iDNA-ABF: multi-scale deep biological language learning model for the interpretable prediction of DNA methylations

MGDDI: A multi-scale graph neural networks for drug-drug interaction prediction

Drug-drug interaction (DDI) prediction is crucial for identifying interactions within drug combinations, especially adverse effects due to physicochemical incompatibility. While current methods have made strides in predicting adverse drug interactions, limitations persist. Most methods rely on handcrafted features, restricting their applicability. They predominantly extract information from individual drugs, neglecting the importance of interaction details between drug pairs. To address these issues, we propose MGDDI, a graph neural network-based model for predicting potential adverse drug interactions. Notably, we use a multiscale graph neural network (MGNN) to learn drug molecule representations, addressing substructure size variations and preventing gradient issues. For capturing interaction details between drug pairs, we integrate a substructure interaction learning module based on attention mechanisms. Our experimental results demonstrate MGDDI's superiority in predicting adverse drug interactions, offering a solution to current methodological limitations.

RDscan: Extracting RNA-disease relationship from the literature based on pre-training model

With the rapid advancements in molecular biology and genomics, a multitude of connections between RNA and diseases has been unveiled, making the efficient and accurate extraction of RNA-disease (RD) relationships from extensive biomedical literature crucial for advancing research in this field. This study introduces RDscan, a novel text mining method developed based on the pre-training and fine-tuning strategy, aimed at automatically extracting RD-related information from a vast corpus of literature using pre-trained biomedical large language models (LLM). Initially, we constructed a dedicated RD corpus by manually curating from literature, comprising 2,082 positive and 2,000 negative sentences, alongside an independent test dataset (comprising 500 positive and 500 negative sentences) for training and evaluating RDscan. Subsequently, by fine-tuning the Bioformer and BioBERT pre-trained models, RDscan demonstrated exceptional performance in text classification and named entity recognition (NER) tasks. In 5-fold cross-validation, RDscan significantly outperformed traditional machine learning methods (Support Vector Machine, Logistic Regression and Random Forest). In addition, we have developed an accessible webserver that assists users in extracting RD relationships from text. In summary, RDscan represents the first text mining tool specifically designed for RD relationship extraction, and is poised to emerge as an invaluable tool for researchers dedicated to exploring the intricate interactions between RNA and diseases. Webserver of RDscan is free available at https://cellknowledge.com.cn/RDscan/.

Moss-m7G: A Motif-Based Interpretable Deep Learning Method for RNA N7-Methlguanosine Site Prediction

N-7methylguanosine (m7G) modification plays a crucial role in various biological processes and is closely associated with the development and progression of many cancers. Accurate identification of m7G modification sites is essential for understanding their regulatory mechanisms and advancing cancer therapy. Previous studies often suffered from insufficient research data, underutilization of motif information, and lack of interpretability. In this work, we designed a novel motif-based interpretable method for m7G modification site prediction, called Moss-m7G. This approach enables the analysis of RNA sequences from a motif-centric perspective. Our proposed word-detection module and motif-embedding module within Moss-m7G extract motif information from sequences, transforming the raw sequences from base-level into motif-level and generating embeddings for these motif sequences. Compared with base sequences, motif sequences contain richer contextual information, which is further analyzed and integrated through the Transformer model. We constructed a comprehensive m7G data set to implement the training and testing process to address the data insufficiency noted in prior research. Our experimental results affirm the effectiveness and superiority of Moss-m7G in predicting m7G modification sites. Moreover, the introduction of the word-detection module enhances the interpretability of the model, providing insights into the predictive mechanisms.

Identification of DNase I hypersensitive sites in the human genome by multiple sequence descriptors

DNase I hypersensitive sites (DHSs) are chromatin regions highly sensitive to DNase I enzymes. Studying DHSs is crucial for understanding complex transcriptional regulation mechanisms and localizing cis-regulatory elements (CREs). Numerous studies have indicated that disease-related loci are often enriched in DHSs regions, underscoring the importance of identifying DHSs. Although wet experiments exist for DHSs identification, they are often labor-intensive. Therefore, there is a strong need to develop computational methods for this purpose. In this study, we used experimental data to construct a benchmark dataset. Seven feature extraction methods were employed to capture information about human DHSs. The F-score was applied to filter the features. By comparing the prediction performance of various classification algorithms through five-fold cross-validation, random forest was proposed to perform the final model construction. The model could produce an overall prediction accuracy of 0.859 with an AUC value of 0.837. We hope that this model can assist scholars conducting DNase research in identifying these sites.

The Prediction of Recombination Hotspot Based on Automated Machine Learning

Meiotic recombination plays a pivotal role in genetic evolution. Genetic variation induced by recombination is a crucial factor in generating biodiversity and a driving force for evolution. At present, the development of recombination hotspot prediction methods has encountered challenges related to insufficient feature extraction and limited generalization capabilities. This paper focused on the research of recombination hotspot prediction methods. We explored deep learning-based recombination hotspot prediction and scrutinized the shortcomings of prevalent models in addressing the challenge of recombination hotspot prediction. To addressing these deficiencies, an automated machine learning approach was utilized to construct recombination hotspot prediction model. The model combined sequence information with physicochemical properties by employing TF-IDF-Kmer and DNA composition components to acquire more effective feature data. Experimental results validate the effectiveness of the feature extraction method and automated machine learning technology used in this study. The final model was validated on three distinct datasets and yielded accuracy rates of 97.14%, 79.71%, and 98.73%, surpassing the current leading models by 2%, 2.56%, and 4%, respectively. In addition, we incorporated tools such as SHAP and AutoGluon to analyze the interpretability of black-box models, delved into the impact of individual features on the results, and investigated the reasons behind misclassification of samples. Finally, an application of recombination hotspot prediction was established to facilitate easy access to necessary information and tools for researchers. The research outcomes of this paper underscore the enormous potential of automated machine learning methods in gene sequence prediction.

Prediction of cell-type-specific cohesin-mediated chromatin loops based on chromatin state

Chromatin loop is of crucial importance for the regulation of gene transcription. Cohesin is a type of chromatin-associated protein that mediates the interaction of chromatin through the loop extrusion. Cohesin-mediated chromatin interactions have strong cell-type specificity, posing a challenge for predicting chromatin loops. Existing computational methods perform poorly in predicting cell-type-specific chromatin loops. To address this issue, we propose a random forest model to predict cell-type-specific cohesin-mediated chromatin loops based on chromatin states identified by ChromHMM and the occupancy of related factors. Our results show that chromatin state is responsible for cell-type-specificity of loops. Using only chromatin states as features, the model achieved high accuracy in predicting cell-type-specific loops between two cell types and can be applied to different cell types. Furthermore, when chromatin states are combined with the occurrence frequency of CTCF, RAD21, YY1, and H3K27ac ChIP-seq peaks, more accurate prediction can be achieved. Our feature extraction method provides novel insights into predicting cell-type-specific chromatin loops and reveals the relationship between chromatin state and chromatin loop formation.

Prediction of miRNAs and diseases association based on sparse autoencoder and MLP

Introduction: MicroRNAs (miRNAs) are small and non-coding RNA molecules which have multiple important regulatory roles within cells. With the deepening research on miRNAs, more and more researches show that the abnormal expression of miRNAs is closely related to various diseases. The relationship between miRNAs and diseases is crucial for discovering the pathogenesis of diseases and exploring new treatment methods. Methods: Therefore, we propose a new sparse autoencoder and MLP method (SPALP) to predict the association between miRNAs and diseases. In this study, we adopt advanced deep learning technologies, including sparse autoencoder and multi-layer perceptron (MLP), to improve the accuracy of predicting miRNA-disease associations. Firstly, the SPALP model uses a sparse autoencoder to perform feature learning and extract the initial features of miRNAs and diseases separately, obtaining the latent features of miRNAs and diseases. Then, the latent features combine miRNAs functional similarity data with diseases semantic similarity data to construct comprehensive miRNAs-diseases datasets. Subsequently, the MLP model can predict the unknown association among miRNAs and diseases. Result: To verify the performance of our model, we set up several comparative experiments. The experimental results show that, compared with traditional methods and other deep learning prediction methods, our method has significantly improved the accuracy of predicting miRNAs-disease associations, with 94.61% accuracy and 0.9859 AUC value. Finally, we conducted case study of SPALP model. We predicted the top 30 miRNAs that might be related to Lupus Erythematosus, Ecute Myeloid Leukemia, Cardiovascular, Stroke, Diabetes Mellitus five elderly diseases and validated that 27, 29, 29, 30, and 30 of the top 30 are indeed associated. Discussion: The SPALP approach introduced in this study is adept at forecasting the links between miRNAs and diseases, addressing the complexities of analyzing extensive bioinformatics datasets and enriching the comprehension contribution to disease progression of miRNAs.

STM-ac4C: a hybrid model for identification of N4-acetylcytidine (ac4C) in human mRNA based on selective kernel convolution, temporal convolutional network, and multi-head self-attention

N4-acetylcysteine (ac4C) is a chemical modification in mRNAs that alters the structure and function of mRNA by adding an acetyl group to the N4 position of cytosine. Researchers have shown that ac4C is closely associated with the occurrence and development of various cancers. Therefore, accurate prediction of ac4C modification sites on human mRNA is crucial for revealing its role in diseases and developing new diagnostic and therapeutic strategies. However, existing deep learning models still have limitations in prediction accuracy and generalization ability, which restrict their effectiveness in handling complex biological sequence data. This paper introduces a deep learning-based model, STM-ac4C, for predicting ac4C modification sites on human mRNA. The model combines the advantages of selective kernel convolution, temporal convolutional networks, and multi-head self-attention mechanisms to effectively extract and integrate multi-level features of RNA sequences, thereby achieving high-precision prediction of ac4C sites. On the independent test dataset, STM-ac4C showed improvements of 1.81%, 3.5%, and 0.37% in accuracy, Matthews correlation coefficient, and area under the curve, respectively, compared to the existing state-of-the-art technologies. Moreover, its performance on additional balanced and imbalanced datasets also confirmed the model's robustness and generalization ability. Various experimental results indicate that STM-ac4C outperforms existing methods in predictive performance. In summary, STM-ac4C excels in predicting ac4C modification sites on human mRNA, providing a powerful new tool for a deeper understanding of the biological significance of mRNA modifications and cancer treatment. Additionally, the model reveals key sequence features that influence the prediction of ac4C sites through sequence region impact analysis, offering new perspectives for future research. The source code and experimental data are available at https://github.com/ymy12341/STM-ac4C.

msBERT-Promoter: a multi-scale ensemble predictor based on BERT pre-trained model for the two-stage prediction of DNA promoters and their strengths

BACKGROUND

A promoter is a specific sequence in DNA that has transcriptional regulatory functions, playing a role in initiating gene expression. Identifying promoters and their strengths can provide valuable information related to human diseases. In recent years, computational methods have gained prominence as an effective means for identifying promoter, offering a more efficient alternative to labor-intensive biological approaches.

RESULTS

In this study, a two-stage integrated predictor called "msBERT-Promoter" is proposed for identifying promoters and predicting their strengths. The model incorporates multi-scale sequence information through a tokenization strategy and fine-tunes the DNABERT model. Soft voting is then used to fuse the multi-scale information, effectively addressing the issue of insufficient DNA sequence information extraction in traditional models. To the best of our knowledge, this is the first time an integrated approach has been used in the DNABERT model for promoter identification and strength prediction. Our model achieves accuracy rates of 96.2% for promoter identification and 79.8% for promoter strength prediction, significantly outperforming existing methods. Furthermore, through attention mechanism analysis, we demonstrate that our model can effectively combine local and global sequence information, enhancing its interpretability.

CONCLUSIONS

msBERT-Promoter provides an effective tool that successfully captures sequence-related attributes of DNA promoters and can accurately identify promoters and predict their strengths. This work paves a new path for the application of artificial intelligence in traditional biology.

Cotton transposon-related variome reveals roles of transposon-related variations in modern cotton cultivation

INTRODUCTION

Transposon plays a vital role in cotton genome evolution, contributing to the expansion and divergence of genomes within the Gossypium genus. However, knowledge of transposon activity in modern cotton cultivation is limited.

OBJECTIVES

In this study, we aimed to construct transposon-related variome within Gossypium genus and reveal role of transposon-related variations during cotton cultivation. In addition, we try to identify valuable transposon-related variations for cotton breeding.

METHODS

We utilized graphical genome construction to build up the graphical transposon-related variome. Based on the graphical variome, we integrated t-test, eQTL analysis and Mendelian Randomization (MR) to identify valuable transposon activities and elite genes. In addition, a convolutional neural network (CNN) model was constructed to evaluate epigenomic effects of transposon-related variations.

RESULTS

We identified 35,980 transposon activities among 10 cotton genomes, and the diversity of genomic and epigenomic features was observed among 21 transposon categories. The graphical cotton transposon-related variome was constructed, and 9,614 transposon-related variations with plasticity in the modern cotton cohort were used for eQTL, phenotypic t-test and Mendelian Randomization. 128 genes were identified as gene resources improving fiber length and strength simultaneously. 4 genes were selected from 128 genes to construct the elite gene panel whose utility has been validated in a natural cotton cohort and 2 accessions with phenotypic divergence. Based on the eQTL analysis results, we identified transposon-related variations involved in cotton's environmental adaption and human domestication, providing evidence of their role in cotton's adaption-domestication cooperation.

CONCLUSIONS

The cotton transposon-related variome revealed the role of transposon-related variations in modern cotton cultivation, providing genomic resources for cotton molecular breeding.

Improvements in viral gene annotation using large language models and soft alignments

BACKGROUND

The annotation of protein sequences in public databases has long posed a challenge in molecular biology. This issue is particularly acute for viral proteins, which demonstrate limited homology to known proteins when using alignment, k-mer, or profile-based homology search approaches. A novel methodology employing Large Language Models (LLMs) addresses this methodological challenge by annotating protein sequences based on embeddings.

RESULTS

Central to our contribution is the soft alignment algorithm, drawing from traditional protein alignment but leveraging embedding similarity at the amino acid level to bypass the need for conventional scoring matrices. This method not only surpasses pooled embedding-based models in efficiency but also in interpretability, enabling users to easily trace homologous amino acids and delve deeper into the alignments. Far from being a black box, our approach provides transparent, BLAST-like alignment visualizations, combining traditional biological research with AI advancements to elevate protein annotation through embedding-based analysis while ensuring interpretability. Tests using the Virus Orthologous Groups and ViralZone protein databases indicated that the novel soft alignment approach recognized and annotated sequences that both blastp and pooling-based methods, which are commonly used for sequence annotation, failed to detect.

CONCLUSION

The embeddings approach shows the great potential of LLMs for enhancing protein sequence annotation, especially in viral genomics. These findings present a promising avenue for more efficient and accurate protein function inference in molecular biology.

Prediction of blood-brain barrier penetrating peptides based on data augmentation with Augur

BACKGROUND

The blood-brain barrier serves as a critical interface between the bloodstream and brain tissue, mainly composed of pericytes, neurons, endothelial cells, and tightly connected basal membranes. It plays a pivotal role in safeguarding brain from harmful substances, thus protecting the integrity of the nervous system and preserving overall brain homeostasis. However, this remarkable selective transmission also poses a formidable challenge in the realm of central nervous system diseases treatment, hindering the delivery of large-molecule drugs into the brain. In response to this challenge, many researchers have devoted themselves to developing drug delivery systems capable of breaching the blood-brain barrier. Among these, blood-brain barrier penetrating peptides have emerged as promising candidates. These peptides had the advantages of high biosafety, ease of synthesis, and exceptional penetration efficiency, making them an effective drug delivery solution. While previous studies have developed a few prediction models for blood-brain barrier penetrating peptides, their performance has often been hampered by issue of limited positive data.

RESULTS

In this study, we present Augur, a novel prediction model using borderline-SMOTE-based data augmentation and machine learning. we extract highly interpretable physicochemical properties of blood-brain barrier penetrating peptides while solving the issues of small sample size and imbalance of positive and negative samples. Experimental results demonstrate the superior prediction performance of Augur with an AUC value of 0.932 on the training set and 0.931 on the independent test set.

CONCLUSIONS

This newly developed Augur model demonstrates superior performance in predicting blood-brain barrier penetrating peptides, offering valuable insights for drug development targeting neurological disorders. This breakthrough may enhance the efficiency of peptide-based drug discovery and pave the way for innovative treatment strategies for central nervous system diseases.

iDNA-OpenPrompt: OpenPrompt learning model for identifying DNA methylation

Introduction: DNA methylation is a critical epigenetic modification involving the addition of a methyl group to the DNA molecule, playing a key role in regulating gene expression without changing the DNA sequence. The main difficulty in identifying DNA methylation sites lies in the subtle and complex nature of methylation patterns, which may vary across different tissues, developmental stages, and environmental conditions. Traditional methods for methylation site identification, such as bisulfite sequencing, are typically labor-intensive, costly, and require large amounts of DNA, hindering high-throughput analysis. Moreover, these methods may not always provide the resolution needed to detect methylation at specific sites, especially in genomic regions that are rich in repetitive sequences or have low levels of methylation. Furthermore, current deep learning approaches generally lack sufficient accuracy. Methods: This study introduces the iDNA-OpenPrompt model, leveraging the novel OpenPrompt learning framework. The model combines a prompt template, prompt verbalizer, and Pre-trained Language Model (PLM) to construct the prompt-learning framework for DNA methylation sequences. Moreover, a DNA vocabulary library, BERT tokenizer, and specific label words are also introduced into the model to enable accurate identification of DNA methylation sites. Results and Discussion: An extensive analysis is conducted to evaluate the predictive, reliability, and consistency capabilities of the iDNA-OpenPrompt model. The experimental outcomes, covering 17 benchmark datasets that include various species and three DNA methylation modifications (4mC, 5hmC, 6mA), consistently indicate that our model surpasses outstanding performance and robustness approaches.

A BERT-based model for the prediction of lncRNA subcellular localization in Homo sapiens

Understanding the subcellular localization of lncRNAs is crucial for comprehending their regulation activities. The conventional detection of lncRNA subcellular location usually uses in situ detection techniques, which are resource intensive. Some machine learning-based algorithms have been proposed for lncRNA subcellular location prediction in mammals. However, due to the low level of conservation of lncRNA sequence, the performance of cross-species models remains unsatisfactory. In this study, we curated a novel dataset containing subcellular location information of lncRNAs in Homo sapiens. Subsequently, based on the BERT pre-trained language algorithm, we developed a model for lncRNA subcellular location prediction. Our model achieved a micro-average area under the receiver operating characteristic (AUROC) of 0.791 on the training set and an AUROC of 0.700 on the testing nucleus set. Additionally, we conducted cross-species validation and motif discovery to further investigate underlying patterns. In summary, our study provides valuable guidance and computational analysis tools for exploring the mechanisms of lncRNA subcellular localization and the dynamic spatial changes of RNA in abnormal physiological states.

Fuzzy kernel evidence Random Forest for identifying pseudouridine sites

Pseudouridine is an RNA modification that is widely distributed in both prokaryotes and eukaryotes, and plays a critical role in numerous biological activities. Despite its importance, the precise identification of pseudouridine sites through experimental approaches poses significant challenges, requiring substantial time and resources.Therefore, there is a growing need for computational techniques that can reliably and quickly identify pseudouridine sites from vast amounts of RNA sequencing data. In this study, we propose fuzzy kernel evidence Random Forest (FKeERF) to identify pseudouridine sites. This method is called PseU-FKeERF, which demonstrates high accuracy in identifying pseudouridine sites from RNA sequencing data. The PseU-FKeERF model selected four RNA feature coding schemes with relatively good performance for feature combination, and then input them into the newly proposed FKeERF method for category prediction. FKeERF not only uses fuzzy logic to expand the original feature space, but also combines kernel methods that are easy to interpret in general for category prediction. Both cross-validation tests and independent tests on benchmark datasets have shown that PseU-FKeERF has better predictive performance than several state-of-the-art methods. This new method not only improves the accuracy of pseudouridine site identification, but also provides a certain reference for disease control and related drug development in the future.

CODENET: A deep learning model for COVID-19 detection

Conventional COVID-19 testing methods have some flaws: they are expensive and time-consuming. Chest X-ray (CXR) diagnostic approaches can alleviate these flaws to some extent. However, there is no accurate and practical automatic diagnostic framework with good interpretability. The application of artificial intelligence (AI) technology to medical radiography can help to accurately detect the disease, reduce the burden on healthcare organizations, and provide good interpretability. Therefore, this study proposes a new deep neural network (CNN) based on CXR for COVID-19 diagnosis - CodeNet. This method uses contrastive learning to make full use of latent image data to enhance the model's ability to extract features and generalize across different data domains. On the evaluation dataset, the proposed method achieves an accuracy as high as 94.20%, outperforming several other existing methods used for comparison. Ablation studies validate the efficacy of the proposed method, while interpretability analysis shows that the method can effectively guide clinical professionals. This work demonstrates the superior detection performance of a CNN using contrastive learning techniques on CXR images, paving the way for computer vision and artificial intelligence technologies to leverage massive medical data for disease diagnosis.

Interpretable Multi-Scale Deep Learning for RNA Methylation Analysis across Multiple Species

RNA modification plays a crucial role in cellular regulation. However, traditional high-throughput sequencing methods for elucidating their functional mechanisms are time-consuming and labor-intensive, despite extensive research. Moreover, existing methods often limit their focus to specific species, neglecting the simultaneous exploration of RNA modifications across diverse species. Therefore, a versatile computational approach is necessary for interpretable analysis of RNA modifications across species. A multi-scale biological language-based deep learning model is proposed for interpretable, sequential-level prediction of diverse RNA modifications. Benchmark comparisons across species demonstrate the model's superiority in predicting various RNA methylation types over current state-of-the-art methods. The cross-species validation and attention weight visualization also highlight the model's capability to capture sequential and functional semantics from genomic backgrounds. Our analysis of RNA modifications helps us find the potential existence of "biological grammars" in each modification type, which could be effective for mapping methylation-related sequential patterns and understanding the underlying biological mechanisms of RNA modifications.

A new method for handling heterogeneous data in bioinformatics

Heterogeneous data, especially a mixture of numerical and categorical data, widely exist in bioinformatics. Most of works focus on defining new distance metrics rather than learning discriminative metrics for mixed data. Here, we create a new support vector heterogeneous metric learning framework for mixed data. A heterogeneous sample pair kernel is defined for mixed data and metric learning is then converted to a sample pair classification problem. The suggested approach lends itself well to effective resolution through conventional support vector machine solvers. Empirical assessments conducted on mixed data benchmarks and cancer datasets affirm the exceptional efficacy demonstrated by the proposed modeling technique.

PSAC-6mA: 6mA site identifier using self-attention capsule network based on sequence-positioning

DNA N6-methyladenine (6mA) modifications play a pivotal role in the regulation of growth, development, and diseases in organisms. As a significant epigenetic marker, 6mA modifications extensively participate in the intricate regulatory networks of the genome. Hence, gaining a profound understanding of how 6mA is intricately involved in these biological processes is imperative for deciphering the gene regulatory networks within organisms. In this study, we propose PSAC-6mA (Position-self-attention Capsule-6mA), a sequence-location-based self-attention capsule network. The positional layer in the model enables positional relationship extraction and independent parameter setting for each base position, avoiding parameter sharing inherent in convolutional approaches. Simultaneously, the self-attention capsule network enhances dimensionality, capturing correlation information between capsules and achieving exceptional results in feature extraction across multiple spatial dimensions within the model. Experimental results demonstrate the superior performance of PSAC-6mA in recognizing 6mA motifs across various species.

METnet: A novel deep learning model predicting MET dysregulation in non-small-cell lung cancer on computed tomography images

BACKGROUND

Mesenchymal epithelial transformation (MET) is a key molecular target for diagnosis and treatment of non-small cell lung cancer (NSCLC). The corresponding molecularly targeted therapeutics have been approved by Food and Drug Administration (FDA), achieving promising results. However, current detection of MET dysregulation requires biopsy and gene sequencing, which is invasive, time-consuming and difficult to obtain tumor samples.

METHODS

To address the above problems, we developed a noninvasive and convenient deep learning (DL) model based on Computed tomography (CT) imaging data for prediction of MET dysregulation. We introduced the unsupervised algorithm RK-net for automated image processing and utilized the MedSAM large model to achieve automated tissue segmentation. Based on the processed CT images, we developed a DL model (METnet). The model based on the grouped convolutional block. We evaluated the performance of the model over the internal test dataset using the area under the receiver operating characteristic curve (AUROC) and accuracy. We conducted subgroup analysis on the basis of clinical data of the lung cancer patients and compared the performance of the model in different subgroups.

RESULTS

The model demonstrated a good discriminative ability over the internal test dataset. The accuracy of METnet was 0.746 with an AUC value of 0.793 (95% CI 0.714-0.871). The subgroup analysis revealed that the model exhibited similar performance across different subgroups.

CONCLUSIONS

METnet realizes prediction of MET dysregulation in NSCLC, holding promise for guiding precise tumor diagnosis and treatment at the molecular level.

AGF-PPIS: A protein-protein interaction site predictor based on an attention mechanism and graph convolutional networks

Protein-protein interactions play an important role in various biological processes. Interaction among proteins has a wide range of applications. Therefore, the correct identification of protein-protein interactions sites is crucial. In this paper, we propose a novel predictor for protein-protein interactions sites, AGF-PPIS, where we utilize a multi-head self-attention mechanism (introducing a graph structure), graph convolutional network, and feed-forward neural network. We use the Euclidean distance between each protein residue to generate the corresponding protein graph as the input of AGF-PPIS. On the independent test dataset Test_60, AGF-PPIS achieves superior performance over comparative methods in terms of seven different evaluation metrics (ACC, precision, recall, F1-score, MCC, AUROC, AUPRC), which fully demonstrates the validity and superiority of the proposed AGF-PPIS model. The source codes and the steps for usage of AGF-PPIS are available at https://github.com/fxh1001/AGF-PPIS.

A computational model of circRNA-associated diseases based on a graph neural network: prediction and case studies for follow-up experimental validation

BACKGROUND

Circular RNAs (circRNAs) have been confirmed to play a vital role in the occurrence and development of diseases. Exploring the relationship between circRNAs and diseases is of far-reaching significance for studying etiopathogenesis and treating diseases. To this end, based on the graph Markov neural network algorithm (GMNN) constructed in our previous work GMNN2CD, we further considered the multisource biological data that affects the association between circRNA and disease and developed an updated web server CircDA and based on the human hepatocellular carcinoma (HCC) tissue data to verify the prediction results of CircDA.

RESULTS

CircDA is built on a Tumarkov-based deep learning framework. The algorithm regards biomolecules as nodes and the interactions between molecules as edges, reasonably abstracts multiomics data, and models them as a heterogeneous biomolecular association network, which can reflect the complex relationship between different biomolecules. Case studies using literature data from HCC, cervical, and gastric cancers demonstrate that the CircDA predictor can identify missing associations between known circRNAs and diseases, and using the quantitative real-time PCR (RT-qPCR) experiment of HCC in human tissue samples, it was found that five circRNAs were significantly differentially expressed, which proved that CircDA can predict diseases related to new circRNAs.

CONCLUSIONS

This efficient computational prediction and case analysis with sufficient feedback allows us to identify circRNA-associated diseases and disease-associated circRNAs. Our work provides a method to predict circRNA-associated diseases and can provide guidance for the association of diseases with certain circRNAs. For ease of use, an online prediction server ( http://server.malab.cn/CircDA ) is provided, and the code is open-sourced ( https://github.com/nmt315320/CircDA.git ) for the convenience of algorithm improvement.

FEOpti-ACVP: identification of novel anti-coronavirus peptide sequences based on feature engineering and optimization

Anti-coronavirus peptides (ACVPs) represent a relatively novel approach of inhibiting the adsorption and fusion of the virus with human cells. Several peptide-based inhibitors showed promise as potential therapeutic drug candidates. However, identifying such peptides in laboratory experiments is both costly and time consuming. Therefore, there is growing interest in using computational methods to predict ACVPs. Here, we describe a model for the prediction of ACVPs that is based on the combination of feature engineering (FE) optimization and deep representation learning. FEOpti-ACVP was pre-trained using two feature extraction frameworks. At the next step, several machine learning approaches were tested in to construct the final algorithm. The final version of FEOpti-ACVP outperformed existing methods used for ACVPs prediction and it has the potential to become a valuable tool in ACVP drug design. A user-friendly webserver of FEOpti-ACVP can be accessed at http://servers.aibiochem.net/soft/FEOpti-ACVP/.

Identification, characterization and expression analysis of circRNA encoded by SARS-CoV-1 and SARS-CoV-2

Virus-encoded circular RNA (circRNA) participates in the immune response to viral infection, affects the human immune system, and can be used as a target for precision therapy and tumor biomarker. The coronaviruses SARS-CoV-1 and SARS-CoV-2 (SARS-CoV-1/2) that have emerged in recent years are highly contagious and have high mortality rates. In coronaviruses, little is known about the circRNA encoded by the SARS-CoV-1/2. Therefore, this study explores whether SARS-CoV-1/2 encodes circRNA and characteristics and functions of circRNA. Based on RNA-seq data of SARS-CoV-1 and SARS-CoV-2 infections, we used circRNA identification tools (circRNA_finder, find_circ and CIRI2) to identify circRNAs. The number of circRNAs encoded by SARS-CoV-1 and SARS-CoV-2 was identified as 151 and 470, respectively. It can be found that SARS-CoV-2 shows more prominent circRNA encoding ability than SARS-CoV-1. Expression analysis showed that only a few circRNAs encoded by SARS-CoV-1/2 showed high expression levels, and the positive strand produced more abundant circRNAs. Then, based on the identified SARS-CoV-1/2-encoded circRNAs, we performed circRNA identification and characterization using the previously developed CirRNAPL. Finally, target gene prediction and functional enrichment analysis were performed. It was found that viral circRNA is closely related to cancer and has a potential role in regulating host cell functions. This study studied the characteristics and functions of viral circRNA encoded by coronavirus SARS-CoV-1/2, providing a valuable resource for further research on the function and molecular mechanism of coronavirus circRNA.

Large language models in bioinformatics: applications and perspectives

Large language models (LLMs) are a class of artificial intelligence models based on deep learning, which have great performance in various tasks, especially in natural language processing (NLP). Large language models typically consist of artificial neural networks with numerous parameters, trained on large amounts of unlabeled input using self-supervised or semi-supervised learning. However, their potential for solving bioinformatics problems may even exceed their proficiency in modeling human language. In this review, we will present a summary of the prominent large language models used in natural language processing, such as BERT and GPT, and focus on exploring the applications of large language models at different omics levels in bioinformatics, mainly including applications of large language models in genomics, transcriptomics, proteomics, drug discovery and single cell analysis. Finally, this review summarizes the potential and prospects of large language models in solving bioinformatic problems.

DisoFLAG: accurate prediction of protein intrinsic disorder and its functions using graph-based interaction protein language model

Intrinsically disordered proteins and regions (IDPs/IDRs) are functionally important proteins and regions that exist as highly dynamic conformations under natural physiological conditions. IDPs/IDRs exhibit a broad range of molecular functions, and their functions involve binding interactions with partners and remaining native structural flexibility. The rapid increase in the number of proteins in sequence databases and the diversity of disordered functions challenge existing computational methods for predicting protein intrinsic disorder and disordered functions. A disordered region interacts with different partners to perform multiple functions, and these disordered functions exhibit different dependencies and correlations. In this study, we introduce DisoFLAG, a computational method that leverages a graph-based interaction protein language model (GiPLM) for jointly predicting disorder and its multiple potential functions. GiPLM integrates protein semantic information based on pre-trained protein language models into graph-based interaction units to enhance the correlation of the semantic representation of multiple disordered functions. The DisoFLAG predictor takes amino acid sequences as the only inputs and provides predictions of intrinsic disorder and six disordered functions for proteins, including protein-binding, DNA-binding, RNA-binding, ion-binding, lipid-binding, and flexible linker. We evaluated the predictive performance of DisoFLAG following the Critical Assessment of protein Intrinsic Disorder (CAID) experiments, and the results demonstrated that DisoFLAG offers accurate and comprehensive predictions of disordered functions, extending the current coverage of computationally predicted disordered function categories. The standalone package and web server of DisoFLAG have been established to provide accurate prediction tools for intrinsic disorders and their associated functions.

4mCPred-GSIMP: Predicting DNA N4-methylcytosine sites in the mouse genome with multi-Scale adaptive features extraction and fusion

The epigenetic modification of DNA N4-methylcytosine (4mC) is vital for controlling DNA replication and expression. It is crucial to pinpoint 4mC's location to comprehend its role in physiological and pathological processes. However, accurate 4mC detection is difficult to achieve due to technical constraints. In this paper, we propose a deep learning-based approach 4mCPred-GSIMP for predicting 4mC sites in the mouse genome. The approach encodes DNA sequences using four feature encoding methods and combines multi-scale convolution and improved selective kernel convolution to adaptively extract and fuse features from different scales, thereby improving feature representation and optimization effect. In addition, we also use convolutional residual connections, global response normalization and pointwise convolution techniques to optimize the model. On the independent test dataset, 4mCPred-GSIMP shows high sensitivity, specificity, accuracy, Matthews correlation coefficient and area under the curve, which are 0.7812, 0.9312, 0.8562, 0.7207 and 0.9233, respectively. Various experiments demonstrate that 4mCPred-GSIMP outperforms existing prediction tools.

LSA-ac4C: A hybrid neural network incorporating double-layer LSTM and self-attention mechanism for the prediction of N4-acetylcytidine sites in human mRNA

N4-acetylcytidine (ac4C) is a vital constituent of the epitranscriptome and plays a crucial role in the regulation of mRNA expression. Numerous studies have established correlations between ac4C and the incidence, progression and prognosis of various cancers. Therefore, accurately predicting ac4C sites is an important step towards comprehending the biological functions of this modification and devising effective therapeutic interventions. Wet experiments are primary methods for studying ac4C, but computational methods have emerged as a promising supplement due to their cost-effectiveness and shorter research cycles. However, current models still have inherent limitations in terms of predictive performance and generalization ability. Here, we utilized automated machine learning technology to establish a reliable baseline and constructed a deep hybrid neural network, LSA-ac4C, which combines double-layer Long Short-Term Memory (LSTM) and self-attention mechanism for accurate ac4C sites prediction. Benchmarking comparisons demonstrate that LSA-ac4C exhibits superior performance compared to the current state-of-the-art method, with ACC, MCC and AUROC improving by 2.89 %, 5.96 % and 1.53 %, respectively, on an independent test set. Overall, LSA-ac4C serves as a powerful tool for predicting ac4C sites in human mRNA, thus benefiting research on RNA modification. For the convenience of the research community, a web server has been established at http://tubic.org/ac4C.

Time series-based hybrid ensemble learning model with multivariate multidimensional feature coding for DNA methylation prediction

BACKGROUND

DNA methylation is a form of epigenetic modification that impacts gene expression without modifying the DNA sequence, thereby exerting control over gene function and cellular development. The prediction of DNA methylation is vital for understanding and exploring gene regulatory mechanisms. Currently, machine learning algorithms are primarily used for model construction. However, several challenges remain to be addressed, including limited prediction accuracy, constrained generalization capability, and insufficient learning capacity.

RESULTS

In response to the aforementioned challenges, this paper leverages the similarities between DNA sequences and time series to introduce a time series-based hybrid ensemble learning model, called Multi2-Con-CAPSO-LSTM. The model utilizes multivariate and multidimensional encoding approach, combining three types of time series encodings with three kinds of genetic feature encodings, resulting in a total of nine types of feature encoding matrices. Convolutional Neural Networks are utilized to extract features from DNA sequences, including temporal, positional, physicochemical, and genetic information, thereby creating a comprehensive feature matrix. The Long Short-Term Memory model is then optimized using the Chaotic Accelerated Particle Swarm Optimization algorithm for predicting DNA methylation.

CONCLUSIONS

Through cross-validation experiments conducted on 17 species involving three types of DNA methylation (6 mA, 5hmC, and 4mC), the results demonstrate the robust predictive capabilities of the Multi2-Con-CAPSO-LSTM model in DNA methylation prediction across various types and species. Compared with other benchmark models, the Multi2-Con-CAPSO-LSTM model demonstrates significant advantages in sensitivity, specificity, accuracy, and correlation. The model proposed in this paper provides valuable insights and inspiration across various disciplines, including sequence alignment, genetic evolution, time series analysis, and structure-activity relationships.

BERT-5mC: an interpretable model for predicting 5-methylcytosine sites of DNA based on BERT

DNA 5-methylcytosine (5mC) is widely present in multicellular eukaryotes, which plays important roles in various developmental and physiological processes and a wide range of human diseases. Thus, it is essential to accurately detect the 5mC sites. Although current sequencing technologies can map genome-wide 5mC sites, these experimental methods are both costly and time-consuming. To achieve a fast and accurate prediction of 5mC sites, we propose a new computational approach, BERT-5mC. First, we pre-trained a domain-specific BERT (bidirectional encoder representations from transformers) model by using human promoter sequences as language corpus. BERT is a deep two-way language representation model based on Transformer. Second, we fine-tuned the domain-specific BERT model based on the 5mC training dataset to build the model. The cross-validation results show that our model achieves an AUROC of 0.966 which is higher than other state-of-the-art methods such as iPromoter-5mC, 5mC_Pred, and BiLSTM-5mC. Furthermore, our model was evaluated on the independent test set, which shows that our model achieves an AUROC of 0.966 that is also higher than other state-of-the-art methods. Moreover, we analyzed the attention weights generated by BERT to identify a number of nucleotide distributions that are closely associated with 5mC modifications. To facilitate the use of our model, we built a webserver which can be freely accessed at: http://5mc-pred.zhulab.org.cn.

StableDNAm: towards a stable and efficient model for predicting DNA methylation based on adaptive feature correction learning

BACKGROUND

DNA methylation, instrumental in numerous life processes, underscores the paramount importance of its accurate prediction. Recent studies suggest that deep learning, due to its capacity to extract profound insights, provides a more precise DNA methylation prediction. However, issues related to the stability and generalization performance of these models persist.

RESULTS

In this study, we introduce an efficient and stable DNA methylation prediction model. This model incorporates a feature fusion approach, adaptive feature correction technology, and a contrastive learning strategy. The proposed model presents several advantages. First, DNA sequences are encoded at four levels to comprehensively capture intricate information across multi-scale and low-span features. Second, we design a sequence-specific feature correction module that adaptively adjusts the weights of sequence features. This improvement enhances the model's stability and scalability, or its generality. Third, our contrastive learning strategy mitigates the instability issues resulting from sparse data. To validate our model, we conducted multiple sets of experiments on commonly used datasets, demonstrating the model's robustness and stability. Simultaneously, we amalgamate various datasets into a single, unified dataset. The experimental outcomes from this combined dataset substantiate the model's robust adaptability.

CONCLUSIONS

Our research findings affirm that the StableDNAm model is a general, stable, and effective instrument for DNA methylation prediction. It holds substantial promise for providing invaluable assistance in future methylation-related research and analyses.

DP-AOP: A novel SVM-based antioxidant proteins identifier

The identification of antioxidant proteins is a challenging yet meaningful task, as they can protect against the damage caused by some free radicals. In addition to time-consuming, laborious, and expensive experimental identification methods, efficient identification of antioxidant proteins through machine learning algorithms has become increasingly common. In recent years, researchers have proposed models for identifying antioxidant proteins; unfortunately, although the accuracy of models is already high, their sensitivity is too low, indicating the possibility of overfitting in the model. Therefore, we developed a new model called DP-AOP for the recognition of antioxidant proteins. We used the SMOTE algorithm to balance the dataset, selected Wei's proposed feature extraction algorithm to obtain 473 dimensional feature vectors, and based on the sorting function in MRMD, scored and ranked each feature to obtain a feature set with contribution values ranging from high to low. To effectively reduce the feature dimension, we combined the dynamic programming idea to make the local eight features the optimal subset. After obtaining the 36 dimensional feature vectors, we finally selected 17 features through experimental analysis. The SVM classification algorithm was used to implement the model through the libsvm tool. The model achieved satisfactory performance, with an accuracy rate of 91.076 %, SN of 96.4 %, SP of 85.8 %, MCC of 82.6 %, and F1 core of 91.5 %. Furthermore, we built a free web server to facilitate researchers' subsequent unfolding studies of antioxidant protein recognition. The website is http://112.124.26.17:8003/#/.

Comparative evaluation and analysis of DNA N4-methylcytosine methylation sites using deep learning

DNA N4-methylcytosine (4mC) is significantly involved in biological processes, such as DNA expression, repair, and replication. Therefore, accurate prediction methods are urgently needed. Deep learning methods have transformed applications that previously require sequencing expertise into engineering challenges that do not require expertise to solve. Here, we compare a variety of state-of-the-art deep learning models on six benchmark datasets to evaluate their performance in 4mC methylation site detection. We visualize the statistical analysis of the datasets and the performance of different deep-learning models. We conclude that deep learning can greatly expand the potential of methylation site prediction.

On the prediction of non-CG DNA methylation using machine learning

DNA methylation can be detected and measured using sequencing instruments after sodium bisulfite conversion, but experiments can be expensive for large eukaryotic genomes. Sequencing nonuniformity and mapping biases can leave parts of the genome with low or no coverage, thus hampering the ability of obtaining DNA methylation levels for all cytosines. To address these limitations, several computational methods have been proposed that can predict DNA methylation from the DNA sequence around the cytosine or from the methylation level of nearby cytosines. However, most of these methods are entirely focused on CG methylation in humans and other mammals. In this work, we study, for the first time, the problem of predicting cytosine methylation for CG, CHG and CHH contexts on six plant species, either from the DNA primary sequence around the cytosine or from the methylation levels of neighboring cytosines. In this framework, we also study the cross-species prediction problem and the cross-context prediction problem (within the same species). Finally, we show that providing gene and repeat annotations allows existing classifiers to significantly improve their prediction accuracy. We introduce a new classifier called AMPS (annotation-based methylation prediction from sequence) that takes advantage of genomic annotations to achieve higher accuracy.

Prediction of miRNA-disease associations by neural network-based deep matrix factorization

MicroRNA(miRNA) is a class of short non-coding RNAs with a length of about 22 nucleotides, which participates in various biological processes of cells. A number of studies have shown that miRNAs are closely related to the occurrence of cancer and various human diseases. Therefore, studying miRNA-disease associations is helpful to understand the pathogenesis of diseases as well as the prevention, diagnosis, treatment and prognosis of diseases. Traditional biological experimental methods for studying miRNA-disease associations have disadvantages such as expensive equipment, time-consuming and labor-intensive. With the rapid development of bioinformatics, more and more researchers are committed to developing effective computational methods to predict miRNA-disease associations in roder to reduce the time and money cost of experiments. In this study, we proposed a neural network-based deep matrix factorization method named NNDMF to predict miRNA-disease associations. To address the problem that traditional matrix factorization methods can only extract linear features, NNDMF used neural network to perform deep matrix factorization to extract nonlinear features, which makes up for the shortcomings of traditional matrix factorization methods. We compared NNDMF with four previous classical prediction models (IMCMDA, GRMDA, SACMDA and ICFMDA) in global LOOCV and local LOOCV, respectively. The AUCs achieved by NNDMF in two cross-validation methods were 0.9340 and 0.8763, respectively. Furthermore, we conducted case studies on three important human diseases (lymphoma, colorectal cancer and lung cancer) to validate the effectiveness of NNDMF. In conclusion, NNDMF could effectively predict the potential miRNA-disease associations.

M6A-BERT-Stacking: A Tissue-Specific Predictor for Identifying RNA N6-Methyladenosine Sites Based on BERT and Stacking Strategy

As the most abundant RNA methylation modification, N6-methyladenosine (m6A) could regulate asymmetric and symmetric division of hematopoietic stem cells and play an important role in various diseases. Therefore, the precise identification of m6A sites around the genomes of different species is a critical step to further revealing their biological functions and influence on these diseases. However, the traditional wet-lab experimental methods for identifying m6A sites are often laborious and expensive. In this study, we proposed an ensemble deep learning model called m6A-BERT-Stacking, a powerful predictor for the detection of m6A sites in various tissues of three species. First, we utilized two encoding methods, i.e., di ribonucleotide index of RNA (DiNUCindex_RNA) and k-mer word segmentation, to extract RNA sequence features. Second, two encoding matrices together with the original sequences were respectively input into three different deep learning models in parallel to train three sub-models, namely residual networks with convolutional block attention module (Resnet-CBAM), bidirectional long short-term memory with attention (BiLSTM-Attention), and pre-trained bidirectional encoder representations from transformers model for DNA-language (DNABERT). Finally, the outputs of all sub-models were ensembled based on the stacking strategy to obtain the final prediction of m6A sites through the fully connected layer. The experimental results demonstrated that m6A-BERT-Stacking outperformed most of the existing methods based on the same independent datasets.

Pretoria: An effective computational approach for accurate and high-throughput identification of CD8+ t-cell epitopes of eukaryotic pathogens

T-cells recognize antigenic epitopes present on major histocompatibility complex (MHC) molecules, triggering an adaptive immune response in the host. T-cell epitope (TCE) identification is challenging because of the extensive number of undetermined proteins found in eukaryotic pathogens, as well as MHC polymorphisms. In addition, conventional experimental approaches for TCE identification are time-consuming and expensive. Thus, computational approaches that can accurately and rapidly identify CD8⁺ T-cell epitopes (TCEs) of eukaryotic pathogens based solely on sequence information may facilitate the discovery of novel CD8⁺ TCEs in a cost-effective manner. Here, Pretoria (Predictor of CD8⁺ TCEs of eukaryotic pathogens) is proposed as the first stack-based approach for accurate and large-scale identification of CD8⁺ TCEs of eukaryotic pathogens. In particular, Pretoria enabled the extraction and exploration of crucial information embedded in CD8⁺ TCEs by employing a comprehensive set of 12 well-known feature descriptors extracted from multiple groups, including physicochemical properties, composition-transition-distribution, pseudo-amino acid composition, and amino acid composition. These feature descriptors were then utilized to construct a pool of 144 different machine learning (ML)-based classifiers based on 12 popular ML algorithms. Finally, the feature selection method was used to effectively determine the important ML classifiers for the construction of our stacked model. The experimental results indicated that Pretoria is an accurate and effective computational approach for CD8⁺ TCE prediction; it was superior to several conventional ML classifiers and the existing method in terms of the independent test, with an accuracy of 0.866, MCC of 0.732, and AUC of 0.921. Additionally, to maximize user convenience for high-throughput identification of CD8⁺ TCEs of eukaryotic pathogens, a user-friendly web server of Pretoria (http://pmlabstack.pythonanywhere.com/Pretoria) was developed and made freely available.

Constructing discriminative feature space for LncRNA-protein interaction based on deep autoencoder and marginal fisher analysis

Long non-coding RNAs (lncRNAs) play important roles by regulating proteins in many biological processes and life activities. To uncover molecular mechanisms of lncRNA, it is very necessary to identify interactions of lncRNA with proteins. Recently, some machine learning methods were proposed to detect lncRNA-protein interactions according to the distribution of known interactions. The performances of these methods were largely dependent upon: (1) how exactly the distribution of known interactions was characterized by feature space; (2) how discriminative the feature space was for distinguishing lncRNA-protein interactions. Because the known interactions may be multiple and complex model, it remains a challenge to construct discriminative feature space for lncRNA-protein interactions. To resolve this problem, a novel method named DFRPI was developed based on deep autoencoder and marginal fisher analysis in this paper. Firstly, some initial features of lncRNA-protein interactions were extracted from the primary sequences and secondary structures of lncRNA and protein. Secondly, a deep autoencoder was exploited to learn encode parameters of the initial features to describe the known interactions precisely. Next, the marginal fisher analysis was employed to optimize the encode parameters of features to characterize a discriminative feature space of the lncRNA-protein interactions. Finally, a random forest-based predictor was trained on the discriminative feature space to detect lncRNA-protein interactions. Verified by a series of experiments, the results showed that our predictor achieved the precision of 0.920, recall of 0.916, accuracy of 0.918, MCC of 0.836, specificity of 0.920, sensitivity of 0.916 and AUC of 0.906 respectively, which outperforms the concerned methods for predicting lncRNA-protein interaction. It may be suggested that the proposed method can generate a reasonable and effective feature space for distinguishing lncRNA-protein interactions accurately. The code and data are available on https://github.com/D0ub1e-D/DFRPI.

Hybrid gMLP model for interaction prediction of MHC-peptide and TCR

Understanding the interaction of T-cell receptor (TCR) with major histocompatibility-peptide (MHC-peptide) complex is extremely important in human immunotherapy and vaccine development. However, due to the limited available data, the performance of existing models for predicting the interaction of T-cell receptors (TCR) with major histocompatibility-peptide complexes is still unsatisfactory. Deep learning models have been applied to prediction tasks in various fields and have achieved better results compared with other traditional models. In this study, we leverage the gMLP model combined with attention mechanism to predict the interaction of MHC-peptide and TCR. Experiments show that our model can predict TCR-peptide interactions accurately and can handle the problems caused by different TCR lengths. Moreover, we demonstrate that the models trained with paired CDR3β-chain and CDR3α-chain data are better than those trained with only CDR3β-chain or with CDR3α-chain data. We also demonstrate that the hybrid model has greater potential than the traditional convolutional neural network.

iEnhancer-ELM: improve enhancer identification by extracting position-related multiscale contextual information based on enhancer language models

Motivation

Enhancers are important cis-regulatory elements that regulate a wide range of biological functions and enhance the transcription of target genes. Although many feature extraction methods have been proposed to improve the performance of enhancer identification, they cannot learn position-related multiscale contextual information from raw DNA sequences.

Results

In this article, we propose a novel enhancer identification method (iEnhancer-ELM) based on BERT-like enhancer language models. iEnhancer-ELM tokenizes DNA sequences with multi-scale k-mers and extracts contextual information of different scale k-mers related with their positions via an multi-head attention mechanism. We first evaluate the performance of different scale k-mers, then ensemble them to improve the performance of enhancer identification. The experimental results on two popular benchmark datasets show that our model outperforms state-of-the-art methods. We further illustrate the interpretability of iEnhancer-ELM. For a case study, we discover 30 enhancer motifs via a 3-mer-based model, where 12 of motifs are verified by STREME and JASPAR, demonstrating our model has a potential ability to unveil the biological mechanism of enhancer.

Availability and implementation

The models and associated code are available at https://github.com/chen-bioinfo/iEnhancer-ELM.

Supplementary information

Supplementary data are available at Bioinformatics Advances online.

Prediction of apoptosis protein subcellular location based on amphiphilic pseudo amino acid composition

Introduction: Apoptosis proteins play an important role in the process of cell apoptosis, which makes the rate of cell proliferation and death reach a relative balance. The function of apoptosis protein is closely related to its subcellular location, it is of great significance to study the subcellular locations of apoptosis proteins. Many efforts in bioinformatics research have been aimed at predicting their subcellular location. However, the subcellular localization of apoptotic proteins needs to be carefully studied. Methods: In this paper, based on amphiphilic pseudo amino acid composition and support vector machine algorithm, a new method was proposed for the prediction of apoptosis proteins\x{2019} subcellular location. Results and Discussion: The method achieved good performance on three data sets. The Jackknife test accuracy of the three data sets reached 90.5%, 93.9% and 84.0%, respectively. Compared with previous methods, the prediction accuracies of APACC_SVM were improved.

MuLan-Methyl-multiple transformer-based language models for accurate DNA methylation prediction

Transformer-based language models are successfully used to address massive text-related tasks. DNA methylation is an important epigenetic mechanism, and its analysis provides valuable insights into gene regulation and biomarker identification. Several deep learning-based methods have been proposed to identify DNA methylation, and each seeks to strike a balance between computational effort and accuracy. Here, we introduce MuLan-Methyl, a deep learning framework for predicting DNA methylation sites, which is based on 5 popular transformer-based language models. The framework identifies methylation sites for 3 different types of DNA methylation: N6-adenine, N4-cytosine, and 5-hydroxymethylcytosine. Each of the employed language models is adapted to the task using the "pretrain and fine-tune" paradigm. Pretraining is performed on a custom corpus of DNA fragments and taxonomy lineages using self-supervised learning. Fine-tuning aims at predicting the DNA methylation status of each type. The 5 models are used to collectively predict the DNA methylation status. We report excellent performance of MuLan-Methyl on a benchmark dataset. Moreover, we argue that the model captures characteristic differences between different species that are relevant for methylation. This work demonstrates that language models can be successfully adapted to applications in biological sequence analysis and that joint utilization of different language models improves model performance. Mulan-Methyl is open source, and we provide a web server that implements the approach.

ToxMVA: An end-to-end multi-view deep autoencoder method for protein toxicity prediction

Effectively predicting protein toxicity plays an essential step in the early stage of protein-based drug discovery, which is of great help to speed up novel drug screening and reduce costs. Recently, several relevant datasets have been designed, and then machine learning-based methods have been proposed to predict the toxicity of the protein and have shown satisfactory performance. However, previous studies generally directly concatenate different protein features, which may introduce irrelevant information and decrease model performance. In this study, we present a novel end-to-end deep learning-based method called ToxMVA, to predict protein toxicity. To be specific, we first build comprehensive feature profiles of proteins based on primary sequences, including sequential, physicochemical, and contextual semantic information. Next, an autoencoder network is introduced to integrate the multi-view information for obtaining a more concise and accurate feature representation. Extensive experimental results on three datasets demonstrate that ToxMVA has superior performance for protein toxicity prediction and shows better robustness among three different datasets.