascular inflammation.

A method for efficient calculation of thermal stability of proteins upon point mutations.

A method for efficient prediction of the relative stability of a protein due to a single amino acid point mutation is presented. In this approach, we calculate the free energy change due to an arbitrary point mutation of a protein from a single MD trajectory of the wild type protein. The method is tested on 27 diverse protein systems with a total of 853 mutations and the calculated relative free energies show a generally good correlation with the experimental values (a correlation coefficient of 0.63). Comparison with the free energy perturbation (FEP) method and the recently developed machine learning methods on two different benchmark data sets shows that the current method is computationally efficient and also numerically reliable for predicting the changes in thermostability upon an arbitrary point mutation of a protein. A discussion is provided on how to further improve the accuracy of the method for the prediction of thermostability of proteins.