Folding Stability Research Articles

COPII cage assembly factor Sec13 integrates information flow regulating endomembrane function in response to human variation

How information flow is coordinated for managing transit of 1/3 of the genome through endomembrane pathways by the coat complex II (COPII) system in response to human variation remains an enigma. By examining the interactome of the COPII cage-assembly component Sec13, we show that it is simultaneously associated with multiple protein complexes that facilitate different features of a continuous program of chromatin organization, transcription, translation, trafficking, and degradation steps that are differentially sensitive to Sec13 levels. For the trafficking step, and unlike other COPII components, reduction of Sec13 expression decreased the ubiquitination and degradation of wild-type (WT) and F508del variant cargo protein cystic fibrosis transmembrane conductance regulator (CFTR) leading to a striking increase in fold stability suggesting that the events differentiating export from degradation are critically dependent on COPII cage assembly at the ER Golgi intermediate compartment (ERGIC) associated recycling and degradation step linked to COPI exchange. Given Sec13’s multiple roles in protein complex assemblies that change in response to its expression, we suggest that Sec13 serves as an unanticipated master regulator coordinating information flow from the genome to the proteome to facilitate spatial covariant features initiating and maintaining design and function of membrane architecture in response to human variation.

Disentanglement of Evolutionary Constraints in Statistical Models of Proteins

The exponential growth of protein sequences in the post-genomic era has revolutionized the application of generative sequence models for pivotal tasks such as contact prediction, protein design, alignment, and homology search. Despite remarkable progress in these areas, the interpretability of the modeled pairwise parameters remains limited due to complexities arising from coevolution, phylogeny, and entropy. While post-correction methods for contact prediction have been developed to eliminate entropy-related contributions from predicted contact maps, there is currently no direct approach to correct entropy in other applications reliant on raw parameters. In this paper, we investigate the sources of entropy signal and propose a novel spectral regularizer, LH (an abbreviation of Henri Lebesgue), to mitigate its impact during model fitting. By incorporating this regularizer into the GREMLIN framework (utilizing a Markov random field or Potts model), we enable the accurate inference of sparse contact maps while simultaneously improving interpretability and addressing overfitting concerns critical for sequence evaluation and design. To validate the efficacy of our approach, we design multiple protein sequences based on GREMLIN with both L2 and LH regularizers, and subsequently experimentally measure their using cDNA display proteolysis. Our findings demonstrate that proteins designed using the LH regularizer exhibit increased diversity and enhanced folding stability. Published by the American Physical Society 2024

Comparative analysis and classification of highly divergent mouse rDNA units based on their intergenic spacer (IGS) variability.

Ribosomal DNA (rDNA) repeat units are organized into tandem clusters in eukaryotic cells. In mice, these clusters are located on at least eight chromosomes and show extensive variation in the number of repeats between mouse genomes. To analyze intra- and inter-genomic variation of mouse rDNA repeats, we selectively isolated 25 individual rDNA units using Transformation-Associated Recombination (TAR) cloning. Long-read sequencing and subsequent comparative sequence analysis revealed that each full-length unit comprises an intergenic spacer (IGS) and a ∼13.4 kb long transcribed region encoding the three rRNAs, but with substantial variability in rDNA unit size, ranging from ∼35to ∼46 kb. Within the transcribed regions of rDNA units, we found 209 variants, 70 of which are in external transcribed spacers (ETSs); but the rDNA size differences are driven primarily by IGS size heterogeneity, due to indels containing repetitive elements and some functional signals such as enhancers. Further evolutionary analysis categorized rDNA units into distinct clusters with characteristic IGS lengths; numbers of enhancers; and presence/absence of two common SNPs in promoter regions, one of which is located within promoter (p)RNA and may influence pRNA folding stability. These characteristic features of IGSs also correlated significantly with 5'ETS variant patterns described previously and associated with differential expression of rDNA units. Our results suggest that variant rDNA units are differentially regulated and open a route to investigate the role of rDNA variation on nucleolar formation and possible associations with pathology.

High-throughput Selection of Human de novo-emerged sORFs with High Folding Potential.

De novo genes emerge from previously noncoding stretches of the genome. Their encoded de novo proteins are generally expected to be similar to random sequences and, accordingly, with no stable tertiary fold and high predicted disorder. However, structural properties of de novo proteins and whether they differ during the stages of emergence and fixation have not been studied in depth and rely heavily on predictions. Here we generated a library of short human putative de novo proteins of varying lengths and ages and sorted the candidates according to their structural compactness and disorder propensity. Using Förster resonance energy transfer combined with Fluorescence-activated cell sorting, we were able to screen the library for most compact protein structures, as well as most elongated and flexible structures. We find that compact de novo proteins are on average slightly shorter and contain lower predicted disorder than less compact ones. The predicted structures for most and least compact de novo proteins correspond to expectations in that they contain more secondary structure content or higher disorder content, respectively. Our experiments indicate that older de novo proteins have higher compactness and structural propensity compared with young ones. We discuss possible evolutionary scenarios and their implications underlying the age-dependencies of compactness and structural content of putative de novo proteins.

An odorant-binding protein based electrical sensor to detect volatile organic compounds

Artificial olfaction systems, such as electronic nose devices (ENs), can be used in various fields, from healthcare to food industry and the environmental sector. ENs consist of an array of sensors composed of different sensing materials. Incorporating Odorant Binding Proteins (OBPs) into gas-sensing materials can increase volatile organic compounds (VOCs) recognition and selectivity. OBPs are small soluble proteins found in vertebrates and insects, responsible for VOCs solubilization and transportation. In this work, OBP3 from Rattus norvegicus was selected to develop an OBP-based electrical VOC sensor, by optimizing protein production and immobilization on sensors surface. OBP3 was successfully produced in E. coli host cells (94 mg/L) and purified with high purity (88% purification yield, 96% purity). The protein folding and thermal stability were assessed by circular dichroism (Tm=71±1ºC) whereas ligand binding activity was verified in solution by fluorescence displacement against diisopropylphenol (Kd=0.24 µM). For the immobilization of OBP3 on gold interdigitated electrodes modified with reduced graphene oxide, we explored two strategies (covalent and non-covalent), establishing a reproducible and cost-effective methodology to develop OBP3-based electrical sensors. The non-covalent immobilization of a linker to the graphene-modified surface showed improved outcomes compared to the carbodiimide crosslinking chemistry. OBP3-sensors presented selectivity towards distinct model compounds in the gas phase (diisopropylphenol, dimethylpyrazine, menthone and decanol), in correlation with the dissociation constants measured by fluorescence displacement assays in solution. As a result, this study expands the practical applications of OBPs for gas-phase sensors, showcasing their potential for enhancing VOC detection.

High-Throughput Mass Spectrometry Analysis of N-Glycans and Protein Markers after FUT8 Knockdown in the Syngeneic SW480/SW620 Colorectal Cancer Cell Model.

Disruption of the glycosylation machinery is a common feature in many types of cancer, and colorectal cancer (CRC) is no exception. Core fucosylation is mediated by the enzyme fucosyltransferase 8 (FucT-8), which catalyzes the addition of α1,6-l-fucose to the innermost GlcNAc residue of N-glycans. We and others have documented the involvement of FucT-8 and core-fucosylated proteins in CRC progression, in which we addressed core fucosylation in the syngeneic CRC model formed by SW480 and SW620 tumor cell lines from the perspective of alterations in their N-glycosylation profile and protein expression as an effect of the knockdown of the FUT8 gene that encodes FucT-8. Using label-free, semiquantitative mass spectrometry (MS) analysis, we found noticeable differences in N-glycosylation patterns in FUT8-knockdown cells, affecting core fucosylation and sialylation, the Hex/HexNAc ratio, and antennarity. Furthermore, stable isotopic labeling of amino acids in cell culture (SILAC)-based proteomic screening detected the alteration of species involved in protein folding, endoplasmic reticulum (ER) and Golgi post-translational stabilization, epithelial polarity, and cellular response to damage and therapy. This data is available via ProteomeXchange with identifier PXD050012. Overall, the results obtained merit further investigation to validate their feasibility as biomarkers of progression and malignization in CRC, as well as their potential usefulness in clinical practice.

Folding mechanics of a bistable composite tape-spring for flexible mechanical hinge

A bistable composite tape-spring (CTS) structure is stable in both extended and coiled configurations, which can be fully coiled or folded. Its bistable coiling feature has been employed in a Roll-Out Solar Array and successfully deployed in space; its foldable characteristics is analogous to a flexible mechanical hinge, showing great potential to be applied in an aircraft landing gear. Both the structural coiling and folding mechanics are dependent on tape geometry; whilst the correlated scale effect has not been investigated, which significantly constrains its foldable mechanical hinge designs and smart driving analysis in order to further reduce weight and complexity for conventional mechanical hinges. Here, we studied the folding stability mechanics of the CTS structure towards its full tape-length range, where novel stress and shape transitional mechanisms are revealed. This is achieved by investigating the quasi-static folding process of the CTS, where new stable shape features in terms of critical transitional length and stable folded angle are observed and identified through experimental observations, finite element model, as well as theoretical analysis. Theoretical criteria were then determined from the strain energy analysis in comparison to the predefined folded tape shape features. The folding stability mechanisms towards its full tape-length range were proposed in order to facilitate customized flexible hinge design and in-situ smart driving analysis in order to replace conventional mechanical hinges.

PKHD1L1, a gene involved in the stereocilia coat, causes autosomal recessive nonsyndromic hearing loss

Identification of genes associated with nonsyndromic hearing loss is a crucial endeavor given the substantial number of individuals who remain without a diagnosis after even the most advanced genetic testing. PKHD1L1 was established as necessary for the formation of the cochlear hair-cell stereociliary coat and causes hearing loss in mice and zebrafish when mutated. We sought to determine if biallelic variants in PKHD1L1 also cause hearing loss in humans. Exome sequencing was performed on DNA of four families segregating autosomal recessive nonsyndromic sensorineural hearing loss. Compound heterozygous p.[(Gly129Ser)];p.[(Gly1314Val)] and p.[(Gly605Arg)];p[(Leu2818TyrfsTer5)], homozygous missense p.(His2479Gln) and nonsense p.(Arg3381Ter) variants were identified in PKHD1L1 that were predicted to be damaging using in silico pathogenicity prediction methods. In vitro functional analysis of two missense variants was performed using purified recombinant PKHD1L1 protein fragments. We then evaluated protein thermodynamic stability with and without the missense variants found in one of the families and performed a minigene splicing assay for another variant. In silico molecular modeling using AlphaFold2 and protein sequence alignment analysis were carried out to further explore potential variant effects on structure. In vitro functional assessment indicated that both engineered PKHD1L1 p.(Gly129Ser) and p.(Gly1314Val) mutant constructs significantly reduced the folding and structural stabilities of the expressed protein fragments, providing further evidence to support pathogenicity of these variants. Minigene assay of the c.1813G>A p.(Gly605Arg) variant, located at the boundary of exon 17, revealed exon skipping leading to an in-frame deletion of 48 amino acids. In silico molecular modeling exposed key structural features that might suggest PKHD1L1 protein destabilization. Multiple lines of evidence collectively associate PKHD1L1 with nonsyndromic mild–moderate to severe sensorineural hearing loss. PKHD1L1 testing in individuals with mild–moderate hearing loss may identify further affected families.

Understanding activity-stability tradeoffs in biocatalysts by enzyme proximity sequencing

Understanding the complex relationships between enzyme sequence, folding stability and catalytic activity is crucial for applications in industry and biomedicine. However, current enzyme assay technologies are limited by an inability to simultaneously resolve both stability and activity phenotypes and to couple these to gene sequences at large scale. Here we present the development of enzyme proximity sequencing, a deep mutational scanning method that leverages peroxidase-mediated radical labeling with single cell fidelity to dissect the effects of thousands of mutations on stability and catalytic activity of oxidoreductase enzymes in a single experiment. We use enzyme proximity sequencing to analyze how 6399 missense mutations influence folding stability and catalytic activity in a D-amino acid oxidase from Rhodotorula gracilis. The resulting datasets demonstrate activity-based constraints that limit folding stability during natural evolution, and identify hotspots distant from the active site as candidates for mutations that improve catalytic activity without sacrificing stability. Enzyme proximity sequencing can be extended to other enzyme classes and provides valuable insights into biophysical principles governing enzyme structure and function.

Molecular Compressive Force Sensor for Mapping Forces at the Cell-Substrate Interface.

Mechanical forces are crucial for biological processes such as T cell antigen recognition. A suite of molecular tension probes to measure pulling forces have been reported over the past decade; however, there are no reports of molecular probes for measuring compressive forces, representing a gap in the current mechanobiology toolbox. To address this gap, we report a molecular compression reporter using pseudostable hairpins (M-CRUSH). The design principle was based on a pseudostable DNA structure that folds in response to an external compressive force. We created a library of DNA stem-loop hairpins with varying thermodynamic stability, and then used Förster Resonance Energy Transfer (FRET) to quantify hairpin folding stability as a function of temperature and crowding. We identified an optimal pseudostable DNA hairpin highly sensitive to molecular crowding that displayed a shift in melting temperature (Tm) of 7 °C in response to a PEG crowding agent. When immobilized on surfaces, this optimized DNA hairpin showed a 29 ± 6% increase in FRET index in response to 25% w/w PEG 8K. As a proof-of-concept demonstration, we employed M-CRUSH to map the compressive forces generated by primary naïve T cells. We noted dynamic compressive forces that were highly sensitive to antigen presentation and coreceptor engagement. Importantly, mechanical forces are generated by cytoskeletal protrusions caused by acto-myosin activity. This was confirmed by treating cells with cytoskeletal inhibitors, which resulted in a lower FRET response when compared to untreated cells. Furthermore, we showed that M-CRUSH signal is dependent on probe density with greater density probes showing enhanced signal. Finally, we demonstrated that M-CRUSH probes are modular and can be applied to different cell types by displaying a compressive signal observed under human platelets. M-CRUSH offers a powerful tool to complement tension sensors and map out compressive forces in living systems.

Computational insights into dynamics and conformational stability of N-acetylmannosamine kinase mutations

The activity of UDP-GlcNAc 2-epimerase/ManNAc kinase (GNE) is essential for the biosynthesis of sialic acid, which is involved in cellular processes in health and diseases. GNE contains an N-terminal epimerase domain and a C-terminal kinase domain (N-acetylmannosamine kinase, MNK). Mutations of the GNE protein led to hypoactivity of the enzyme and cause sialurea or autosomal recessive inclusion body myopathy/Nonaka myopathy. Here, we used all-atom molecular dynamics (MD) simulations to comprehend the folding, dynamics and conformational stability of MNK variants, including the wild type (WT) and three mutants (H677R, V696M and H677R/V696M). The deleterious and destabilizing nature of MNK mutants were predicted using different prediction tools. Results predicted that mutations modulate the stability, flexibility and function of MNK. The effect of mutations on the conformational stability and dynamics of MNK was next studied through the free-energy landscape (FEL), hydrogen-bonds and secondary structure changes. The FEL results show that the mutations interfere with various conformational transitions in both WT and mutants, exposing the structural underpinnings of protein destabilization and unfolding brought on by mutation. We discover that, when compared to the other two mutations, V696M and H677R/V696M, H677R has the most harmful effects. These findings have a strong correlation with published experimental studies that demonstrate how these mutations disrupt MNK activity. Hence, this computational study describes the structural details to unravel the mutant effects at the atomistic resolution and has implications for understanding the GNE's physiological and pathological role. Communicated by Ramaswamy H. Sarma

Dissecting the Conformational Stability of a Glycan Hairpin.

Systematic structural studies of model oligopeptides revealed important aspects of protein folding and offered design principles to access non-natural materials. In the same way, the rules that regulate glycan folding could be established by studying synthetic oligosaccharide models. However, their analysis is often limited due to the synthetic and analytical complexity. By utilizing a glycan capable of spontaneously folding into a hairpin conformation as a model system, we investigated the factors that contribute to its conformational stability in aqueous solution. The modular design of the hairpin model featured a trisaccharide turn unit and two β-1,4-oligoglucoside stacking strands that allowed for systematic chemical modifications of the glycan sequence, including the introduction of NMR labels and staples. Nuclear magnetic resonance assisted by molecular dynamics simulations revealed that stereoelectronic effects and multiple glycan-glycan interactions are the major determinants of folding stabilization. Chemical modifications in the glycan primary sequence (e.g., strand elongation) can be employed to fine-tune the rigidity of structural motifs distant from the modification sites. These results could inspire the design of other glycan architectures, with implications in glycobiology and material sciences.

A multistable composite hinge structure

A composite tape-spring (CTS) structure is a thin-walled open slit tube with fibres oriented at ±45°, which is stable in both extended and coiled configurations. The governing factors of its bistability include composite constitutive behaviour, initial geometrical proportions, and geometrically non-linear structural behaviour. Its bistable principle can be employed to produce a flexible multistable hinge structure with tailorable stability. This is achieved by introducing variable stiffness design within a cylindrical shell structure, where folding stability is dependent on central functional patch region, and then connected to linking ploy regions. Thus, a novel multistable composite hinge structure can be designed with positive Gaussian curvature deformation, and its multistability is highly tailorable: a lengthy one-dimensional mechanical arm can be designed to coil and fold multiple times to enable large folding ratio. An analytical model was established based on the strain energy principle, in order to determine effects from functional tape length; the typical structural stability and stable configurations were then predicted with respect to regional length of the functional layer. It is found that the stability of a multistable composite hinge structure is dependent on geometry and combination of both the functional patch region, and connecting ploy region; the stable criteria are then proposed and show good agreement with experimental observations and FE analysis. These enrich the diversities of functional deployable structures to benefit novel requirements for various deployable mechanisms, and enable customised design, as well as smart driving for flexible and multifunctional mechanical composite hinge applications.

Comparing effects of attractive interactions in crowded systems: nonspecific, hydrophobic, and hydrogen bond interactions

The equilibrium stability of a protein is determined by its amino acid sequence and the solution conditions, such as temperature, pH and presence of chemical denaturant. The stability of a single protein in two identical solutions can nonetheless differ if other macromolecules, termed cosolutes or crowders, are present in one of the solutions at concentrations high enough to occupy a substantial fraction of the solution volume. This effect, due to the presence of the crowders, decreases or increases the stability depending on the interactions between the protein and crowders. Hard-core steric repulsions, which are responsible for the reduction in free volume, are expected to entropically stabilize the protein while attractive interactions can be destabilizing. Here we use a coarse-grained protein model to assess the impact of different types of crowder-protein interactions on the stability of a 35-amino acid model sequence folding into a helical bundle. We find that, for the same interaction strength and concentration, spherical crowders with a hydrophobic character are more destabilizing than crowders interacting nonspecifically with the protein. However, the two types of interactions differ in the degree of association between crowders and protein. At an interaction strength for which the attractive interactions roughly counteracts the stabilizing hard-core repulsions, the nonspecific interactions lead to much stronger crowder-protein association than the hydrophobic interactions. Additionally, we study crowders in the form of polypeptide chains, which are capable of hydrogen bonding with the protein. These peptide crowders have a destabilizing effect even at relatively low crowder concentrations, especially if the sequence of the peptide crowders includes hydrophobic amino acids. Our findings emphasize the importance of the interplay between different types of attractive crowder-protein interactions and entropic effects in determining the net effect on protein stability.

Tying a true topological protein knot by cyclization

Knotted proteins are fascinating to biophysicists because of their robust ability to fold into intricately defined three-dimensional structures with complex and topologically knotted arrangements. Exploring the biophysical properties of the knotted proteins is of significant interest, as they could offer enhanced chemical, thermal, and mechanostabilities. A true mathematical knot requires a closed path; in contrast, knotted protein structures have open N- and C-termini. To address the question of how a truly knotted protein differs from the naturally occurring counterpart, we enzymatically cyclized a 31 knotted YibK protein from Haemophilus influenza (HiYibK) to investigate the impact of path closure on its structure-function relationship and folding stability. Through the use of a multitude of structural and biophysical tools, including X-ray crystallography, NMR spectroscopy, small angle X-ray scattering, differential scanning calorimetry, and isothermal calorimetry, we showed that the path closure minimally perturbs the native structure and ligand binding of HiYibK. Nevertheless, the cyclization did alter the folding stability and mechanism according to chemical and thermal unfolding analysis. These molecular insights contribute to our fundamental understanding of protein folding and knotting that could have implications in the protein design with higher stabilities.

Estimating RNA Secondary Structure Folding Free Energy Changes with efn2.

A number of analyses require estimates of the folding free energy changes of specific RNA secondary structures. These predictions are often based on a set of nearest neighbor parameters that models the folding stability of a RNA secondary structure as the sum of folding stabilities of the structural elements that comprise the secondary structure. In the software suite RNAstructure, the free energy change calculation is implemented in the program efn2. The efn2 program estimates the folding free energy change and the experimental uncertainty in the folding free energy change. It can be run through the graphical user interface for RNAstructure, from the command line, or a web server. This chapter provides detailed protocols for using efn2.

Competitive Influence of Alkali Metals in the Ion Atmosphere on Nucleic Acid Duplex Stability.

The nonspecific atmosphere around nucleic acids, often termed the ion atmosphere, encompasses a collection of weak ion-nucleic acid interactions. Although nonspecific, the ion atmosphere has been shown to influence nucleic acid folding and structural stability. Studies investigating the composition of the ion atmosphere have shown competitive occupancy of the atmosphere between metal ions in the same solution. Many studies have investigated single ion effects on nucleic acid secondary structure stability; however, no comprehensive studies have investigated how the competitive occupancy of mixed ions in the ion atmosphere influences nucleic acid secondary structure stability. Here, six oligonucleotides were optically melted in buffers containing molar quantities, or mixtures, of either XCl (X = Li, K, Rb, or Cs) or NaCl. A correction factor was developed to better predict RNA duplex stability in solutions containing mixed XCl/NaCl. For solutions containing a 1:1 mixture of XCl/NaCl, one alkali metal chloride contributed more to duplex stability than the other. Overall, there was a 54% improvement in predictive capabilities with the correction factor compared with the standard 1.0 M NaCl nearest-neighbor models. This correction factor can be used in models to better predict RNA secondary structure in solutions containing mixed XCl/NaCl.

An Investigation of Acoustic Back-Coupling in Human Phonation on a Synthetic Larynx Model.

In the human phonation process, acoustic standing waves in the vocal tract can influence the fluid flow through the glottis as well as vocal fold oscillation. To investigate the amount of acoustic back-coupling, the supraglottal flow field has been recorded via high-speed particle image velocimetry (PIV) in a synthetic larynx model for several configurations with different vocal tract lengths. Based on the obtained velocity fields, acoustic source terms were computed. Additionally, the sound radiation into the far field was recorded via microphone measurements and the vocal fold oscillation via high-speed camera recordings. The PIV measurements revealed that near a vocal tract resonance frequency fR, the vocal fold oscillation frequency fo (and therefore also the flow field's fundamental frequency) jumps onto fR. This is accompanied by a substantial relative increase in aeroacoustic sound generation efficiency. Furthermore, the measurements show that fo-fR-coupling increases vocal efficiency, signal-to-noise ratio, harmonics-to-noise ratio and cepstral peak prominence. At the same time, the glottal volume flow needed for stable vocal fold oscillation decreases strongly. All of this results in an improved voice quality and phonation efficiency so that a person phonating with fo-fR-coupling can phonate longer and with better voice quality.

A sodium/potassium switch for G4-prone G/C-rich sequences.

Metal ions are essential components for the survival of living organisms. For most species, intracellular and extracellular ionic conditions differ significantly. As G-quadruplexes (G4s) areion-dependent structures, changes in the [Na+]/[K+] ratio may affect the folding of genomic G4s. More than 11000 putative G4 sequences in the human genome (hg19) contain at least two runs of three continuous cytosines, and these mixed G/C-rich sequences may form a quadruplex or a competing hairpin structure based on G-C base pairing. In this study, we examine how the [Na+]/[K+] ratio influences the structures of G/C-rich sequences. The natural G4 structure with a9-nt long central loop, CEBwt, was chosen as a model sequence, and the loop bases were gradually replaced by cytosines. The series of CEB mutations revealed that the presence of cytosines in G4 loops does not prevent G4 folding or decrease G4 stability but increases the probability of forming a competing structure, either a hairpin or an intermolecular duplex. Slow conversion to the quadruplex in vitro (in a potassium-rich buffer) and cells was demonstrated by NMR. 'Shape-shifting' sequences may respond to [Na+]/[K+] changes with delayed kinetics.

Statistical modeling to quantify the uncertainty of FoldX-predicted protein folding and binding stability

BackgroundComputational methods of predicting protein stability changes upon missense mutations are invaluable tools in high-throughput studies involving a large number of protein variants. However, they are limited by a wide variation in accuracy and difficulty of assessing prediction uncertainty. Using a popular computational tool, FoldX, we develop a statistical framework that quantifies the uncertainty of predicted changes in protein stability.ResultsWe show that multiple linear regression models can be used to quantify the uncertainty associated with FoldX prediction for individual mutations. Comparing the performance among models with varying degrees of complexity, we find that the model precision improves significantly when we utilize molecular dynamics simulation as part of the FoldX workflow. Based on the model that incorporates information from molecular dynamics, biochemical properties, as well as FoldX energy terms, we can generally expect upper bounds on the uncertainty of folding stability predictions of ± 2.9 kcal/mol and ± 3.5 kcal/mol for binding stability predictions. The uncertainty for individual mutations varies; our model estimates it using FoldX energy terms, biochemical properties of the mutated residue, as well as the variability among snapshots from molecular dynamics simulation.ConclusionsUsing a linear regression framework, we construct models to predict the uncertainty associated with FoldX prediction of stability changes upon mutation. This technique is straightforward and can be extended to other computational methods as well.