The ability to consistently distinguish real protein structures from computationally generated model decoys is not yet a solved problem. One route to distinguish real protein structures from decoys is to delineate the important physical features that specify a real protein. For example, it has long been appreciated that the hydrophobic cores of proteins contribute significantly to their stability. We used two sources to obtain datasets of decoys to compare with real protein structures: submissions to the biennial Critical Assessment of protein Structure Prediction competition, in which researchers attempt to predict the structure of a protein only knowing its amino acid sequence, and also decoys generated by 3DRobot, which have user‐specified global root‐mean‐squared deviations from experimentally determined structures. Our analysis revealed that both sets of decoys possess cores that do not recapitulate the key features that define real protein cores. In particular, the model structures appear more densely packed (because of energetically unfavorable atomic overlaps), contain too few residues in the core, and have improper distributions of hydrophobic residues throughout the structure. Based on these observations, we developed a feed‐forward neural network, which incorporates key physical features of protein cores, to predict how well a computational model recapitulates the real protein structure without knowledge of the structure of the target sequence. By identifying the important features of protein structure, our method is able to rank decoy structures with similar accuracy to that obtained by state‐of‐the‐art methods that incorporate many additional features. The small number of physical features makes our model interpretable, emphasizing the importance of protein packing and hydrophobicity in protein structure prediction.
- Award ID(s):
- 1900061
- NSF-PAR ID:
- 10343790
- Date Published:
- Journal Name:
- IEEE/ACM Transactions on Computational Biology and Bioinformatics
- ISSN:
- 1545-5963
- Page Range / eLocation ID:
- 1 to 1
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract -
more » « less
Abstract Protein
S -sulfenylation, which results from oxidation of free thiols on cysteine residues, has recently emerged as an important post-translational modification that regulates the structure and function of proteins involved in a variety of physiological and pathological processes. By altering the size and physiochemical properties of modified cysteine residues, sulfenylation can impact the cellular function of proteins in several different ways. Thus, the ability to rapidly and accurately identify putative sulfenylation sites in proteins will provide important insights into redox-dependent regulation of protein function in a variety of cellular contexts. Though bottom-up proteomic approaches, such as tandem mass spectrometry (MS/MS), provide a wealth of information about global changes in the sulfenylation state of proteins, MS/MS-based experiments are often labor-intensive, costly and technically challenging. Therefore, to complement existing proteomic approaches, researchers have developed a series of computational tools to identify putative sulfenylation sites on proteins. However, existing methods often suffer from low accuracy, specificity, and/or sensitivity. In this study, we developed SVM-SulfoSite, a novel sulfenylation prediction tool that uses support vector machines (SVM) to identify key determinants of sulfenylation among five feature classes: binary code, physiochemical properties, k-space amino acid pairs, amino acid composition and high-quality physiochemical indices. Using 10-fold cross-validation, SVM-SulfoSite achieved 95% sensitivity and 83% specificity, with an overall accuracy of 89% and Matthew’s correlation coefficient (MCC) of 0.79. Likewise, using an independent test set of experimentally identified sulfenylation sites, our method achieved scores of 74%, 62%, 80% and 0.42 for accuracy, sensitivity, specificity and MCC, with an area under the receiver operator characteristic (ROC) curve of 0.81. Moreover, in side-by-side comparisons, SVM-SulfoSite performed as well as or better than existing sulfenylation prediction tools. Together, these results suggest that our method represents a robust and complementary technique for advanced exploration of protein S-sulfenylation. -
more » « less
ResNet and, more recently, AlphaFold2 have demonstrated that deep neural networks can now predict a tertiary structure of a given protein amino-acid sequence with high accuracy. This seminal development will allow molecular biology researchers to advance various studies linking sequence, structure, and function. Many studies will undoubtedly focus on the impact of sequence mutations on stability, fold, and function. In this paper, we evaluate the ability of AlphaFold2 to predict accurate tertiary structures of wildtype and mutated sequences of protein molecules. We do so on a benchmark dataset in mutation modeling studies. Our empirical evaluation utilizes global and local structure analyses and yields several interesting observations. It shows, for instance, that AlphaFold2 performs similarly on wildtype and variant sequences. The placement of the main chain of a protein molecule is highly accurate. However, while AlphaFold2 reports similar confidence in its predictions over wildtype and variant sequences, its performance on placements of the side chains suffers in comparison to main-chain predictions. The analysis overall supports the premise that AlphaFold2-predicted structures can be utilized in further downstream tasks, but that further refinement of these structures may be necessary.
-
more » « less
Recent advances in Artificial Intelligence and Machine Learning (
e.g ., AlphaFold, RosettaFold, and ESMFold) enable prediction of three-dimensional (3D) protein structures from amino acid sequences alone at accuracies comparable to lower-resolution experimental methods. These tools have been employed to predict structures across entire proteomes and the results of large-scale metagenomic sequence studies, yielding an exponential increase in available biomolecular 3D structural information. Given the enormous volume of this newly computed biostructure data, there is an urgent need for robust tools to manage, search, cluster, and visualize large collections of structures. Equally important is the capability to efficiently summarize and visualize metadata, biological/biochemical annotations, and structural features, particularly when working with vast numbers of protein structures of both experimental origin from the Protein Data Bank (PDB) and computationally-predicted models. Moreover, researchers require advanced visualization techniques that support interactive exploration of multiple sequences and structural alignments. This paper introduces a suite of tools provided on the RCSB PDB research-focused web portal RCSB. org, tailor-made for efficient management, search, organization, and visualization of this burgeoning corpus of 3D macromolecular structure data. -
null (Ed.)The novel coronavirus severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the cause of COVID-19. The main receptor of SARS-CoV-2, angiotensin I converting enzyme 2 (ACE2), is now undergoing extensive scrutiny to understand the routes of transmission and sensitivity in different species. Here, we utilized a unique dataset of ACE2 sequences from 410 vertebrate species, including 252 mammals, to study the conservation of ACE2 and its potential to be used as a receptor by SARS-CoV-2. We designed a five-category binding score based on the conservation properties of 25 amino acids important for the binding between ACE2 and the SARS-CoV-2 spike protein. Only mammals fell into the medium to very high categories and only catarrhine primates into the very high category, suggesting that they are at high risk for SARS-CoV-2 infection. We employed a protein structural analysis to qualitatively assess whether amino acid changes at variable residues would be likely to disrupt ACE2/SARS-CoV-2 spike protein binding and found the number of predicted unfavorable changes significantly correlated with the binding score. Extending this analysis to human population data, we found only rare (frequency <0.001) variants in 10/25 binding sites. In addition, we found significant signals of selection and accelerated evolution in the ACE2 coding sequence across all mammals, and specific to the bat lineage. Our results, if confirmed by additional experimental data, may lead to the identification of intermediate host species for SARS-CoV-2, guide the selection of animal models of COVID-19, and assist the conservation of animals both in native habitats and in human care.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1109/TCBB.2021.3134103
