Predicting lysine methylation sites using a convolutional neural network

All large scale LC-MS/MS post-translational methylation site discovery experiments require methylpeptide spectrum matches (methyl-PSMs) to be identified at acceptably low false discovery rates (FDRs). To meet estimated methyl-PSM FDRs, methyl-PSM filtering criteria are often determined using the target-decoy approach. The efficacy of this methyl-PSM filtering approach has, however, yet to be thoroughly evaluated. Here, we conduct a systematic analysis of methyl-PSM FDRs across a range of sample preparation workflows (each differing in their exposure to the alcohols methanol and isopropyl alcohol) and mass spectrometric instrument platforms (each employing a different mode of MS/MS dissociation). Through (13)CD3-methionine labeling (heavy-methyl SILAC) of Saccharomyces cerevisiae cells and in-depth manual data inspection, accurate lists of true positive methyl-PSMs were determined, allowing methyl-PSM FDRs to be compared with target-decoy approach-derived methyl-PSM FDR estimates. These results show that global FDR estimates produce extremely unreliable methyl-PSM filtering criteria; we demonstrate that this is an unavoidable consequence of the high number of amino acid combinations capable of producing peptide sequences that are isobaric to methylated peptides of a different sequence. Separate methyl-PSM FDR estimates were also found to be unreliable due to prevalent sources of false positive methyl-PSMs that produce high peptide identity score distributions. Incorrect methylation site localizations, peptides containing cysteinyl-S-β-propionamide, and methylated glutamic or aspartic acid residues can partially, but not wholly, account for these false positive methyl-PSMs. Together, these results indicate that the target-decoy approach is an unreliable means of estimating methyl-PSM FDRs and methyl-PSM filtering criteria. We suggest that orthogonal methylpeptide validation (e.g. heavy-methyl SILAC or its offshoots) should be considered a prerequisite for obtaining high confidence methyl-PSMs in large scale LC-MS/MS methylation site discovery experiments and make recommendations on how to reduce methyl-PSM FDRs in samples not amenable to heavy isotope labeling. Data are available via ProteomeXchange with the data identifier PXD002857.

N6-methyladenine (6mA) is an important DNA methylation modification that serves a crucial function in various biological activities. Accurate prediction of 6mA sites is essential for elucidating its biological function and underlying mechanism. Although existing methods have achieved great success, there remains a pressing need for improved prediction accuracy and generalization cap ability across diverse species. This study aimed to develop a robust method to address these challenges. We proposed HD-6mAPred, a hybrid deep learning model that combines bidirectional gated recurrent unit (BiGRU), convolutional neural network (CNN) and attention mechanism, along with various DNA sequence coding schemes. Firstly, DNA sequences were encoded using four different ways: one-hot encoding, electron-ion interaction pseudo-potential (EIIP), enhanced nucleic acid composition (ENAC) and nucleotide chemical properties (NCP). Secondly, a hold-out search strategy was employed to identify the optimal features or feature combinations for both BiGRU and CNN. Finally, the attention mechanism was introduced to weigh the importance of features derived from the BiGRU and CNN. A series of experiments on the Rosaceae, rice and Arabidopsis datasets were conducted to demonstrate the superiority of HD-6mAPred. In Rosaceae, the HD-6mAPred model achieved excellent performance: accuracy (ACC) of 0.996, Matthew correlation coefficient (MCC) of 0.993, sensitivity (SN) and specificity (SP) of 0.995 and 0.998, respectively. In rice, the evaluation metrics are 0.952 (ACC), 0.905 (MCC), 0.955 (SN), and 0.949 (SP). In Arabidopsis, the corresponding metrics are 0.937 (ACC), 0.875 (MCC), 0.927 (SN), and 0.948 (SP). Compared to existing methods, these results demonstrate that HD-6mAPred achieves state-of-the-art performance in predicting 6mA sites across three plant species. Furthermore, HD-6mAPred not only improves the accuracy of 6mA site prediction, but also shows excellent generalization capability across species. The source code utilized in this study is publicly accessible at https://doi.org/10.5281/zenodo.15355131.

The leather industry plays a crucial role in the production of high-quality leather products to remain competitive in the market. Due to various reasons such as material handling and turning, leather defects can appear at different stages of the production process. This paper proposed deep learning techniques, specifically transfer learning, deep convolutional neural networks (D-CNN), and ensemble learning for automating leather defect detection and classification. The dataset was taken from the open source Kaggle. The dataset was preprocessed from the beginning. However, we performed Error Level Analysis (ELA) to detect possible changes or fraud in the image. We also used data augmentation techniques (resizing, rescaling, flipping, rotation, zooming, and contrasting) on the dataset to improve the model performance. This study investigated deep learning architectures for automatic leather defect detection, comparing them with seven deep convolutional neural networks (D-CNN) models and transfer learning architectures. Furthermore, we introduce and evaluate two novel ensemble models: MER built on transfer learning architecture (MobileNetV2, EfficientNetB0, and ResNet50) and AZL built on D-CNN architecture (AlexNet, ZfNet, and LeNet). A key contribution of this study is the demonstration of significant performance improvements achieved by strategically integrating squeeze and excitation and label smoothing techniques into transfer learning, D-CNN, and ensemble frameworks. Our proposed MER ensemble method achieves a state-of-the-art of 96.39% for leather defect detection. The proposed model was evaluated using performance metrics including precision, recall, F1-score, AUC, ROC, and Matthews Correlation Coefficient (MCC). Additionally, Grad-CAM was employed for explainable AI to visualize and interpret the model’s decision-making process. Comparative analysis against separate transfer learning and D-CNN models fails to establish the superior performance of our MER ensemble, which highlights the potential to significantly advance automated quality inspection in leather production. This study provides insights into evaluating the best deep learning techniques for leather defect classification, paving the way for more accurate, efficient, and reliable industrial applications. Proposes a robust deep learning-based framework for automated leather defect detection using transfer learning, deep CNNs, and two novel ensemble models (MER and AZL), enhanced with squeeze-and-excitation and label smoothing techniques. Achieves state-of-the-art performance with the proposed MER ensemble model, achieve 96.39% accuracy, and demonstrates superior results across precision, recall, F1-score, AUC–ROC, and MCC. Integrates Error Level Analysis (ELA), extensive data augmentation, and Grad-CAM–based explainable AI to ensure data reliability, improve generalization, and provide transparent visual interpretation of defect regions for industrial quality inspection.

Protein lysine methylation is a critical and dynamic post-translational modification that can regulate protein stability and function. This post-translational modification is regulated by lysine methyltransferases and lysine demethylases. Recent studies using mass-spectrometric techniques have revealed that in addition to histones, a great number of transcription factors are also methylated, often at multiple sites and to different degrees (mono-, di-, trimethyl lysine). The biomedical significance of transcription factor methylation in human diseases, including cancer, has been explored recently. Some studies have demonstrated that interfering with transcription factor lysine methylation both in vitro and in vivo can inhibit cancer cell proliferation, thereby reversing tumor progression. The inhibitors targeting lysine methyltransferases and lysine demethylases have been under development for the past two decades, and may be used as potential anticancer agents in the clinic. In this review, we focus on the current findings of transcription factor lysine methylation, and the effects on both transcriptional activity and target gene expression. We outlined the biological significance of transcription factor lysine methylation on tumor progression and highlighted its clinical value in cancer therapy.

Protein lysine methylation is a post-translational modification (PTM) that promotes protein complex formation to regulate DNA replication, gene expression, and repair mechanisms. The Su(Var)3–9, Enhancer-of-zeste and Trithorax (SET) and Myeloid, Nervy, and DEAF-1 (MYND) domain-containing proteins SMYD are lysine methyltransferases that catalyze the methylation of various histone and non-histone proteins. There are five members of this SMYD family, and all of these have conserved SET and MYND domains. The SET domain is divided into two segments by the MYND domain (the S-sequence and a core SET domain). SMYD Family performs a key role in numerous biological functions, including growth, development, apoptosis, and proliferation. SMYD family members are associated with skeletal and cardiac muscle physiology and pathology. Several studies have shown that aberrant lysine methylation plays a significant role in oncogenesis. Recently, the SMYD family has gained importance for its role in various mechanisms involved in cancer development and progressions, such as methylation and modification of tumor suppressor proteins (p53 and pRb), transcriptional factors (STAT3, NF-κB), nuclear proteins (PARP1), chaperons (Hsp90), protein kinases (MAPK, ERK), and cell cycle regulatory proteins (CDKN2). SMYD family proteins drive oncogenesis, lead the way to metastasis, and develop chemoresistance, allowing cancer cells to grow, invade the neighboring tissues, and resist therapeutics. In this review, we summarize SMYD family members' role in different cancers by focusing on their histone and non-histone methylation targets and illustrating the mechanism of SMYD family-mediated oncogenesis.

BackgroundSignaling via protein lysine methylation has been proposed to play a central role in the regulation of many physiologic and pathologic programs. In contrast to other post-translational modifications such as phosphorylation, proteome-wide approaches to investigate lysine methylation networks do not exist.ResultsIn the current study, we used the ProtoArray® platform, containing over 9,500 human proteins, and developed and optimized a system for proteome-wide identification of novel methylation events catalyzed by the protein lysine methyltransferase (PKMT) SETD6. This enzyme had previously been shown to methylate the transcription factor RelA, but it was not known whether SETD6 had other substrates. By using two independent detection approaches, we identified novel candidate substrates for SETD6, and verified that all targets tested in vitro and in cells were genuine substrates.ConclusionsWe describe a novel proteome-wide methodology for the identification of new PKMT substrates. This technological advance may lead to a better understanding of the enzymatic activity and substrate specificity of the large number (more than 50) PKMTs present in the human proteome, most of which are uncharacterized.

A cross-immunity model for influenza A(H3N2) based on convolutional neural networks (CNNs) was developed and validated under temporally structured conditions that mimic real-world forecasting. Antigenic distance was derived from hemagglutination inhibition (HI) titers. The model was trained on WHO data (2011-2023) and tested in a time-split fashion on independent recent data (2022-2024). Hemagglutinin sequences (HA/HA1) were encoded into 3D tensors using five physicochemical indices from AAindex. Two- and three-layer CNN architectures were tested. Performance was evaluated using Accuracy, Sensitivity, Specificity, and Matthews Correlation Coefficient (MCC) with 95% confidence intervals. Validation on the classic Smith's dataset showed high accuracy (Accuracy = 0.9996, MCC = 0.9964), serving as a necessary sanity check. Testing on current data yielded lower but robust results (Accuracy: 0.73-0.81, MCC: 0.48-0.60), reflecting real-world forecasting complexity. ROC analysis confirmed the strong discriminative ability (AUC ≥ 0.805) and good calibration (Brier scores ≤ 0.192). The three-layer CNN demonstrated greater robustness on challenging data. This CNN model is an effective tool for assessing influenza A(H3N2) antigenic distances and holds promise for integration into epidemiological models to aid vaccine strain selection. Further accuracy improvements may arise from modeling the structural impact of amino acid substitutions and polyclonal immune responses.

Robust multi-class model constructed for rapid quality control of Cordyceps sinensis

Cell-penetrating peptides (CPPs) are short peptides capable of penetrating cell membranes, making them valuable for drug delivery and intracellular targeting. Accurate prediction of CPPs can streamline experimental validation in the lab. This study aims to assess pretrained protein language models (pLMs) for their effectiveness in representing CPPs and develop a reliable model for CPP classification. We evaluated peptide embeddings generated from BEPLER, CPCProt, SeqVec, various ESM variants (ESM, ESM-2 with expanded feature set, ESM-1b, and ESM-1v), ProtT5-XL UniRef50, ProtT5-XL BFD, and ProtBERT. We developed pLM4CCPs, a novel deep learning architecture using convolutional neural networks (CNNs) as the classifier for binary classification of CPPs. pLM4CCPs demonstrated superior performance over existing state-of-the-art CPP prediction models, achieving improvements in accuracy (ACC) by 4.9-5.5%, Matthews correlation coefficient (MCC) by 9.3-10.2%, and sensitivity (Sn) by 14.1-19.6%. Among all the tested models, ESM-1280 and ProtT5-XL BFD demonstrated the highest overall performance on the kelm data set. ESM-1280 achieved an ACC of 0.896, an MCC of 0.796, a Sn of 0.844, and a specificity (Sp) of 0.978. ProtT5-XL BFD exhibited superior performance with an ACC of 0.901, an MCC of 0.802, an Sn of 0.885, and an Sp of 0.917. pLM4CCPs combine predictions from multiple models to provide a consensus on whether a given peptide sequence is classified as a CPP or non-CPP. This approach will enhance prediction reliability by leveraging the strengths of each individual model. A user-friendly web server for bioactivity predictions, along with data sets, is available at https://ry2acnp6ep.us-east-1.awsapprunner.com. The source code and protocol for adapting pLM4CPPs can be accessed on GitHub at https://github.com/drkumarnandan/pLM4CPPs. This platform aims to advance CPP prediction and peptide functionality modeling, aiding researchers in exploring peptide functionality effectively.

We study and propose, for the first time, an autonomous classification of the estrous cycle (the reproductive cycle in rats). This cycle consists of 4 stages: Proestrus, Estrus, Metestrus and Diestrus. The short duration of the cycle in rats makes them an ideal model for research about changes that occur during the reproductive cycle. Classification is based on the cytology shown by vaginal smear. For this reason, we use manual and automatic feature extraction; these features are classified with support vector machines, multilayer perceptron networks and convolutional neural networks. A dataset of 412 images of estrous cycle was used. It was divided into two sets. The first contains all four stages, the second contains two classes. The first class is formed by the stages Proestrus and Estrus and the second class is formed by the stages Metestrus and Diestrus. The two sets were formed to solve the main problems, the research of the reproductive cycle and the reproduction control of rodents. For the first set, we obtained an 82% of validation accuracy and 98.38% of validation accuracy for the second set using convolutional neural networks. The results were validated through cross validation and F1 metric.

Because it is so difficult to distinguish handwritten digits, digit identification is one of the most critical applications in computer vision. This is one of the reasons why it is so tough. The field of handwritten character recognition is one in which a great deal of application of numerous deep learning models has occurred. The startling parallels that can be drawn between deep learning and the brain are primarily responsible for its meteoric rise in popularity. In this study, the Artificial Neural Network and the Convolutional Neural Network, two of the most used Deep Learning algorithms, were investigated with an eye toward the recognition process's feature extraction and classification phases. With the assistance of the categorical cross-entropy loss and the ADAM optimizer, the models were trained on the MNIST dataset. Backpropagation and gradient descent are the two methods utilized during the training process of neural networks that contain reLU activations and carry out automatic feature extraction. In computer vision, one of the most common and widely used classifiers is the Convolution Neural Network, sometimes referred to as ConvNets or Convolutional neural networks. This network is used for the recognition and categorization of images.

Limited experimental data remains a key challenge in applying machine learning to drug discovery, particularly for cancer-related targets. In this study, we present a data-efficient active meta-deep learning framework to predict mitogen-activated protein kinase 1 (MAPK1) inhibitors, which are promising candidates for cancer-related therapies. Our approach integrates active learning (AL) with a meta-model that combines four deep architectures: a convolutional neural network, an attention, a graph convolutional network, and a graph neural network-attention, trained on molecular descriptors and graph-based representations. These models generate four probability-based features that feed into an attention-based meta-learner, improving predictive performance by 5.12% in the area under the precision-recall curve (AUPRC) and 5.48% in the Matthews correlation coefficient (MCC) using only 10% of the training data. Among the AL sampling strategies evaluated, entropy sampling showed competitive performance in selecting informative molecules for model improvement. Overall, our framework achieves an AUPRC of 0.835 ± 0.017 and MCC of 0.817 ± 0.017, on par with a traditional training method despite using only 26.7% of the training data. Compared to a conventional random forest model trained on brute-force, a 100% full training set, our approach shows a 10.6% improvement in AUPRC and modest gains in MCC, confirming the effectiveness of the proposed framework. Under severe class imbalance, balanced accuracy steadily increased across AL iterations, reaching values greater than 0.85 at the final iteration for all uncertainty-driven strategies. Molecular docking confirmed successful prioritization of the top four predicted compounds. Evaluation on an external MAPK1 data set demonstrated generalizability, with our approach achieving an AUPRC of 0.818 and an MCC of 0.403, comparable to the independent test set. These results highlight the potential of combining intelligent data selection with deep learning architectures through the meta-model to accelerate predictive performance in data-scarce drug discovery. Scientific contribution: This study contributes a novel, data-efficient active meta-deep learning framework for predicting MAPK1 inhibitors, addressing the challenge of limited experimental data in a cancer-specific target. By integrating AL with a meta-model composed of four deep architectures, the approach significantly enhances the predictive performance using only a fraction of the training data. The framework achieves superior metrics compared to traditional training methods, highlighting its potential to accelerate drug discovery in data-scarce settings.

Explainable 3D VGG-style convolutional neural network for pediatric hydrocephalus detection on computed tomography: A segmentation-free and fully volumetric deep learning framework.

Genres are one of the key features that categorize music based on specific series of patterns. However, the Arabic music content on the web is poorly defined into its genres, making the automatic classification of Arabic audio genres challenging. For this reason, in this research, our objective is first to construct a well-annotated dataset of five of the most well-known Arabic music genres, which are: Eastern Takht, Rai, Muwashshah, the poem, and Mawwal, and finally present a comprehensive empirical comparison of deep Convolutional Neural Networks (CNNs) architectures on Arabic music genres classification. In this work, to utilize CNNs to develop a practical classification system, the audio data is transformed into a visual representation (spectrogram) using Short Time Fast Fourier Transformation (STFT), then several audio features are extracted using Mel Frequency Cepstral Coefficients (MFCC). Performance evaluation of classifiers is measured with the accuracy score, time to build, and Matthew's correlation coefficient (MCC). The concluded results demonstrated that AlexNet is considered among the top-performing five CNNs classifiers studied: LeNet5, AlexNet, VGG, ResNet-50, and LSTM-CNN, with an overall accuracy of 96%.

Machine learning based predictions of protein–protein interactions (PPIs) could provide valuable insights into protein functions, disease occurrence, and therapy design on a large scale. The intensive feature engineering in most of these methods makes the prediction task more tedious and trivial. The emerging deep learning technology enabling automatic feature engineering is gaining great success in various fields. However, the over-fitting and generalization of its models are not yet well investigated in most scenarios. Here, we present a deep neural network framework (DNN-PPI) for predicting PPIs using features learned automatically only from protein primary sequences. Within the framework, the sequences of two interacting proteins are sequentially fed into the encoding, embedding, convolution neural network (CNN), and long short-term memory (LSTM) neural network layers. Then, a concatenated vector of the two outputs from the previous layer is wired as the input of the fully connected neural network. Finally, the Adam optimizer is applied to learn the network weights in a back-propagation fashion. The different types of features, including semantic associations between amino acids, position-related sequence segments (motif), and their long- and short-term dependencies, are captured in the embedding, CNN and LSTM layers, respectively. When the model was trained on Pan’s human PPI dataset, it achieved a prediction accuracy of 98.78% at the Matthew’s correlation coefficient (MCC) of 97.57%. The prediction accuracies for six external datasets ranged from 92.80% to 97.89%, making them superior to those achieved with previous methods. When performed on Escherichia coli, Drosophila, and Caenorhabditis elegans datasets, DNN-PPI obtained prediction accuracies of 95.949%, 98.389%, and 98.669%, respectively. The performances in cross-species testing among the four species above coincided in their evolutionary distances. However, when testing Mus Musculus using the models from those species, they all obtained prediction accuracies of over 92.43%, which is difficult to achieve and worthy of note for further study. These results suggest that DNN-PPI has remarkable generalization and is a promising tool for identifying protein interactions.