DAMpred: Recognizing Disease-Associated nsSNPs through Bayes-Guided Neural-Network Model Built on Low-Resolution Structure Prediction of Proteins and Protein-Protein Interactions

FCMSTrans: Accurate Prediction of Disease-Associated nsSNPs by Utilizing Multiscale Convolution and Deep Feature Combination within a Transformer Framework

Nonsynonymous single-nucleotide polymorphisms (nsSNPs), implicated in over 6000 diseases, necessitate accurate prediction for expedited drug discovery and improved disease diagnosis. In this study, we propose FCMSTrans, a novel nsSNP predictor that innovatively combines the transformer framework and multiscale modules for comprehensive feature extraction. The distinctive attribute of FCMSTrans resides in a deep feature combination strategy. This strategy amalgamates evolutionary-scale modeling (ESM) and ProtTrans (PT) features, providing an understanding of protein biochemical properties, and position-specific scoring matrix, secondary structure, predicted relative solvent accessibility, and predicted disorder (PSPP) features, which are derived from four protein sequences and structure-oriented characteristics. This feature combination offers a comprehensive view of the molecular dynamics involving nsSNPs. Our model employs the transformer's self-attention mechanisms across multiple layers, extracting higher-level and abstract representations. Simultaneously, varied-level features are captured by multiscale convolutions, enriching feature abstraction at multiple echelons. Our comparative analyses with existing methodologies highlight significant improvements made possible by the integrated feature fusion approach adopted in FCMSTrans. This is further substantiated by performance assessments based on diverse data sets, such as PredictSNP, MMP, and PMD, with areas under the curve (AUCs) of 0.869, 0.819, and 0.693, respectively. Furthermore, FCMSTrans shows robustness and superiority by outperforming the current best predictor, PROVEAN, in a blind test conducted on a third-party data set, achieving an impressive AUC score of 0.7838. The Python code of FCMSTrans is available at https://github.com/gc212/FCMSTrans for academic usage.

MMPatho: Leveraging Multilevel Consensus and Evolutionary Information for Enhanced Missense Mutation Pathogenic Prediction

Understanding the pathogenicity of missense mutation (MM) is essential for shed light on genetic diseases, gene functions, and individual variations. In this study, we propose a novel computational approach, called MMPatho, for enhancing missense mutation pathogenic prediction. First, we established a large-scale nonredundant MM benchmark data set based on the entire Ensembl database, complemented by a focused blind test set specifically for pathogenic GOF/LOF MM. Based on this data set, for each mutation, we utilized Ensembl VEP v104 and dbNSFP v4.1a to extract variant-level, amino acid-level, individuals' outputs, and genome-level features. Additionally, protein sequences were generated using ENSP identifiers with the Ensembl API, and then encoded. The mutant sites' ESM-1b and ProtTrans-T5 embeddings were subsequently extracted. Then, our model group (MMPatho) was developed by leveraging upon these efforts, which comprised ConsMM and EvoIndMM. To be specific, ConsMM employs individuals' outputs and XGBoost with SHAP explanation analysis, while EvoIndMM investigates the potential enhancement of predictive capability by incorporating evolutionary information from ESM-1b and ProtT5-XL-U50, large protein language embeddings. Through rigorous comparative experiments, both ConsMM and EvoIndMM were capable of achieving remarkable AUROC (0.9836 and 0.9854) and AUPR (0.9852 and 0.9902) values on the blind test set devoid of overlapping variations and proteins from the training data, thus highlighting the superiority of our computational approach in the prediction of MM pathogenicity. Our Web server, available at http://csbio.njust.edu.cn/bioinf/mmpatho/, allows researchers to predict the pathogenicity (alongside the reliability index score) of MMs using the ConsMM and EvoIndMM models and provides extensive annotations for user input. Additionally, the newly constructed benchmark data set and blind test set can be accessed via the data page of our web server.

CRISPR-Mediated Base Editing: Promises and Challenges for a Viable Oncotherapy Strategy

Base editing technology, developed from the CRISPR/Cas9 system, is able to efficiently implement single-base substitutions at specific DNA or RNA sites without generating double-strand breaks with precision and efficiency. Point mutations account for 58% of disease-causing genetic mutations in humans, and single nucleotide variants are an important cause of tumorigenesis, and the advent of base editors offers new hope for the study or treatment of such diseases. Although it has some limitations, base editors have been continuously improved in terms of editing efficiency, specificity, and product purity since their development. In this review, we examine the main base editing technologies and discuss their applications and prospects in tumor research and therapy, as well as elaborate on their mode of delivery.

Structure-based pathogenicity relationship identifier for predicting effects of single missense variants and discovery of higher-order cancer susceptibility clusters of mutations

We report the structure-based pathogenicity relationship identifier (SPRI), a novel computational tool for accurate evaluation of pathological effects of missense single mutations and prediction of higher-order spatially organized units of mutational clusters. SPRI can effectively extract properties determining pathogenicity encoded in protein structures, and can identify deleterious missense mutations of germ line origin associated with Mendelian diseases, as well as mutations of somatic origin associated with cancer drivers. It compares favorably to other methods in predicting deleterious mutations. Furthermore, SPRI can discover spatially organized pathogenic higher-order spatial clusters (patHOS) of deleterious mutations, including those of low recurrence, and can be used for discovery of candidate cancer driver genes and driver mutations. We further demonstrate that SPRI can take advantage of AlphaFold2 predicted structures and can be deployed for saturation mutation analysis of the whole human proteome.

Development and validation of a machine learning algorithm prediction for dense granule proteins in Apicomplexa

BACKGROUND

Apicomplexa consist of numerous pathogenic parasitic protistan genera that invade host cells and reside and replicate within the parasitophorous vacuole (PV). Through this interface, the parasite exchanges nutrients and affects transport and immune modulation. During the intracellular life-cycle, the specialized secretory organelles of the parasite secrete an array of proteins, among which dense granule proteins (GRAs) play a major role in the modification of the PV. Despite this important role of GRAs, a large number of potential GRAs remain unidentified in Apicomplexa.

METHODS

A multi-view attention graph convolutional network (MVA-GCN) prediction model with multiple features was constructed using a combination of machine learning and genomic datasets, and the prediction was performed on selected Neospora caninum protein data. The candidate GRAs were verified by a CRISPR/Cas9 gene editing system, and the complete NcGRA64(a,b) gene knockout strain was constructed and the phenotypes of the mutant were analyzed.

RESULTS

The MVA-GCN prediction model was used to screen N. caninum candidate GRAs, and two novel GRAs (NcGRA64a and NcGRA64b) were verified by gene endogenous tagging. Knockout of complete genes of NcGRA64(a,b) in N. caninum did not affect the parasite's growth and replication in vitro and virulence in vivo.

CONCLUSIONS

Our study showcases the utility of the MVA-GCN deep learning model for mining Apicomplexa GRAs in genomic datasets, and the prediction model also has certain potential in mining other functional proteins of apicomplexan parasites.

Comprehensive in-silico analysis of deleterious SNPs in APOC2 and APOA5 and their differential expression in cancer and cardiovascular diseases conditions

Genetic variations in APOC2 and APOA5 genes involve activating lipoprotein lipase (LPL), responsible for the hydrolysis of triglycerides (TG) in blood and whose impaired functions affect the TG metabolism and are associated with metabolic diseases. In this study, we investigate the biological significance of genetic variations at the DNA sequence and structural level using various computational tools. Subsequently, 8 (APOC2) and 17 (APOA5) non-synonymous SNPs (nsSNPs) were identified as high-confidence deleterious SNPs based on the effects of the mutations on protein conservation, stability, and solvent accessibility. Furthermore, based on our docking results, the interaction of native and mutant forms of the corresponding proteins with LPL depicts differences in root mean square deviation (RMSD), and binding affinities suggest that these mutations may affect their function. Furthermore, in vivo, and in vitro studies have shown that differential expression of these genes in disease conditions due to the influence of nsSNPs abundance may be associated with promoting the development of cancer and cardiovascular diseases. Preliminary screening using computational methods can be a helpful start in understanding the effects of mutations in APOC2 and APOA5 on lipid metabolism; however, further wet-lab experiments would further strengthen the conclusions drawn from the computational study.

Circuit topology predicts pathogenicity of missense mutations

The contact topology of a protein determines important aspects of the folding process. The topological measure of contact order has been shown to be predictive of the rate of folding. Circuit topology is emerging as another fundamental descriptor of biomolecular structure, with predicted effects on the folding rate. We analyze the residue‐based circuit topological environments of 21 K mutations labeled as pathogenic or benign. Multiple statistical lines of reasoning support the conclusion that the number of contacts in two specific circuit topological arrangements, namely inverse parallel and cross relations, with contacts involving the mutated residue have discriminatory value in determining the pathogenicity of human variants. We investigate how results vary with residue type and according to whether the gene is essential. We further explore the relationship to a number of structural features and find that circuit topology provides nonredundant information on protein structures and pathogenicity of mutations. Results may have implications for the polymer physics of protein folding and suggest that “local” topological information, including residue‐based circuit topology and residue contact order, could be useful in improving state‐of‐the‐art machine learning algorithms for pathogenicity prediction.

Prediction of disease-associated nsSNPs by integrating multi-scale ResNet models with deep feature fusion

More than 6000 human diseases have been recorded to be caused by non-synonymous single nucleotide polymorphisms (nsSNPs). Rapid and accurate prediction of pathogenic nsSNPs can improve our understanding of the principle and design of new drugs, which remains an unresolved challenge. In the present work, a new computational approach, termed MSRes-MutP, is proposed based on ResNet blocks with multi-scale kernel size to predict disease-associated nsSNPs. By feeding the serial concatenation of the extracted four types of features, the performance of MSRes-MutP does not obviously improve. To address this, a second model FFMSRes-MutP is developed, which utilizes deep feature fusion strategy and multi-scale 2D-ResNet and 1D-ResNet blocks to extract relevant two-dimensional features and physicochemical properties. FFMSRes-MutP with the concatenated features achieves a better performance than that with individual features. The performance of FFMSRes-MutP is benchmarked on five different datasets. It achieves the Matthew's correlation coefficient (MCC) of 0.593 and 0.618 on the PredictSNP and MMP datasets, which are 0.101 and 0.210 higher than that of the existing best method PredictSNP1. When tested on the HumDiv and HumVar datasets, it achieves MCC of 0.9605 and 0.9507, and area under curve (AUC) of 0.9796 and 0.9748, which are 0.1747 and 0.2669, 0.0853 and 0.1335, respectively, higher than the existing best methods PolyPhen-2 and FATHMM (weighted). In addition, on blind test using a third-party dataset, FFMSRes-MutP performs as the second-best predictor (with MCC and AUC of 0.5215 and 0.7633, respectively), when compared with the other four predictors. Extensive benchmarking experiments demonstrate that FFMSRes-MutP achieves effective feature fusion and can be explored as a useful approach for predicting disease-associated nsSNPs. The webserver is freely available at http://csbio.njust.edu.cn/bioinf/ffmsresmutp/ for academic use.

MutTMPredictor: Robust and accurate cascade XGBoost classifier for prediction of mutations in transmembrane proteins

•

Prediction of mutations in transmembrane proteins is of significance for diseases diagnosis.

•

Building on the evolutionary information, proposed the Gaussian WAPSSM algorithm.

•

Based on WAPSSM and sequence and structure-based features, proposed the cascade XGBoost algorithm.

•

Webserver is freely at (http://csbio.njust.edu.cn/bioinf/ffmsresmutp/).

•

Implement MutTMPredictor to predict mutations in transmembrane proteins.

Transmembrane proteins have critical biological functions and play a role in a multitude of cellular processes including cell signaling, transport of molecules and ions across membranes. Approximately 60% of transmembrane proteins are considered as drug targets. Missense mutations in such proteins can lead to many diverse diseases and disorders, such as neurodegenerative diseases and cystic fibrosis. However, there are limited studies on mutations in transmembrane proteins. In this work, we first design a new feature encoding method, termed weight attenuation position-specific scoring matrix (WAPSSM), which builds upon the protein evolutionary information. Then, we propose a new mutation prediction algorithm (cascade XGBoost) by leveraging the idea learned from consensus predictors and gcForest. Multi-level experiments illustrate the effectiveness of WAPSSM and cascade XGBoost algorithms. Finally, based on WAPSSM and other three types of features, in combination with the cascade XGBoost algorithm, we develop a new transmembrane protein mutation predictor, named MutTMPredictor. We benchmark the performance of MutTMPredictor against several existing predictors on seven datasets. On the 546 mutations dataset, MutTMPredictor achieves the accuracy (ACC) of 0.9661 and the Matthew’s Correlation Coefficient (MCC) of 0.8950. While on the 67,584 dataset, MutTMPredictor achieves an MCC of 0.7523 and area under curve (AUC) of 0.8746, which are 0.1625 and 0.0801 respectively higher than those of the existing best predictor (fathmm). Besides, MutTMPredictor also outperforms two specific predictors on the Pred-MutHTP datasets. The results suggest that MutTMPredictor can be used as an effective method for predicting and prioritizing missense mutations in transmembrane proteins. The MutTMPredictor webserver and datasets are freely accessible at http://csbio.njust.edu.cn/bioinf/muttmpredictor/ for academic use.

ADDRESS: A Database of Disease-associated Human Variants Incorporating Protein Structure and Folding Stabilities

Numerous human diseases are caused by mutations in genomic sequences. Since amino acid changes affect protein function through mechanisms often predictable from protein structure, the integration of structural and sequence data enables us to estimate with greater accuracy whether and how a given mutation will lead to disease. Publicly available annotated databases enable hypothesis assessment and benchmarking of prediction tools. However, the results are often presented as summary statistics or black box predictors, without providing full descriptive information. We developed a new semi-manually curated human variant database presenting information on the protein contact-map, sequence-to-structure mapping, amino acid identity change, and stability prediction for the popular UniProt database. We found that the profiles of pathogenic and benign missense polymorphisms can be effectively deduced using decision trees and comparative analyses based on the presented dataset. The database is made publicly available through https://zhanglab.ccmb.med.umich.edu/ADDRESS.

TargetMM: Accurate Missense Mutation Prediction by Utilizing Local and Global Sequence Information with Classifier Ensemble

Aim and Objective:

Missense mutation (MM) may lead to various human diseases by disabling proteins. Accurate prediction of MM is important and challenging for both protein function annotation and drug design. Although several computational methods yielded acceptable success rates, there is still room for further enhancing the prediction performance of MM.

Materials and Methods:

In the present study, we designed a new feature extracting method, which considers the impact degree of residues in the microenvironment range to the mutation site. Stringent cross-validation and independent test on benchmark datasets were performed to evaluate the efficacy of the proposed feature extracting method. Furthermore, three heterogeneous prediction models were trained and then ensembled for the final prediction. By combining the feature representation method and classifier ensemble technique, we reported a novel MM predictor called TargetMM for identifying the pathogenic mutations from the neutral ones.

Results:

Comparison outcomes based on statistical evaluation demonstrate that TargetMM outperforms the prior advanced methods on the independent test data. The source codes and benchmark datasets of TargetMM are freely available at https://github.com/sera616/TargetMM.git for academic use.

Novel disease-causing variants in a cohort of Iranian patients with metachromatic leukodystrophy and in silico analysis of their pathogenicity

OBJECTIVE

Metachromatic leukodystrophy (MLD) is an autosomal recessive leukodystrophy caused by deficiency of aryl sulfatase A (ASA) activity affecting the nervous system. MLD and mutations in ARSA have not been widely studied in non-European cohorts. The genotype-phenotype spectrum of MLD patients was investigated in this study of a cohort of Iranian leukodystrophy patients. In silico analysis was performed to investigate the pathogenicity of the variants.

METHODS

Genetic analysis for 25 patients was performed with direct sequencing of the ARSA gene. The missense variants underwent in silico analysis to characterize the pathogenicity based on predicted structural and stability changes.

RESULTS

19 patients had variants in ARSA genes, including 18 homozygotes and one compound heterozygote individual. In 6 individuals no mutations were found in ARSA gene, suggesting an alternative cause of their leukodystrophy. We found 5 novel disease causing variants: p.Phe64Ile, p.Ser292Alafs*34, p.Arg99Profs*35, p.Phe400Leu and p.Leu429Pro. 32 % of the patients had p.Gly311Ser substitution and resulted in juvenile MLD type. Different in silico analysis showed variable pathogenic effect for the variants.

CONCLUSION

c.931 G > A (p.Gly311Ser) and c.465 + 1 G > A variants are the most frequent alleles among Iranian MLD patients and five mutations appear to be confined to the Iranian patients. Population screening for these variants may be helpful to reduce the burden of the disease in this part of the world.

Evolution of Sequence-based Bioinformatics Tools for Protein-protein Interaction Prediction

Protein-protein interactions (PPIs) are the physical connections between two or more proteins via electrostatic forces or hydrophobic effects. Identification of the PPIs is pivotal, which contributes to many biological processes including protein function, disease incidence, and therapy design. The experimental identification of PPIs via high-throughput technology is time-consuming and expensive. Bioinformatics approaches are expected to solve such restrictions. In this review, our main goal is to provide an inclusive view of the existing sequence-based computational prediction of PPIs. Initially, we briefly introduce the currently available PPI databases and then review the state-of-the-art bioinformatics approaches, working principles, and their performances. Finally, we discuss the caveats and future perspective of the next generation algorithms for the prediction of PPIs.

Mutations in transmembrane proteins: diseases, evolutionary insights, prediction and comparison with globular proteins

Membrane proteins are unique in that they interact with lipid bilayers, making them indispensable for transporting molecules and relaying signals between and across cells. Due to the significance of the protein's functions, mutations often have profound effects on the fitness of the host. This is apparent both from experimental studies, which implicated numerous missense variants in diseases, as well as from evolutionary signals that allow elucidating the physicochemical constraints that intermembrane and aqueous environments bring. In this review, we report on the current state of knowledge acquired on missense variants (referred to as to single amino acid variants) affecting membrane proteins as well as the insights that can be extrapolated from data already available. This includes an overview of the annotations for membrane protein variants that have been collated within databases dedicated to the topic, bioinformatics approaches that leverage evolutionary information in order to shed light on previously uncharacterized membrane protein structures or interaction interfaces, tools for predicting the effects of mutations tailored specifically towards the characteristics of membrane proteins as well as two clinically relevant case studies explaining the implications of mutated membrane proteins in cancer and cardiomyopathy.

Germline Mutations in FAF1 Are Associated With Hereditary Colorectal Cancer

BACKGROUND & AIMS

A significant proportion of colorectal cancer (CRC) cases have familial aggregation but little is known about the genetic factors that contribute to these cases. We performed an exhaustive functional characterization of genetic variants associated with familial CRC.

METHODS

We performed whole-exome sequencing analyses of 75 patients from 40 families with a history of CRC (including early-onset cases) of an unknown germline basis (discovery cohort). We also sequenced specific genes in DNA from an external replication cohort of 473 families, including 488 patients with colorectal tumors that had normal expression of mismatch repair proteins (validation cohort). We disrupted the Fas-associated factor 1 gene (FAF1) in DLD-1 CRC cells using CRISPR/Cas9 gene editing; some cells were transfected with plasmids that express FAF1 missense variants. Cells were analyzed by immunoblots, quantitative real-time polymerase chain reaction, and functional assays monitoring apoptosis, proliferation, and assays for Wnt signaling or nuclear factor (NF)-kappa-B activity.

RESULTS

We identified predicted pathogenic variant in the FAF1 gene (c.1111G>A; p.Asp371Asn) in the discovery cohort; it was present in 4 patients of the same family. We identified a second variant in FAF1 in the validation cohort (c.254G>C; p.Arg85Pro). Both variants encoded unstable FAF1 proteins. Expression of these variants in CRC cells caused them to become resistant to apoptosis, accumulate beta-catenin in the cytoplasm, and translocate NF-kappa-B to the nucleus.

CONCLUSIONS

In whole-exome sequencing analyses of patients from families with a history of CRC, we identified variants in FAF1 that associate with development of CRC. These variants encode unstable forms of FAF1 that increase resistance of CRC cells to apoptosis and increase activity of beta-catenin and NF-kappa-B.