Using a hybrid neural network architecture for DNA sequence representation: A study on N4-methylcytosine sites.

N4-methylcytosine (4mC) is a modified form of cytosine found in DNA, contributing to epigenetic regulation. It exists in various genomes, including the Rosaceae family encompassing significant fruit crops like apples, cherries, and roses. Previous investigations have examined the distribution and functional implications of 4mC sites within the Rosaceae genome, focusing on their potential roles in gene expression regulation, environmental adaptation, and evolution. This research aims to improve the accuracy of predicting 4mC sites within the genome of Fragaria vesca, a Rosaceae plant species. Building upon the original 4mc-w2vec method, which combines word embedding processing and a convolutional neural network (CNN), we have incorporated additional feature encoding techniques and leveraged pre-trained natural language processing (NLP) models with different deep learning architectures including different forms of CNN, recurrent neural networks (RNN) and long short-term memory (LSTM). Our assessments have shown that the best model is derived from a CNN model using fastText encoding. This model demonstrates enhanced performance, achieving a sensitivity of 0.909, specificity of 0.77, and accuracy of 0.879 on an independent dataset. Furthermore, our model surpasses previously published works on the same dataset, thus showcasing its superior predictive capabilities.

Recent progress in machine learning approaches for predicting carcinogenicity in drug development.

INTRODUCTION: This review explores the transformative impact of machine learning (ML) on carcinogenicity prediction within drug development. It discusses the historical context and recent advancements, emphasizing the significance of ML methodologies in overcoming challenges related to data interpretation, ethical considerations, and regulatory acceptance. AREAS COVERED: The review comprehensively examines the integration of ML, deep learning, and diverse artificial intelligence (AI) approaches in various aspects of drug development safety assessments. It explores applications ranging from early-phase compound screening to clinical trial optimization, highlighting the versatility of ML in enhancing predictive accuracy and efficiency. EXPERT OPINION: Through the analysis of traditional approaches such as in vivo rodent bioassays and in vitro assays, the review underscores the limitations and resource intensity associated with these methods. It provides expert insights into how ML offers innovative solutions to address these challenges, revolutionizing safety assessments in drug development.

An interpretable deep learning model for classifying adaptor protein complexes from sequence information.

Adaptor proteins (APs) are a family of proteins that aids in intracellular membrane trafficking, and their impairments or defects are closely related to various disorders. Traditional methods to identify and classify APs require time and complex techniques, which were then advanced by machine learning and computational approaches to facilitate the APs recognition task. However, most studies focused on recognizing separate ones in the APs family or the APs in general with non-APs, lacking one comprehensive strategy to distinguish the complexes of AP subtypes. Herein, we proposed a novel method to implement one novel task as discriminating the AP complexes in the APs family, utilizing an interpretable deep neural network architecture on sequence-based encoding features. This work also introduced a benchmark data set of AP complexes originating from the UniProt and GeneOntology databases. To assess the robustness of our proposed method, we compared our performance to various machine learning algorithms and feature extraction strategies. Furthermore, the interpretation of the model's prediction performance was implemented using t-distributed stochastic neighbor embedding (t-SNE), uniform manifold approximation and projection (UMAP), and SHapley Additive exPlanations (SHAP) analysis to show the distribution of AP complexes on optimal features. The promising performance of our architecture can assist scientists not only in AP complexes distinction but also in general protein sequences. Moreover, we have also made our work publicly on GitHub https://github.com/khanhlee/adaptor-dnn.

BERT-Promoter: An improved sequence-based predictor of DNA promoter using BERT pre-trained model and SHAP feature selection.

A promoter is a sequence of DNA that initializes the process of transcription and regulates whenever and wherever genes are expressed in the organism. Because of its importance in molecular biology, identifying DNA promoters are challenging to provide useful information related to its functions and related diseases. Several computational models have been developed to early predict promoters from high-throughput sequencing over the past decade. Although some useful predictors have been proposed, there remains short-falls in those models and there is an urgent need to enhance the predictive performance to meet the practice requirements. In this study, we proposed a novel architecture that incorporated transformer natural language processing (NLP) and explainable machine learning to address this problem. More specifically, a pre-trained Bidirectional Encoder Representations from Transformers (BERT) model was employed to encode DNA sequences, and SHapley Additive exPlanations (SHAP) analysis served as a feature selection step to look at the top-rank BERT encodings. At the last stage, different machine learning classifiers were implemented to learn the top features and produce the prediction outcomes. This study not only predicted the DNA promoters but also their activities (strong or weak promoters). Overall, several experiments showed an accuracy of 85.5 % and 76.9 % for these two levels, respectively. Our performance showed a superiority to previously published predictors on the same dataset in most measurement metrics. We named our predictor as BERT-Promoter and it is freely available at https://github.com/khanhlee/bert-promoter.

SuccSite: Incorporating Amino Acid Composition and Informative k-spaced Amino Acid Pairs to Identify Protein Succinylation Sites.

Protein succinylation is a biochemical reaction in which a succinyl group (-CO-CH2-CH2-CO-) is attached to the lysine residue of a protein molecule. Lysine succinylation plays important regulatory roles in living cells. However, studies in this field are limited by the difficulty in experimentally identifying the substrate site specificity of lysine succinylation. To facilitate this process, several tools have been proposed for the computational identification of succinylated lysine sites. In this study, we developed an approach to investigate the substrate specificity of lysine succinylated sites based on amino acid composition. Using experimentally verified lysine succinylated sites collected from public resources, the significant differences in position-specific amino acid composition between succinylated and non-succinylated sites were represented using the Two Sample Logo program. These findings enabled the adoption of an effective machine learning method, support vector machine, to train a predictive model with not only the amino acid composition, but also the composition of k-spaced amino acid pairs. After the selection of the best model using a ten-fold cross-validation approach, the selected model significantly outperformed existing tools based on an independent dataset manually extracted from published research articles. Finally, the selected model was used to develop a web-based tool, SuccSite, to aid the study of protein succinylation. Two proteins were used as case studies on the website to demonstrate the effective prediction of succinylation sites. We will regularly update SuccSite by integrating more experimental datasets. SuccSite is freely accessible at http://csb.cse.yzu.edu.tw/SuccSite/.

SNARE-CNN: a 2D convolutional neural network architecture to identify SNARE proteins from high-throughput sequencing data.

Deep learning has been increasingly and widely used to solve numerous problems in various fields with state-of-the-art performance. It can also be applied in bioinformatics to reduce the requirement for feature extraction and reach high performance. This study attempts to use deep learning to predict SNARE proteins, which is one of the most vital molecular functions in life science. A functional loss of SNARE proteins has been implicated in a variety of human diseases (e.g., neurodegenerative, mental illness, cancer, and so on). Therefore, creating a precise model to identify their functions is a crucial problem for understanding these diseases, and designing the drug targets. Our SNARE-CNN model which uses two-dimensional convolutional neural networks and position-specific scoring matrix profiles could identify SNARE proteins with achieved sensitivity of 76.6%, specificity of 93.5%, accuracy of 89.7%, and MCC of 0.7 in cross-validation dataset. We also evaluate the performance of our model via an independent dataset and the result shows that we are able to solve the overfitting problem. Compared with other state-of-the-art methods, this approach achieved significant improvement in all of the metrics. Throughout the proposed study, we provide an effective model for identifying SNARE proteins and a basis for further research that can apply deep learning in bioinformatics, especially in protein function prediction. SNARE-CNN are freely available at https://github.com/khanhlee/snare-cnn.

A New Scheme to Characterize and Identify Protein Ubiquitination Sites.

Protein ubiquitination, involving the conjugation of ubiquitin on lysine residue, serves as an important modulator of many cellular functions in eukaryotes. Recent advancements in proteomic technology have stimulated increasing interest in identifying ubiquitination sites. However, most computational tools for predicting ubiquitination sites are focused on small-scale data. With an increasing number of experimentally verified ubiquitination sites, we were motivated to design a predictive model for identifying lysine ubiquitination sites for large-scale proteome dataset. This work assessed not only single features, such as amino acid composition (AAC), amino acid pair composition (AAPC) and evolutionary information, but also the effectiveness of incorporating two or more features into a hybrid approach to model construction. The support vector machine (SVM) was applied to generate the prediction models for ubiquitination site identification. Evaluation by five-fold cross-validation showed that the SVM models learned from the combination of hybrid features delivered a better prediction performance. Additionally, a motif discovery tool, MDDLogo, was adopted to characterize the potential substrate motifs of ubiquitination sites. The SVM models integrating the MDDLogo-identified substrate motifs could yield an average accuracy of 68.70 percent. Furthermore, the independent testing result showed that the MDDLogo-clustered SVM models could provide a promising accuracy (78.50 percent) and perform better than other prediction tools. Two cases have demonstrated the effective prediction of ubiquitination sites with corresponding substrate motifs.

UbiNet: an online resource for exploring the functional associations and regulatory networks of protein ubiquitylation.

Protein ubiquitylation catalyzed by E3 ubiquitin ligases are crucial in the regulation of many cellular processes. Owing to the high throughput of mass spectrometry-based proteomics, a number of methods have been developed for the experimental determination of ubiquitylation sites, leading to a large collection of ubiquitylation data. However, there exist no resources for the exploration of E3-ligase-associated regulatory networks of for ubiquitylated proteins in humans. Therefore, the UbiNet database was developed to provide a full investigation of protein ubiquitylation networks by incorporating experimentally verified E3 ligases, ubiquitylated substrates and protein-protein interactions (PPIs). To date, UbiNet has accumulated 43 948 experimentally verified ubiquitylation sites from 14 692 ubiquitylated proteins of humans. Additionally, we have manually curated 499 E3 ligases as well as two E1 activating and 46 E2 conjugating enzymes. To delineate the regulatory networks among E3 ligases and ubiquitylated proteins, a total of 430 530 PPIs were integrated into UbiNet for the exploration of ubiquitylation networks with an interactive network viewer. A case study demonstrated that UbiNet was able to decipher a scheme for the ubiquitylation of tumor proteins p63 and p73 that is consistent with their functions. Although the essential role of Mdm2 in p53 regulation is well studied, UbiNet revealed that Mdm2 and additional E3 ligases might be implicated in the regulation of other tumor proteins by protein ubiquitylation. Moreover, UbiNet could identify potential substrates for a specific E3 ligase based on PPIs and substrate motifs. With limited knowledge about the mechanisms through which ubiquitylated proteins are regulated by E3 ligases, UbiNet offers users an effective means for conducting preliminary analyses of protein ubiquitylation. The UbiNet database is now freely accessible via http://csb.cse.yzu.edu.tw/UbiNet/ The content is regularly updated with the literature and newly released data.Database URL: http://csb.cse.yzu.edu.tw/UbiNet/.

Characterization and identification of ubiquitin conjugation sites with E3 ligase recognition specificities.

BACKGROUND: In eukaryotes, ubiquitin-conjugation is an important mechanism underlying proteasome-mediated degradation of proteins, and as such, plays an essential role in the regulation of many cellular processes. In the ubiquitin-proteasome pathway, E3 ligases play important roles by recognizing a specific protein substrate and catalyzing the attachment of ubiquitin to a lysine (K) residue. As more and more experimental data on ubiquitin conjugation sites become available, it becomes possible to develop prediction models that can be scaled to big data. However, no development that focuses on the investigation of ubiquitinated substrate specificities has existed. Herein, we present an approach that exploits an iteratively statistical method to identify ubiquitin conjugation sites with substrate site specificities. RESULTS: In this investigation, totally 6259 experimentally validated ubiquitinated proteins were obtained from dbPTM. After having filtered out homologous fragments with 40% sequence identity, the training data set contained 2658 ubiquitination sites (positive data) and 5532 non-ubiquitinated sites (negative data). Due to the difficulty in characterizing the substrate site specificities of E3 ligases by conventional sequence logo analysis, a recursively statistical method has been applied to obtain significant conserved motifs. The profile hidden Markov model (profile HMM) was adopted to construct the predictive models learned from the identified substrate motifs. A five-fold cross validation was then used to evaluate the predictive model, achieving sensitivity, specificity, and accuracy of 73.07%, 65.46%, and 67.93%, respectively. Additionally, an independent testing set, completely blind to the training data of the predictive model, was used to demonstrate that the proposed method could provide a promising accuracy (76.13%) and outperform other ubiquitination site prediction tool. CONCLUSION: A case study demonstrated the effectiveness of the characterized substrate motifs for identifying ubiquitination sites. The proposed method presents a practical means of preliminary analysis and greatly diminishes the total number of potential targets required for further experimental confirmation. This method may help unravel their mechanisms and roles in E3 recognition and ubiquitin-mediated protein degradation.