MusiteDeep: a deep-learning based webserver for protein post-translational modification site prediction and visualization

Protein post-translational modifications (PTMs) are chemical modifications of a protein after its translation. Owing to its play an important role in deep understanding of various biological processes and the development of effective drugs, PTM site prediction have become a hot topic in bioinformatics. Recently, many online tools are developed to prediction various types of PTM sites, most of which are based on local sequence and some biological information. However, few of existing tools consider the relations between different PTMs for their prediction task. Here, we develop a web server called PTM-ssMP to predict PTM site, which adopts site-specific modification profile (ssMP) to efficiently extract and encode the information of both proximal PTMs and local sequence simultaneously. In PTM-ssMP we provide efficient prediction of multiple types of PTM site including phosphorylation, lysine acetylation, ubiquitination, sumoylation, methylation, O-GalNAc, O-GlcNAc, sulfation and proteolytic cleavage. To assess the performance of PTM-ssMP, a large number of experimentally verified PTM sites are collected from several sources and used to train and test the prediction models. Our results suggest that ssMP consistently contributes to remarkable improvement of prediction performance. In addition, results of independent tests demonstrate that PTM-ssMP compares favorably with other existing tools for different PTM types. PTM-ssMP is implemented as an online web server with user-friendly interface, which is freely available at http://bioinformatics.ustc.edu.cn/PTM-ssMP/index/.

We present here the recent update of AutoMotif Server (AMS 2.0) that predicts post-translational modification sites in protein sequences. The support vector machine (SVM) algorithm was trained on data gathered in 2007 from various sets of proteins containing experimentally verified chemical modifications of proteins. Short sequence segments around a modification site were dissected from a parent protein, and represented in the training set as binary or profile vectors. The updated efficiency of the SVM classification for each type of modification and the predictive power of both representations were estimated using leave-one-out tests for model of general phosphorylation and for modifications catalyzed by several specific protein kinases. The accuracy of the method was improved in comparison to the previous version of the service (Plewczynski et al., "AutoMotif server: prediction of single residue post-translational modifications in proteins", Bioinformatics 21: 2525-7, 2005). The precision of the updated version reached over 90% for selected types of phosphorylation and was optimized in trade of lower recall value of the classification model. The AutoMotif Server version 2007 is freely available at http://ams2.bioinfo.pl/ . Additionally, the reference dataset for optimization of prediction of phosphorylation sites, collected from the UniProtKB was also provided and can be accessed at http://ams2.bioinfo.pl/data/ .

Protein post-translational modifications (PTMs), particularly phosphorylation, serve as the primary "molecular switches" that orchestrate cellular signaling and drug response. While PTM dysregulation is a hallmark of cancer and neurodegeneration, the lack of standardized, drug-perturbed datasets has hindered the development of predictive models capable of capturing context-dependent PTM responses. Effective predictive modeling must therefore integrate multidimensional data, including the specific drug, dosage, treatment duration, cellular background, and the modified site. However, existing PTM resources remain largely static and fail to capture drug-induced regulation across these critical dimensions. To address this gap, we present DrugPTM-Bench, a curated, large-scale benchmark derived from decryptM-derived dose-dependent PTM measurements, standardizing site-level drug response across 7 cancer cell lines, 27 drugs, and 11,167 proteins. Comprising 99.5% phosphorylation events, the dataset includes six time points, 16 dosage levels, and pEC50 potency values (half-maximal effective concentration). We formulate a classification task to identify upregulated, downregulated, or unchanged PTM sites (following a drug treatment), a critical step in deciphering drug Mechanism of Action (MoA) and target engagement. Our evaluation reveals that in protein-disjoint out-of-distribution (OOD) setting, baseline machine learning and deep learning models struggle to recover minority regulation classes, while standard rebalancing strategies improve recall only at the cost of precision and overall F1-score. These results indicate that current methods do not learn robust decision boundaries between regulated and unchanged PTM events. DrugPTM-Bench provides a phosphoproteomics benchmark for modeling drug-induced PTM regulation in imbalanced biological settings. Beyond classification, DrugPTM-Bench's retention of pEC50 values, drug perturbation profiles, and site-level sequence context enables additional predictive tasks including drug potency regression, mechanism-of-action prediction from PTM fingerprints, and drug-specific PTM site sensitivity ranking, establishing a multi-task benchmark for PTM-centric drug discovery. Ultimately, DrugPTM-Bench establishes a rigorous framework for developing robust, context-aware models to elucidate drug MoA and signaling dynamics.

Phosphorylation ranks among the most crucial post-translational modifications (PTMs), significantly influencing the conformation, activity, and functionality of proteins, and is intricately associated with numerous pathophysiological processes. Therefore, proposing a scientifically valid computational method for precise prediction of phosphorylation sites carries substantial importance. In this study, we propose and evaluate a novel method, PEGN-PSP, based on graph techniques and language models for the prediction of general phosphorylation sites. This method employs an adaptive feature fusion strategy, combining sequence embeddings and pretrained model embeddings to enhance the model's capability in representing features. The use of graph neural network attention mechanisms not only captures local patterns of protein sequences, but also effectively captures long-range dependency information between residues in protein sequences. Independent test results indicate that, in comparison to the current state-of-the-art methods for general phosphorylation site prediction, PEGN-PSP improves the Matthews correlation coefficient for S/T sites by 4.6% and for Y sites by 6.5%. Additionally, PEGN-PSP has good robustness in predicting lysine crotonylation sites, indicating that our method PEGN-PSP has strong potential in predicting other protein post-translational modification sites.

The use of multiple ion chromatograms in on-line HPLC-MS for the characterization of post-translational and chemical modifications of proteins

Most proteins perform their functions after post-translational modifications (PTMs). Post-translational modifications play crucial roles in various cell functions and biological process. Computational identification of PTMs is an important way in contrast to experimental approach with its convenience. This review gave the process for prediction of Post-translational modification sites including benchmark dataset, feature construction, algorithm, evaluation measurement, and online web-server. Feature construction which reflects intrinsic correlation with proteins was emphasized mainly. Seven feature construction methods containing physiochemical properties and sequence information have been illustrated principally. Last nitro tyrosine was an application to design a site-specific predictor. The procedure was clearly described.

In the field of bioinformatics, the protein Post-Translational Modification (PTM) site prediction has been widely studied and Web Information Systems (WIS) has been deployed by researchers for this task. Through a literature review and benchmarking process, we identified the requirements which included quick predictions, efficient memory usage, and input validations. However, no detailed designs have been proposed so far, which may have contributed to some requirements not being implemented in some of the websites. Therefore, we propose a detailed WIS conceptual design, which can be used for predicting the sites of multiple PTM types, equipped with a validation algorithm and compared the usage of various string searching algorithms as well as file storage formats. Experiment results showed that the linear search algorithm is the fastest for this task and storing the protein data in npz format when performing multi-PTMs site prediction can assist in reducing memory usage. The proposed design can be implemented into user-friendly web tools that are both efficient in speed and memory usage in future studies.

Post-translational modification (PTM) sites have become popular for predictor development. However, with the exception of phosphorylation and a handful of other examples, PTMs suffer from a limited number of available training examples and sparsity in protein sequences. Here, proline hydroxylation is taken as an example to compare different methods and evaluate their performance on new experimentally determined sites. As a guide for effective experimental design, predictors require both high specificity and sensitivity. However, the self-reported performance may often not be indicative of prediction quality and detection of new sites is not guaranteed. We have benchmarked seven published hydroxylation site predictors on two newly constructed independent datasets. The self-reported performance is found to widely overestimate the real accuracy measured on independent datasets. No predictor performs better than random on new examples, indicating the refined models do not sufficiently generalize to detect new sites. The number of false positives is high and precision low, in particular for non-collagen proteins whose motifs are not conserved. As hydroxylation site predictors do not generalize for new data, caution is advised when using PTM predictors in the absence of independent evaluations, in particular for highly specific sites involved in signalling.

We present the first large scale study characterizing both N- and O-linked glycosylation in a site-specific manner on hundreds of proteins. We demonstrate that a lectin-affinity fractionation step using wheat germ agglutinin enriches not only peptides carrying intracellular O-GlcNAc, but also those bearing ER/Golgi-derived N- and O-linked carbohydrate structures. Liquid chromatography-MS (LC/MS) analysis with high accuracy precursor mass measurements and high sensitivity ion trap electron-transfer dissociation (ETD) were utilized for structural characterization of glycopeptides. Our results reveal both the identity of the precise sites of glycosylation and information on the oligosaccharide structures possible on these proteins. We report a novel iterative approach that allowed us to interpret the ETD data set directly without making prior assumptions about the nature and distribution of oligosaccharides present in our glycopeptide mixture. Over 2500 unique N- and O-linked glycopeptides were identified on 453 proteins. The extent of microheterogeneity varied extensively, and up to 19 different oligosaccharides were attached at a given site. We describe the presence of the well-known mucin-type structures for O-glycosylation, an EGF-domain-specific fucosylation and a rare O-mannosylation on the transmembrane phosphatase Ptprz1. Finally, we identified three examples of O-glycosylation on tyrosine residues.

Proteins through post translational modifications perform their biological process and cellular functions. Post-translational modifications play important roles in various biological process and cell functions. Identifying the PTMs sites in proteins is very significant to basic research and drug design. Experimental technique to identify post translational modifications is laborious. Computational identification of post translational modifications is a complementary way for its convenience. This review gives the processing to predict post-translational modification sites in proteins including feature construction, algorithms, evaluation measurement, and online webserver. There are two types of post translational modification in proteins. In the prediction of single PTM sites, we transformed it into binary classification learning. While in the prediction of crosstalk PTM sites, we transformed it into multi-label learning. This review summarized the steps on the two issues.

Identification of the sites of post-translational modifications (PTMs) in protein, RNA, and DNA sequences is currently a very hot topic. This is because the information thus obtained is very useful for in-depth understanding the biological processes at the cellular level and for developing effective drugs against major diseases including cancers as well. Although this can be done by means of various experimental techniques, it is both time-consuming and costly to determine the PTM sites purely based on experiments. With the avalanche of biological sequences generated in the post-genomic age, it is highly desired to develop bioinformatics tools for rapidly and effectively identifying the PTM sites. In the last few years, many efforts have been made in this regard, and considerable progresses have been achieved. This review is focused on those prediction methods that have the following two features. (1) They have been developed by strictly observing the 5-steps rule so that they each have a user-friendly web-server for the majority of experimental scientists to easily get their desired data without the need to go through the detailed mathematics involved. (2) Their cornerstones have been based on Pseudo Amino Acid Composition (PseAAC) or Pseudo K-tuple Nucleotide Composition (PseKNC), and hence the prediction quality is generally higher than most of the other PTM prediction methods.

Protein posttranslational modifications (PTMs) perform essential roles in the biological regulation of a cell. PTMs are extremely important because they can change a protein's physical or chemical properties, conformation, activity, cellular location, or stability. In fact, most proteins are altered by the addition or removal of a chemical moiety on either an amino acid or the protein's N- or C-terminus. Some PTMs can be added and removed dynamically as a mechanism for reversibly controlling protein function. Thus, identifying the PTM sites is critical to fully understand the biological roles of any given protein. Mass spectrometry (MS) is a widely used analytical strategy to identify PTMs. We have used an automated two-dimensional liquid chromatography (LC) system coupled with electrospray ionization quadrupole ion-trap MS to identify PTMs for indoleamine 2,3-dioxygenase 1 (IDO1), one of the tryptophan catabolic enzymes. IDO1 promotes immune tolerance by suppressing local T-cell responses under various physiological and pathophysiological conditions, such as pregnancy in mammals, tumor resistance, autoimmunity, and chronic inflammation. Although many studies have demonstrated the biological importance of IDO activity, the PTMs of IDO enzymes remain largely unknown. Only a few important PTMs of IDO1 have been found, such as nitration, N-terminal acetylation, and phosphorylation. In this review, we analyze the PTMs of IDO1 using our two-dimensional LC-MS/MS system, and provide an overview of our current understanding.

A wide variety of protein post-translational modifications (PTMs) decorate cellular proteins, regulating their structure, interactions and ultimately their function. The density of co-occurring PTMs on proteins can be very high, where multiple PTMs can positively or negatively influence each other's actions, nowadays termed PTM crosstalk. In this review we highlight recent progress in the area of PTM crosstalk, whereby we focus on crosstalk between protein phosphorylation and O-GlcNAcylation. These two PTMs largely target identical (i.e. Ser and Thr) amino acids in proteins. Phosphorylation/O-GlcNAcylation crosstalk comes in many flavors, for instance by competition for the same site/residue (reciprocal crosstalk), as well as by modifications influencing each other in proximity or even distal on the protein sequence. PTM crosstalk is observed on the writers of these modifications (i.e. kinases and O-GlcNAc transferase), on the erasers (i.e. phosphatases/O-GlcNAcase) and on the readers and the substrates. Although likely not allinclusive, we describe examples of all these different flavors of crosstalk, and additionally the methods that are emerging to better investigate in particular phosphorylation/O-GlcNAcylation crosstalk.

Post-translational protein modifications can be recognized by B and T lymphocytes and can potentially make "self"-proteins appear foreign to the immune system. Such modifications may directly affect major histocompatibility complex-restricted T cell recognition of processed peptides or may perturb the processing events that generate such peptides. Using the tetanus toxin C fragment protein as a test case, we show that spontaneous deamidation of asparagine residues interferes with processing by the enzyme asparagine endopeptidase (AEP) and contributes to diminished antigen presentation. Deamidation inhibits AEP action either directly, when asparagine residues targeted by AEP are modified, or indirectly, when adjacent Asn residues are deamidated. Thus, deamidation of long-lived self-proteins may qualitatively or quantitatively affect the spectrum of self-peptides displayed to T cells and may thereby contribute to the onset or exacerbation of autoimmune disease.

Regulation of rice seed germination has been shown to mainly occur at post-transcriptional levels, of which the changes on proteome status is a major one. Lysine acetylation and succinylation are two prevalent protein post-translational modifications (PTMs) involved in multiple biological processes, especially for metabolism regulation. To investigate the potential mechanism controlling metabolism regulation in rice seed germination, we performed the lysine acetylation and succinylation analyses simultaneously. Using high-accuracy nano-LC-MS/MS in combination with the enrichment of lysine acetylated or succinylated peptides from digested embryonic proteins of 24 h after imbibition (HAI) rice seed, a total of 699 acetylated sites from 389 proteins and 665 succinylated sites from 261 proteins were identified. Among these modified lysine sites, 133 sites on 78 proteins were commonly modified by two PTMs. The overlapped PTM sites were more likely to be in polar acidic/basic amino acid regions and exposed on the protein surface. Both of the acetylated and succinylated proteins cover nearly all aspects of cellular functions. Ribosome complex and glycolysis/gluconeogenesis-related proteins were significantly enriched in both acetylated and succinylated protein profiles through KEGG enrichment and protein-protein interaction network analyses. The acetyl-CoA and succinyl-CoA metabolism-related enzymes were found to be extensively modified by both modifications, implying the functional interaction between the two PTMs. This study provides a rich resource to examine the modulation of the two PTMs on the metabolism pathway and other biological processes in germinating rice seed.