A Review of Machine Learning and Algorithmic Methods for Protein Phosphorylation Site Prediction

Phosphorylation of mammalian cytosolic and mitochondrial malate dehydrogenase: insights into regulation

Malate dehydrogenase (MDH) is a key enzyme in mammalian metabolic pathways in cytosolic and mitochondrial compartments. Regulation of MDH through phosphorylation remains an underexplored area. In this review we consolidate evidence supporting the potential role of phosphorylation in modulating the function of mammalian MDH. Parallels are drawn with the phosphorylation of lactate dehydrogenase, a homologous enzyme, to reveal its regulatory significance and to suggest a similar regulatory strategy for MDH. Comprehensive mining of phosphorylation databases, provides substantial experimental (primarily mass spectrometry) evidence of MDH phosphorylation in mammalian cells. Experimentally identified phosphorylation sites are overlaid with MDH's functional domains, offering perspective on how these modifications could influence enzyme activity. Preliminary results are presented from phosphomimetic mutations (serine/threonine residues changed to aspartate) generated in recombinant MDH proteins serving as a proof of concept for the regulatory impact of phosphorylation. We also examine and highlight several approaches to probe the structural and cellular impact of phosphorylation. This review highlights the need to explore the dynamic nature of MDH phosphorylation and calls for identifying the responsible kinases and the physiological conditions underpinning this modification. The synthesis of current evidence and experimental data aims to provide insights for future research on understanding MDH regulation, offering new avenues for therapeutic interventions in metabolic disorders and cancer.

Current computational tools for protein lysine acylation site prediction

As a main subtype of post-translational modification (PTM), protein lysine acylations (PLAs) play crucial roles in regulating diverse functions of proteins. With recent advancements in proteomics technology, the identification of PTM is becoming a data-rich field. A large amount of experimentally verified data is urgently required to be translated into valuable biological insights. With computational approaches, PLA can be accurately detected across the whole proteome, even for organisms with small-scale datasets. Herein, a comprehensive summary of 166 in silico PLA prediction methods is presented, including a single type of PLA site and multiple types of PLA sites. This recapitulation covers important aspects that are critical for the development of a robust predictor, including data collection and preparation, sample selection, feature representation, classification algorithm design, model evaluation, and method availability. Notably, we discuss the application of protein language models and transfer learning to solve the small-sample learning issue. We also highlight the prediction methods developed for functionally relevant PLA sites and species/substrate/cell-type-specific PLA sites. In conclusion, this systematic review could potentially facilitate the development of novel PLA predictors and offer useful insights to researchers from various disciplines.

The expanding landscape of canonical and non-canonical protein phosphorylation

Protein phosphorylation is a crucial regulatory mechanism in cell signaling, acting as a molecular switch that modulates protein function. Catalyzed by protein kinases and reversed by phosphoprotein phosphatases, it is essential in both normal physiological and pathological states. Recent advances have uncovered a vast and intricate landscape of protein phosphorylation that include histidine phosphorylation and more unconventional events, such as pyrophosphorylation and polyphosphorylation. Many questions remain about the true size of the phosphoproteome and, more importantly, its site-specific functional relevance. The involvement of unconventional actors such as pseudokinases and pseudophosphatases adds further complexity to be resolved. This review explores recent discoveries and ongoing challenges, highlighting the need for continued research to fully elucidate the roles and regulation of protein phosphorylation.

Protein modification in neurodegenerative diseases

Posttranslational modifications play a crucial role in governing cellular functions and protein behavior. Researchers have implicated dysregulated posttranslational modifications in protein misfolding, which results in cytotoxicity, particularly in neurodegenerative diseases such as Alzheimer disease, Parkinson disease, and Huntington disease. These aberrant posttranslational modifications cause proteins to gather in certain parts of the brain that are linked to the development of the diseases. This leads to neuronal dysfunction and the start of neurodegenerative disease symptoms. Cognitive decline and neurological impairments commonly manifest in neurodegenerative disease patients, underscoring the urgency of comprehending the posttranslational modifications' impact on protein function for targeted therapeutic interventions. This review elucidates the critical link between neurodegenerative diseases and specific posttranslational modifications, focusing on Tau, APP, α-synuclein, Huntingtin protein, Parkin, DJ-1, and Drp1. By delineating the prominent aberrant posttranslational modifications within Alzheimer disease, Parkinson disease, and Huntington disease, the review underscores the significance of understanding the interplay among these modifications. Emphasizing 10 key abnormal posttranslational modifications, this study aims to provide a comprehensive framework for investigating neurodegenerative diseases holistically. The insights presented herein shed light on potential therapeutic avenues aimed at modulating posttranslational modifications to mitigate protein aggregation and retard neurodegenerative disease progression.

In Silico Analysis of the Missense Variants of Uncertain Significance of CTNNB1 Gene Reported in GnomAD Database

CTNNB1 pathogenic variants are related to the improper functioning of the WNT/β-catenin pathway, promoting the development of different types of cancer of somatic origin. Bioinformatics analyses of genetic variation are a great tool to understand the possible consequences of these variants on protein structure and function and their probable implication in pathologies. The objective of this study is to describe the impact of the missense variants of uncertain significance (VUS) of the CTNNB1 gene on structure and function of the β-catenin protein. The CTNNB1 variants were obtained from the GnomAD v2.1.1 database; subsequently, a bioinformatic analysis was performed using the VarSome, UCSC Genome Browser, UniProt, the Kinase Library database, and DynaMut2 platforms to evaluate clinical significance, gene conservation, consensus sites for post-translational modifications, and the dynamics and stability of proteins. The GnomAD v2.1.1 database included 826 variants of the CTNNB1 gene, of which 385 were in exons and exon/intron boundaries. Among these variants, 214 were identified as missense, of which 146 were classified as VUS. Notably, 12 variants were in proximity to consensus sites for post-translational modifications (PTMs). The in silico analysis showed a slight tendency towards probably pathogenic for c.59C>T (p.Ala20Val) and c.983T>C (p.Met328Thr) missense VUS. These findings provide possible functional implications of these variants in some types of cancer.

The HABD: Home of All Biological Databases Empowering Biological Research With Cutting-Edge Database Systems

The emergence of computer technologies and computing power has led to the development of several database systems that provide standardized access to vast quantities of data, making it possible to collect, search, index, evaluate, and extract useful knowledge across various fields. The Home of All Biological Databases (HABD) has been established as a continually expanding platform that aims to store, organize, and distribute biological data in a searchable manner, removing all dead and non-accessible data. The platform meticulously categorizes data into various categories, such as COVID-19 Pandemic Database (CO-19PDB), Database relevant to Human Research (DBHR), Cancer Research Database (CRDB), Latest Database of Protein Research (LDBPR), Fungi Databases Collection (FDBC), and many other databases that are categorized based on biological phenomena. It currently provides a total of 22 databases, including 6 published, 5 submitted, and the remaining in various stages of development. These databases encompass a range of areas, including phytochemical-specific and plastic biodegradation databases. HABD is equipped with search engine optimization (SEO) analyzer and Neil Patel tools, which ensure excellent SEO and high-speed value. With timely updates, HABD aims to facilitate the processing and visualization of data for scientists, providing a one-stop-shop for all biological databases. Computer platforms, such as PhP, html, CSS, Java script and Biopython, are used to build all the databases. © 2024 Wiley Periodicals LLC.

Interaction of Soybean (Glycine max (L.) Merr.) Class II ACBPs with MPK2 and SAPK2 Kinases: New Insights into the Regulatory Mechanisms of Plant ACBPs

Plant acyl-CoA-binding proteins (ACBPs) function in plant development and stress responses, with some ACBPs interacting with protein partners. This study tested the interaction between two Class II GmACBPs (Glycine max ACBPs) and seven kinases, using yeast two-hybrid (Y2H) assays and bimolecular fluorescence complementation (BiFC). The results revealed that both GmACBP3.1 and GmACBP4.1 interact with two soybean kinases, a mitogen-activated protein kinase MPK2, and a serine/threonine-protein kinase SAPK2, highlighting the significance of the ankyrin-repeat (ANK) domain in facilitating protein-protein interactions. Moreover, an in vitro kinase assay and subsequent Phos-tag SDS-PAGE determined that GmMPK2 and GmSAPK2 possess the ability to phosphorylate Class II GmACBPs. Additionally, the kinase-specific phosphosites for Class II GmACBPs were predicted using databases. The HDOCK server was also utilized to predict the binding models of Class II GmACBPs with these two kinases, and the results indicated that the affected residues were located in the ANK region of Class II GmACBPs in both docking models, aligning with the findings of the Y2H and BiFC experiments. This is the first report describing the interaction between Class II GmACBPs and kinases, suggesting that Class II GmACBPs have potential as phospho-proteins that impact signaling pathways.

MehNet: a vigesimal-based model by amino acid melting points generates unique ID numbers for protein sequences

The amino acid encoding plays a pivotal role in machine learning-based methods for predicting protein structure and function, as well as in protein mapping techniques. Additionally, the classification of protein sequences presents its own challenges. The current study aims to assign a constant value to each amino acid, thereby creating distinctions among protein sequences. The datasets used in this study were obtained from the UniProt Knowledgebase. Subsequently, these datasets underwent preprocessing steps, and identical sequences were categorized under the same headings. Each amino acid was ranked based on its respective melting point and was assigned a vigesimal digit. These generated vigesimal digits were subsequently converted to decimal values. The centerpiece of this methodology was the melting point hashing table, which was given the name 'MehNet'. Ultimately, each protein sequence was assigned a unique identification number. This approach successfully digitized protein sequences. Notably, experiments involving randomly distributed vigesimal digits for amino acids did not yield results as promising as those achieved with MehNet. The model's classification phase, which utilizes a k-nearest neighbors (kNN) classifier, demonstrates exceptional performance in miscellaneous viral sequences. It achieves high accuracy rates, with an overall accuracy of 99.75%. Notably, it achieves an outstanding accuracy of 99.92% for the Influenza C class, highlighting its ability to distinguish closely related viral sequences.Communicated by Ramaswamy H. Sarma.

Machine learning-based approaches for ubiquitination site prediction in human proteins

Protein ubiquitination is a critical post-translational modification (PTMs) involved in numerous cellular processes. Identifying ubiquitination sites (Ubi-sites) on proteins offers valuable insights into their function and regulatory mechanisms. Due to the cost- and time-consuming nature of traditional approaches for Ubi-site detection, there has been a growing interest in leveraging artificial intelligence for computer-aided Ubi-site prediction. In this study, we collected experimentally verified Ubi-sites of human proteins from the dbPTM database, then conducted comprehensive state-of-the art computational methods along with standard evaluation metrics and a proper validation strategy for Ubi-site prediction. We presented the effectiveness of our framework by comparing ten machine learning (ML) based approaches in three different categories: feature-based conventional ML methods, end-to-end sequence-based deep learning (DL) techniques, and hybrid feature-based DL models. Our results revealed that DL approaches outperformed the classical ML methods, achieving a 0.902 F1-score, 0.8198 accuracy, 0.8786 precision, and 0.9147 recall as the best performance for a DL model using both raw amino acid sequences and hand-crafted features. Interestingly, our experimental results disclosed that the performance of DL methods had a positive correlation with the length of amino acid fragments, suggesting that utilizing the entire sequence can lead to more accurate predictions in future research endeavors. Additionally, we developed a meticulously curated benchmark for Ubi-site prediction in human proteins. This benchmark serves as a valuable resource for future studies, enabling fair and accurate comparisons between different methods. Overall, our work highlights the potential of ML, particularly DL techniques, in predicting Ubi-sites and furthering our knowledge of protein regulation through ubiquitination in cells.