BMC Bioinformatics最新文献

{"title":"LDAGM: prediction lncRNA-disease asociations by graph convolutional auto-encoder and multilayer perceptron based on multi-view heterogeneous networks.","authors":"Bing Zhang, Haoyu Wang, Chao Ma, Hai Huang, Zhou Fang, Jiaxing Qu","doi":"10.1186/s12859-024-05950-z","DOIUrl":"https://doi.org/10.1186/s12859-024-05950-z","url":null,"abstract":"<p><strong>Background: </strong>Long non-coding RNAs (lncRNAs) can prevent, diagnose, and treat a variety of complex human diseases, and it is crucial to establish a method to efficiently predict lncRNA-disease associations.</p><p><strong>Results: </strong>In this paper, we propose a prediction method for the lncRNA-disease association relationship, named LDAGM, which is based on the Graph Convolutional Autoencoder and Multilayer Perceptron model. The method first extracts the functional similarity and Gaussian interaction profile kernel similarity of lncRNAs and miRNAs, as well as the semantic similarity and Gaussian interaction profile kernel similarity of diseases. It then constructs six homogeneous networks and deeply fuses them using a deep topology feature extraction method. The fused networks facilitate feature complementation and deep mining of the original association relationships, capturing the deep connections between nodes. Next, by combining the obtained deep topological features with the similarity network of lncRNA, disease, and miRNA interactions, we construct a multi-view heterogeneous network model. The Graph Convolutional Autoencoder is employed for nonlinear feature extraction. Finally, the extracted nonlinear features are combined with the deep topological features of the multi-view heterogeneous network to obtain the final feature representation of the lncRNA-disease pair. Prediction of the lncRNA-disease association relationship is performed using the Multilayer Perceptron model. To enhance the performance and stability of the Multilayer Perceptron model, we introduce a hidden layer called the aggregation layer in the Multilayer Perceptron model. Through a gate mechanism, it controls the flow of information between each hidden layer in the Multilayer Perceptron model, aiming to achieve optimal feature extraction from each hidden layer.</p><p><strong>Conclusions: </strong>Parameter analysis, ablation studies, and comparison experiments verified the effectiveness of this method, and case studies verified the accuracy of this method in predicting lncRNA-disease association relationships.</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 1","pages":"332"},"PeriodicalIF":2.9,"publicationDate":"2024-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11481433/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142457213","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"DNASimCLR: a contrastive learning-based deep learning approach for gene sequence data classification.","authors":"Minghao Yang, Zehua Wang, Zizhuo Yan, Wenxiang Wang, Qian Zhu, Changlong Jin","doi":"10.1186/s12859-024-05955-8","DOIUrl":"https://doi.org/10.1186/s12859-024-05955-8","url":null,"abstract":"<p><strong>Background: </strong>The rapid advancements in deep neural network models have significantly enhanced the ability to extract features from microbial sequence data, which is critical for addressing biological challenges. However, the scarcity and complexity of labeled microbial data pose substantial difficulties for supervised learning approaches. To address these issues, we propose DNASimCLR, an unsupervised framework designed for efficient gene sequence data feature extraction.</p><p><strong>Results: </strong>DNASimCLR leverages convolutional neural networks and the SimCLR framework, based on contrastive learning, to extract intricate features from diverse microbial gene sequences. Pre-training was conducted on two classic large scale unlabelled datasets encompassing metagenomes and viral gene sequences. Subsequent classification tasks were performed by fine-tuning the pretrained model using the previously acquired model. Our experiments demonstrate that DNASimCLR is at least comparable to state-of-the-art techniques for gene sequence classification. For convolutional neural network-based approaches, DNASimCLR surpasses the latest existing methods, clearly establishing its superiority over the state-of-the-art CNN-based feature extraction techniques. Furthermore, the model exhibits superior performance across diverse tasks in analyzing biological sequence data, showcasing its robust adaptability.</p><p><strong>Conclusions: </strong>DNASimCLR represents a robust and database-agnostic solution for gene sequence classification. Its versatility allows it to perform well in scenarios involving novel or previously unseen gene sequences, making it a valuable tool for diverse applications in genomics.</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 1","pages":"328"},"PeriodicalIF":2.9,"publicationDate":"2024-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11476100/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142457212","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multi-task graph deep learning model to predict drugs combination of synergy and sensitivity scores.","authors":"Samar Monem, Aboul Ella Hassanien, Alaa H Abdel-Hamid","doi":"10.1186/s12859-024-05925-0","DOIUrl":"10.1186/s12859-024-05925-0","url":null,"abstract":"<p><strong>Background: </strong>Drug combination treatments have proven to be a realistic technique for treating challenging diseases such as cancer by enhancing efficacy and mitigating side effects. To achieve the therapeutic goals of these combinations, it is essential to employ multi-targeted drug combinations, which maximize effectiveness and synergistic effects.</p><p><strong>Results: </strong>This paper proposes 'MultiComb', a multi-task deep learning (MTDL) model designed to simultaneously predict the synergy and sensitivity of drug combinations. The model utilizes a graph convolution network to represent the Simplified Molecular-Input Line-Entry (SMILES) of two drugs, generating their respective features. Also, three fully connected subnetworks extract features of the cancer cell line. These drug and cell line features are then concatenated and processed through an attention mechanism, which outputs two optimized feature representations for the target tasks. The cross-stitch model learns the relationship between these tasks. At last, each learned task feature is fed into fully connected subnetworks to predict the synergy and sensitivity scores. The proposed model is validated using the O'Neil benchmark dataset, which includes 38 unique drugs combined to form 17,901 drug combination pairs and tested across 37 unique cancer cells. The model's performance is tested using some metrics like mean square error ( <math><mrow><mi>MSE</mi></mrow> </math> ), mean absolute error ( <math><mrow><mi>MAE</mi></mrow> </math> ), coefficient of determination ( <math> <msup><mrow><mi>R</mi></mrow> <mn>2</mn></msup> </math> ), Spearman, and Pearson scores. The mean synergy scores of the proposed model are 232.37, 9.59, 0.57, 0.76, and 0.73 for the previous metrics, respectively. Also, the values for mean sensitivity scores are 15.59, 2.74, 0.90, 0.95, and 0.95, respectively.</p><p><strong>Conclusion: </strong>This paper proposes an MTDL model to predict synergy and sensitivity scores for drug combinations targeting specific cancer cell lines. The MTDL model demonstrates superior performance compared to existing approaches, providing better results.</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 1","pages":"327"},"PeriodicalIF":2.9,"publicationDate":"2024-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11468365/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142399244","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MethylSeqLogo: DNA methylation smart sequence logos.","authors":"Fei-Man Hsu, Paul Horton","doi":"10.1186/s12859-024-05896-2","DOIUrl":"10.1186/s12859-024-05896-2","url":null,"abstract":"<p><strong>Background: </strong>Some transcription factors, MYC for example, bind sites of potentially methylated DNA. This may increase binding specificity as such sites are (1) highly under-represented in the genome, and (2) offer additional, tissue specific information in the form of hypo- or hyper-methylation. Fortunately, bisulfite sequencing data can be used to investigate this phenomenon.</p><p><strong>Method: </strong>We developed MethylSeqLogo, an extension of sequence logos which includes new elements to indicate DNA methylation and under-represented dimers in each position of a set binding sites. Our method displays information from both DNA strands, and takes into account the sequence context (CpG or other) and genome region (promoter versus whole genome) appropriate to properly assess the expected background dimer frequency and level of methylation. MethylSeqLogo preserves sequence logo semantics-the relative height of nucleotides within a column represents their proportion in the binding sites, while the absolute height of each column represents information (relative entropy) and the height of all columns added together represents total information RESULTS: We present figures illustrating the utility of using MethylSeqLogo to summarize data from several CpG binding transcription factors. The logos show that unmethylated CpG binding sites are a feature of transcription factors such as MYC and ZBTB33, while some other CpG binding transcription factors, such as CEBPB, appear methylation neutral.</p><p><strong>Conclusions: </strong>Our software enables users to explore bisulfite and ChIP sequencing data sets-and in the process obtain publication quality figures.</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 Suppl 2","pages":"326"},"PeriodicalIF":2.9,"publicationDate":"2024-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11462690/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142387652","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"NeuroimaGene: an R package for assessing the neurological correlates of genetically regulated gene expression.","authors":"Xavier Bledsoe, Eric R Gamazon","doi":"10.1186/s12859-024-05936-x","DOIUrl":"10.1186/s12859-024-05936-x","url":null,"abstract":"<p><strong>Background: </strong>We present the NeuroimaGene resource as an R package designed to assist researchers in identifying genes and neurologic features relevant to psychiatric and neurological health. While recent studies have identified hundreds of genes as potential components of pathophysiology in neurologic and psychiatric disease, interpreting the physiological consequences of this variation is challenging. The integration of neuroimaging data with molecular findings is a step toward addressing this challenge. In addition to sharing associations with both molecular variation and clinical phenotypes, neuroimaging features are intrinsically informative of cognitive processes. NeuroimaGene provides a tool to understand how disease-associated genes relate to the intermediate structure of the brain.</p><p><strong>Results: </strong>We created NeuroimaGene, a user-friendly, open access R package now available for public use. Its primary function is to identify neuroimaging derived brain features that are impacted by genetically regulated expression of user-provided genes or gene sets. This resource can be used to (1) characterize individual genes or gene sets as relevant to the structure and function of the brain, (2) identify the region(s) of the brain or body in which expression of target gene(s) is neurologically relevant, (3) impute the brain features most impacted by user-defined gene sets such as those produced by cohort level gene association studies, and (4) generate publication level, modifiable visual plots of significant findings. We demonstrate the utility of the resource by identifying neurologic correlates of stroke-associated genes derived from pre-existing analyses.</p><p><strong>Conclusions: </strong>Integrating neurologic data as an intermediate phenotype in the pathway from genes to brain-based diagnostic phenotypes increases the interpretability of molecular studies and enriches our understanding of disease pathophysiology. The NeuroimaGene R package is designed to assist in this process and is publicly available for use.</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 1","pages":"325"},"PeriodicalIF":2.9,"publicationDate":"2024-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11463069/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142387651","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Crossfeat: a transformer-based cross-feature learning model for predicting drug side effect frequency.","authors":"Bin Baek, Hyunju Lee","doi":"10.1186/s12859-024-05915-2","DOIUrl":"10.1186/s12859-024-05915-2","url":null,"abstract":"<p><strong>Background: </strong>Safe drug treatment requires an understanding of the potential side effects. Identifying the frequency of drug side effects can reduce the risks associated with drug use. However, existing computational methods for predicting drug side effect frequencies heavily depend on known drug side effect frequency information. Consequently, these methods face challenges when predicting the side effect frequencies of new drugs. Although a few methods can predict the side effect frequencies of new drugs, they exhibit unreliable performance owing to the exclusion of drug-side effect relationships.</p><p><strong>Results: </strong>This study proposed CrossFeat, a model based on convolutional neural network-transformer architecture with cross-feature learning that can predict the occurrence and frequency of drug side effects for new drugs, even in the absence of information regarding drug-side effect relationships. CrossFeat facilitates the concurrent learning of drugs and side effect information within its transformer architecture. This simultaneous exchange of information enables drugs to learn about their associated side effects, while side effects concurrently acquire information about the respective drugs. Such bidirectional learning allows for the comprehensive integration of drug and side effect knowledge. Our five-fold cross-validation experiments demonstrated that CrossFeat outperforms existing studies in predicting side effect frequencies for new drugs without prior knowledge.</p><p><strong>Conclusions: </strong>Our model offers a promising approach for predicting the drug side effect frequencies, particularly for new drugs where prior information is limited. CrossFeat's superior performance in cross-validation experiments, along with evidence from case studies and ablation experiments, highlights its effectiveness.</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 1","pages":"324"},"PeriodicalIF":2.9,"publicationDate":"2024-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11459996/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142387650","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"C-ziptf: stable tensor factorization for zero-inflated multi-dimensional genomics data.","authors":"Daniel Chafamo, Vignesh Shanmugam, Neriman Tokcan","doi":"10.1186/s12859-024-05886-4","DOIUrl":"10.1186/s12859-024-05886-4","url":null,"abstract":"<p><p>In the past two decades, genomics has advanced significantly, with single-cell RNA-sequencing (scRNA-seq) marking a pivotal milestone. ScRNA-seq provides unparalleled insights into cellular diversity and has spurred diverse studies across multiple conditions and samples, resulting in an influx of complex multidimensional genomics data. This highlights the need for robust methodologies capable of handling the complexity and multidimensionality of such genomics data. Furthermore, single-cell data grapples with sparsity due to issues like low capture efficiency and dropout effects. Tensor factorizations (TF) have emerged as powerful tools to unravel the complex patterns from multi-dimensional genomics data. Classic TF methods, based on maximum likelihood estimation, struggle with zero-inflated count data, while the inherent stochasticity in TFs further complicates result interpretation and reproducibility. Our paper introduces Zero Inflated Poisson Tensor Factorization (ZIPTF), a novel method for high-dimensional zero-inflated count data factorization. We also present Consensus-ZIPTF (C-ZIPTF), merging ZIPTF with a consensus-based approach to address stochasticity. We evaluate our proposed methods on synthetic zero-inflated count data, simulated scRNA-seq data, and real multi-sample multi-condition scRNA-seq datasets. ZIPTF consistently outperforms baseline matrix and tensor factorization methods, displaying enhanced reconstruction accuracy for zero-inflated data. When dealing with high probabilities of excess zeros, ZIPTF achieves up to <math><mrow><mn>2.4</mn> <mo>×</mo></mrow> </math> better accuracy. Moreover, C-ZIPTF notably enhances the factorization's consistency. When tested on synthetic and real scRNA-seq data, ZIPTF and C-ZIPTF consistently uncover known and biologically meaningful gene expression programs. Access our data and code at: https://github.com/klarman-cell-observatory/scBTF and https://github.com/klarman-cell-observatory/scbtf_experiments .</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 1","pages":"323"},"PeriodicalIF":2.9,"publicationDate":"2024-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11456250/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142378997","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Tabular deep learning: a comparative study applied to multi-task genome-wide prediction.","authors":"Yuhua Fan, Patrik Waldmann","doi":"10.1186/s12859-024-05940-1","DOIUrl":"10.1186/s12859-024-05940-1","url":null,"abstract":"<p><strong>Purpose: </strong>More accurate prediction of phenotype traits can increase the success of genomic selection in both plant and animal breeding studies and provide more reliable disease risk prediction in humans. Traditional approaches typically use regression models based on linear assumptions between the genetic markers and the traits of interest. Non-linear models have been considered as an alternative tool for modeling genomic interactions (i.e. non-additive effects) and other subtle non-linear patterns between markers and phenotype. Deep learning has become a state-of-the-art non-linear prediction method for sound, image and language data. However, genomic data is better represented in a tabular format. The existing literature on deep learning for tabular data proposes a wide range of novel architectures and reports successful results on various datasets. Tabular deep learning applications in genome-wide prediction (GWP) are still rare. In this work, we perform an overview of the main families of recent deep learning architectures for tabular data and apply them to multi-trait regression and multi-class classification for GWP on real gene datasets.</p><p><strong>Methods: </strong>The study involves an extensive overview of recent deep learning architectures for tabular data learning: NODE, TabNet, TabR, TabTransformer, FT-Transformer, AutoInt, GANDALF, SAINT and LassoNet. These architectures are applied to multi-trait GWP. Comprehensive benchmarks of various tabular deep learning methods are conducted to identify best practices and determine their effectiveness compared to traditional methods.</p><p><strong>Results: </strong>Extensive experimental results on several genomic datasets (three for multi-trait regression and two for multi-class classification) highlight LassoNet as a standout performer, surpassing both other tabular deep learning models and the highly efficient tree based LightGBM method in terms of both best prediction accuracy and computing efficiency.</p><p><strong>Conclusion: </strong>Through series of evaluations on real-world genomic datasets, the study identifies LassoNet as a standout performer, surpassing decision tree methods like LightGBM and other tabular deep learning architectures in terms of both predictive accuracy and computing efficiency. Moreover, the inherent variable selection property of LassoNet provides a systematic way to find important genetic markers that contribute to phenotype expression.</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 1","pages":"322"},"PeriodicalIF":2.9,"publicationDate":"2024-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11452967/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142375044","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ribosomal computing: implementation of the computational method.","authors":"Pratima Chatterjee, Prasun Ghosal, Sahadeb Shit, Arindam Biswas, Saurav Mallik, Sarah Allabun, Manal Othman, Almubarak Hassan Ali, E Elshiekh, Ben Othman Soufiene","doi":"10.1186/s12859-024-05945-w","DOIUrl":"10.1186/s12859-024-05945-w","url":null,"abstract":"<p><strong>Background: </strong>Several computational and mathematical models of protein synthesis have been explored to accomplish the quantitative analysis of protein synthesis components and polysome structure. The effect of gene sequence (coding and non-coding region) in protein synthesis, mutation in gene sequence, and functional model of ribosome needs to be explored to investigate the relationship among protein synthesis components further. Ribosomal computing is implemented by imitating the functional property of protein synthesis.</p><p><strong>Result: </strong>In the proposed work, a general framework of ribosomal computing is demonstrated by developing a computational model to present the relationship between biological details of protein synthesis and computing principles. Here, mathematical abstractions are chosen carefully without probing into intricate chemical details of the micro-operations of protein synthesis for ease of understanding. This model demonstrates the cause and effect of ribosome stalling during protein synthesis and the relationship between functional protein and gene sequence. Moreover, it also reveals the computing nature of ribosome molecules and other protein synthesis components. The effect of gene mutation on protein synthesis is also explored in this model.</p><p><strong>Conclusion: </strong>The computational model for ribosomal computing is implemented in this work. The proposed model demonstrates the relationship among gene sequences and protein synthesis components. This model also helps to implement a simulation environment (a simulator) for generating protein chains from gene sequences and can spot the problem during protein synthesis. Thus, this simulator can identify a disease that can happen due to a protein synthesis problem and suggest precautions for it.</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 1","pages":"321"},"PeriodicalIF":2.9,"publicationDate":"2024-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11448306/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142364277","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SAE-Impute: imputation for single-cell data via subspace regression and auto-encoders.","authors":"Liang Bai, Boya Ji, Shulin Wang","doi":"10.1186/s12859-024-05944-x","DOIUrl":"10.1186/s12859-024-05944-x","url":null,"abstract":"<p><strong>Background: </strong>Single-cell RNA sequencing (scRNA-seq) technology has emerged as a crucial tool for studying cellular heterogeneity. However, dropouts are inherent to the sequencing process, known as dropout events, posing challenges in downstream analysis and interpretation. Imputing dropout data becomes a critical concern in scRNA-seq data analysis. Present imputation methods predominantly rely on statistical or machine learning approaches, often overlooking inter-sample correlations.</p><p><strong>Results: </strong>To address this limitation, We introduced SAE-Impute, a new computational method for imputing single-cell data by combining subspace regression and auto-encoders for enhancing the accuracy and reliability of the imputation process. Specifically, SAE-Impute assesses sample correlations via subspace regression, predicts potential dropout values, and then leverages these predictions within an autoencoder framework for interpolation. To validate the performance of SAE-Impute, we systematically conducted experiments on both simulated and real scRNA-seq datasets. These results highlight that SAE-Impute effectively reduces false negative signals in single-cell data and enhances the retrieval of dropout values, gene-gene and cell-cell correlations. Finally, We also conducted several downstream analyses on the imputed single-cell RNA sequencing (scRNA-seq) data, including the identification of differential gene expression, cell clustering and visualization, and cell trajectory construction.</p><p><strong>Conclusions: </strong>These results once again demonstrate that SAE-Impute is able to effectively reduce the droupouts in single-cell dataset, thereby improving the functional interpretability of the data.</p>","PeriodicalId":8958,"journal":{"name":"BMC Bioinformatics","volume":"25 1","pages":"317"},"PeriodicalIF":2.9,"publicationDate":"2024-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11443887/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142360991","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"生物学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}