Related references
Note: Only part of the references are listed.
Article
Biochemical Research Methods
Madalina Ciortan et al.
Summary: Single-cell RNA sequencing (scRNA-seq) allows researchers to study tissue heterogeneity and discover new cell subtypes. However, the high sparsity of scRNA-seq data poses challenges for clustering analysis. In this study, a new method called graph-sc is proposed, which leverages a graph autoencoder network to create embeddings for scRNA-seq cell data. Experimental results show that graph-sc performs favorably compared to competing techniques and is stable, robust, and computationally efficient.
Article
Biochemical Research Methods
David Buterez et al.
Summary: This study introduces the use of graph neural networks for unsupervised exploration of scRNA-seq data, specifically focusing on dimensionality reduction and clustering. The developed model, CellVGAE, shows effectiveness in extracting meaningful features, visualizing and interpreting the data. It outperforms existing methods in terms of interpretability, capturing information, and reducing training times.
Article
Biochemical Research Methods
Xiao Wang et al.
Summary: Researchers propose a novel deep learning approach, which is a hierarchical distribution-matching framework assisted with contrastive learning, to address batch effects in single-cell RNA sequencing data. This method effectively reduces distribution differences between different batches and aligns samples from different batches to recover cell type clusters.
Article
Biochemical Research Methods
Musu Yuan et al.
Summary: The research introduces a robust deep learning-based single-cell Multiple Reference Annotator that effectively transfers knowledge from multiple insufficient reference datasets to unlabeled target data, while also removing batch effects.
Article
Biochemical Research Methods
Yang Xu et al.
Summary: SMILE is an unsupervised deep learning algorithm that successfully integrates multi-source single-cell omics data and removes batch effects. By maximizing mutual information, it can project cell types from different tissues into a shared space and integrate data from different modalities.
Article
Biochemical Research Methods
A. Ali Heydari et al.
Summary: In this work, a novel framework called ACTIVA is proposed for generating realistic synthetic data using a single-stream adversarial variational autoencoder conditioned with cell-type information. Compared to other GAN-based models, ACTIVA generates cells that are more realistic, harder to identify, and have better gene-gene correlation. Data augmentation with ACTIVA significantly improves classification of rare subtypes and reduces runtime compared to other models.
Review
Biochemical Research Methods
Mingon Kang et al.
Summary: High-throughput next-generation sequencing allows for the generation of large amounts of multi-omics data, which has revolutionized biomedical research by providing a more comprehensive understanding of biological systems and disease development mechanisms. Deep learning algorithms have emerged as a promising method in multi-omics data analysis due to their predictive performance and ability to capture nonlinear and hierarchical features. However, integrating and translating multi-omics data into functional insights remains a challenge, but there is a clear trend towards incorporating multi-omics analysis in biomedical research to explain complex relationships between molecular layers.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Biochemical Research Methods
Qingyang Yin et al.
Summary: Single-cell RNA sequencing (scRNA-seq) is used to study cell heterogeneity and enable the construction of cell atlas. Disease-related scRNA-seq data can be used for disease status prediction. The scIAE framework, based on integrative autoencoders, shows effective and robust classification performance across different datasets, platforms, and species.
BRIEFINGS IN BIOINFORMATICS
(2022)
Review
Biochemical Research Methods
Mario Flores et al.
Summary: Single-cell technologies have become mature enough to provide answers to complex research questions, but the computational challenges of processing massive and complicated datasets have increased. Deep learning is positioned as a competitive alternative for single-cell analyses. This article surveys 25 deep learning algorithms and their applicability in the single-cell RNA-seq processing pipeline, offering readers guidance in choosing suitable algorithms for specific objectives.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Biochemical Research Methods
Siqi Bao et al.
Summary: This study summarizes the recent advances and applications of deep learning-based methods in the analysis of scRNA-seq data and specific tools, and investigates the future perspectives and challenges of deep learning techniques in the appropriate analysis and interpretation of scRNA-seq data.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Biochemical Research Methods
Mengyuan Zhao et al.
Summary: In this study, a deep learning framework called DGRNS is developed for inferring gene regulatory networks from single-cell transcriptomic data. It effectively identifies the relationships between gene pairs and overcomes the challenges of sparsity and noise in the data. The results show that DGRNS outperforms other methods and discovers novel regulatory relationships with high confidence, which require further research.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Biotechnology & Applied Microbiology
Alberto Santos et al.
Summary: A knowledge graph platform integrates proteomics with other omics data and biomedical databases, providing support for precision medicine. The Clinical Knowledge Graph (CKG) is a flexible data model that can accommodate a large number of nodes and relationships, accelerating the analysis and interpretation of proteomics data. By incorporating CKG, clinical decision-making can be informed and proteomics data can be enriched.
NATURE BIOTECHNOLOGY
(2022)
Article
Biotechnology & Applied Microbiology
Emma Dann et al.
Summary: Milo is a scalable statistical framework that performs differential abundance testing by assigning cells to partially overlapping neighborhoods on a k-nearest neighbor graph. It can identify perturbations obscured by discretizing cells into clusters and outperforms alternative testing strategies. Milo is based on cell-cell similarity structure and may be applicable to various single-cell data beyond scRNA-seq.
NATURE BIOTECHNOLOGY
(2022)
Article
Biochemical Research Methods
Malte D. Luecken et al.
Summary: This study benchmarked 68 method and preprocessing combinations on 85 batches of gene expression data, highlighting the importance of highly variable gene selection in improving method performance. When dealing with complex integration tasks, scANVI, Scanorama, scVI, and scGen consistently performed well, while the performance of single-cell ATAC-sequencing integration was strongly influenced by the choice of feature space.
Editorial Material
Cell Biology
Qin Ma et al.
Summary: Deep learning has great potential in single-cell data analysis, but there are still many challenges and possible new developments to be explored. In this commentary, the progress, limitations, best practices, and outlook of adapting deep learning methods for analyzing single-cell data are considered.
NATURE REVIEWS MOLECULAR CELL BIOLOGY
(2022)
Article
Biochemical Research Methods
Xiangtao Li et al.
Summary: The article introduces a method called SEDIM, which automatically designs deep neural network architectures for imputing gene expression levels. It improves computational efficiency by constructing an offline surrogate model. Experimental results show that SEDIM significantly improves imputation and clustering performance, and also performs well in other contexts and platforms.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Biochemical Research Methods
Mohit Goyal et al.
Summary: The article introduces a neural-network-based framework JIND for automated cell-type identification. JIND performs integration in a suitably chosen space and employs a novel asymmetric alignment to address batch effects. Results demonstrate that the joint approach of integration and classification in JIND outperforms existing pipelines in accuracy and can identify cells that cannot be reliably classified.
Article
Biotechnology & Applied Microbiology
Mohammad Lotfollahi et al.
Summary: scArches is a deep learning strategy for mapping query datasets on top of a reference, allowing efficient and decentralized reference construction while preserving biological state information and removing batch effects. It generalizes to multimodal reference mapping and can impute missing modalities.
NATURE BIOTECHNOLOGY
(2022)
Article
Biotechnology & Applied Microbiology
T. Lohoff et al.
Summary: Improved integration of spatial and single-cell transcriptomic data through the seqFISH method provides insights into mouse development, revealing cell types across the embryo and uncovering axes of cell differentiation that are not apparent from scRNA-seq data. This approach offers a high-resolution spatial map for studying cell fate decisions in complex tissues and development.
NATURE BIOTECHNOLOGY
(2022)
Article
Biotechnology & Applied Microbiology
Zhi-Jie Cao et al.
Summary: GLUE is a computational framework that bridges the gap between different omics layers by modeling regulatory interactions, and it outperforms state-of-the-art tools in accuracy, robustness, and scalability for heterogeneous single-cell multi-omics data. It has been successfully applied in various challenging tasks.
NATURE BIOTECHNOLOGY
(2022)
Article
Biotechnology & Applied Microbiology
Romain Lopez et al.
Summary: Most spatial transcriptomics technologies have resolution limitations. However, the DestVI method can provide higher-resolution analysis of cellular organization and gene expression changes in tissues.
NATURE BIOTECHNOLOGY
(2022)
Article
Biochemical Research Methods
Ron Zeira et al.
Summary: PASTE is a method for aligning and integrating spatial transcriptomics data from adjacent tissue slices, which utilizes transcriptomic similarity and spatial coordinates to increase downstream analysis power. PASTE accurately aligns spots, constructs 3D alignments or integrates into a consensus slice, and improves the identification of cell types and differentially expressed genes.
Article
Multidisciplinary Sciences
Hui Li et al.
Summary: Single cell RNA sequencing is widely used in biomedical research, generating large volumes of diverse data. To clean the data from noise and artifacts, researchers have developed the AutoClass model, which integrates an autoencoder and a classifier to effectively remove various types of noise and artifacts.
NATURE COMMUNICATIONS
(2022)
Article
Multidisciplinary Sciences
Andrea Riba et al.
Summary: This study utilizes single-cell RNA-sequencing technology to analyze the dynamics of gene regulation during the cell cycle. The authors propose a deep learning method, DeepCycle, to infer the cell cycle state in single cells based on scRNA-seq data. By generating scRNA-seq libraries in different cell systems and observing cycling patterns in cell cycle-related genes, the authors establish a high-resolution map of the entire cell cycle transcriptome and identify major waves of transcription during the G1 phase.
NATURE COMMUNICATIONS
(2022)
Article
Multidisciplinary Sciences
Jamie L. Marshall et al.
Summary: High-resolution spatial transcriptomics technique enables mapping of RNA expression and has the potential for exploring disease processes and therapeutic pathways. In this study, Slide-seqV2 was applied to mouse and human kidneys, revealing the presence of disease-specific cell neighborhoods.
Review
Genetics & Heredity
Stefan Stanojevic et al.
Summary: Recently developed technologies for single-cell genomic data generation have had a revolutionary impact on the field of biology. The use of multi-omics assays provides even greater opportunities for understanding cellular states and biological processes. However, integrating omics data with different dimensionalities and statistical properties remains a challenging task, and computational tools leveraging various techniques from machine translation to network theory are being developed to address this issue.
GENOMICS PROTEOMICS & BIOINFORMATICS
(2022)
Article
Biochemical Research Methods
Arnab Kumar Mondal et al.
Summary: Clustering single-cell RNA sequencing data is challenging due to high-dimensionality and data-sparsity. Regularized Auto-Encoder (RAE) based deep neural network models have achieved success, but suffer from bias-variance trade-off. The proposed method, scRAE, effectively addresses this issue and demonstrates efficacy through experimentation.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2022)
Article
Multidisciplinary Sciences
Lei Xiong et al.
Summary: SCALEX is a deep-learning method for online integration of diverse single-cell data. It accurately aligns different modalities of single-cell data, retains true biological differences, and has superior performance in large-scale single-cell applications.
NATURE COMMUNICATIONS
(2022)
Article
Biochemical Research Methods
Zhengyang Dong et al.
Summary: The study introduces a semi-supervised dimensionality reduction method, netAE, which achieves satisfactory classification accuracy in cell labeling even with a small labeled set, without disrupting the similarity structure of the original space.
Review
Chemistry, Analytical
Maria A. Woerheide et al.
Summary: Recent advances in high-throughput technologies have enabled the profiling of multiple layers of a biological system, generating vast amounts of data that can be integrated in multi-omics studies. Integrative analysis of such datasets is complicated by the high dimensionality, heterogeneity, and lack of universal protocols. While previous reviews focused on integrating two omics layers with a phenotype, this review highlights methods to combine metabolomics data with two or more omics layers without focusing on a specific phenotype.
ANALYTICA CHIMICA ACTA
(2021)
Article
Biochemical Research Methods
Liang Chen et al.
Summary: Inspired by unsupervised domain adaptation, the study introduces a flexible single-cell semi-supervised clustering and annotation framework, scSemiCluster. By integrating reference and target data for training, the model utilizes structure similarity regularization and pairwise constraints to optimize clustering results. Without explicit domain alignment and batch effect correction, scSemiCluster outperforms other state-of-the-art algorithms, making it the first to utilize both deep discriminative clustering and deep generative clustering in the single-cell field.
Article
Biochemical Research Methods
Tian Tian et al.
Summary: Research shows that there is still limited exploration in imputing discrete data, the proposed ZINB model-based autoencoder in this paper effectively improves imputation accuracy and can be applied in pre-processing scRNA-seq data. Analysis of extensive experimental data reveals that zero-inflated reconstruction can effectively enhance cell type separation and improve the accuracy of differential expression analysis.
Article
Biochemical Research Methods
Mojtaba Bahrami et al.
Summary: The single-cell Generative Adversarial Network (scGAN) can acquire cell-type-specific gene expression patterns from raw data while minimizing the confounding effect of technical artifacts. In experiments, scGAN demonstrated superior performance in forming cell type clusters and identifying genes associated with major depressive disorder compared to other methods.
Article
Biochemical Research Methods
Bin Yu et al.
Summary: The scGMAI is a new single-cell Gaussian mixture clustering method that utilizes autoencoder networks and FastICA to accurately cluster and identify cell types from scRNA-Seq data, showing great potential in data analysis.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Chunman Zuo et al.
Summary: The study introduces a single-cell multimodal variational autoencoder model that effectively integrates transcriptomic and chromatin accessibility information, accurately representing the multilayer profiles of cells and demonstrating good capabilities in dissecting cellular heterogeneity, denoising and imputing data, as well as constructing associations between multilayer omics data.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Chenyang Xu et al.
Summary: In this study, a single-cell dropout imputation method (GNNImpute) was proposed, which effectively utilized shared information to impute the dropout in scRNA-seq data. The method was evaluated with different real datasets and showed promising results in terms of MSE, MAE, PCC and CS. The use of graph attention convolution and autoencoder structure demonstrated great potential in single-cell dropout imputation.
BMC BIOINFORMATICS
(2021)
Article
Biochemistry & Molecular Biology
Chenling Xu et al.
Summary: As the number of single-cell transcriptomics datasets increases, integrating data to achieve a common ontology of cell types and states becomes the natural next step. scVI and scANVI are two methods that can effectively integrate data and perform well in cell state annotation, with high accuracy, scalability, and adaptability.
MOLECULAR SYSTEMS BIOLOGY
(2021)
Article
Biotechnology & Applied Microbiology
Robert R. Stickels et al.
Summary: Slide-seqV2 technology achieves an RNA capture efficiency close to 50% of single-cell RNA-seq data, which is 10 times higher than Slide-seq and approaches the detection efficiency of droplet-based single-cell RNA-seq techniques. Leveraging the high detection efficiency of Slide-seqV2, dendritically localized mRNAs in mouse hippocampal neurons were successfully identified, and spatial information was integrated with single-cell trajectory analysis tools to characterize the spatiotemporal development of the mouse neocortex, revealing underlying genetic programs that were poorly sampled with Slide-seq. The combination of near-cellular resolution and high transcript detection efficiency makes Slide-seqV2 applicable across various experimental contexts.
NATURE BIOTECHNOLOGY
(2021)
Article
Mathematical & Computational Biology
Luca Alessandri et al.
Summary: Single-cell RNA sequencing is a crucial tool for studying cellular heterogeneity. Researchers have developed a tool based on Sparsely-Connected Autoencoder for functional mining of single cell clusters, enabling the discovery of hidden features associated with scRNAseq data. New metrics were implemented to quantify the abilities of the tool and evaluate the quality of neural network achievements.
NPJ SYSTEMS BIOLOGY AND APPLICATIONS
(2021)
Article
Biochemical Research Methods
Jinjin Tian et al.
Summary: A new single-cell RNA sequencing data simulator ESCO is proposed in the study, focusing on gene co-expression and evaluating the performance of imputation methods in GCN recovery. The results show that imputation generally helps GCN recovery when data are not too sparse, while simple data aggregating methods are more suitable in the presence of an excessive fraction of zero counts.
Article
Biochemical Research Methods
Xuan Liu et al.
Summary: Single-cell RNA sequencing (scRNA-Seq) is a promising strategy for characterizing immune cell populations, but manually classifying immune cells from transcriptional profiles remains a challenge. ImmClassifier, a machine learning classifier, has shown to be effective in distinguishing fine-grained cell types in scRNA-Seq experiments, outperforming other tools in precision and recall.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Biochemistry & Molecular Biology
Sijin Cheng et al.
Summary: This study conducted a pan-cancer analysis of single myeloid cells from 210 patients across 15 human cancer types, revealing distinct features of TIMs in different cancer types. For example, mast cells in nasopharyngeal cancer were associated with better prognosis, and pro-angiogenic TAMs showed diverse markers across different cancer types. The composition of TIMs appeared to be linked with certain features of somatic mutations and gene expressions.
Article
Biology
Junyi Li et al.
Summary: The newly proposed scRNA-seq clustering framework ScGSLC effectively integrates single-cell RNA sequencing data and protein-protein interaction networks, clustering cells through graph convolution networks and outperforming state-of-the-art methods on nine public datasets in terms of unsupervised clustering.
COMPUTATIONAL BIOLOGY AND CHEMISTRY
(2021)
Article
Biochemistry & Molecular Biology
Jacob C. Kimmel et al.
Summary: scNym is a semi-supervised adversarial neural network that can transfer cell identity annotations between different experiments by learning rich representations of cell identities from both labeled and unlabeled datasets. It shows superior performance in transferring annotations across experiments and can synthesize information from multiple datasets to improve accuracy. Additionally, scNym models are well calibrated, interpretable, and can be enhanced with saliency methods.
Article
Multidisciplinary Sciences
Chen Qiao et al.
Summary: VeloAE is a tailored representation learning method that can robustly estimate cellular transitions in single-cell transcriptome experiments and capture cellular differentiation in various biological systems. It accurately identifies dynamic stimulations in time-series designs.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2021)
Article
Biochemical Research Methods
Adam Gayoso et al.
Summary: totalVI is a framework for end-to-end joint analysis of CITE-seq data which probabilistically represents data as a composite of biological and technical factors, providing a cohesive solution for common analysis tasks. It demonstrates strong performance in tasks such as dimensionality reduction, dataset integration, correlation estimation, and differential expression testing.
Article
Multidisciplinary Sciences
Duc Tran et al.
Summary: Accurate analysis of single-cell RNA sequencing (scRNA-seq) data is crucial for various research fields, but is often hindered by technical noise and high dropout rates. The hierarchical autoencoder, scDHA, introduced in this study, outperforms existing methods in various aspects of scRNA-seq analysis, including cell segregation and classification.
NATURE COMMUNICATIONS
(2021)
Article
Multidisciplinary Sciences
Zixiang Luo et al.
Summary: Dimensionality reduction is crucial for visualizing and interpreting high-dimensional single-cell RNA sequencing data, and the single-cell graph autoencoder (scGAE) introduced in this study preserves both topological structure and feature information simultaneously. By extending scGAE for data visualization, clustering, and trajectory inference, it outperforms recent deep learning methods in accurately reconstructing developmental trajectories and separating cell clusters. Implementation on empirical data provides new insights into cell developmental lineages and maintains inter-cluster distances.
SCIENTIFIC REPORTS
(2021)
Article
Biochemical Research Methods
Raktim Mitra et al.
Summary: RVAgene is a recurrent variational autoencoder for modeling gene expression dynamics, accurately reconstructing temporal gene profiles, and discovering biological features through a low-dimensional representation of data. It can generate new gene expression data by sampling the latent space and learn a low-error latent space representation through cross-validation. It shows potential for unsupervised discovery through clustering and gene ontology term enrichment analysis.
Article
Biochemical Research Methods
Jinglong Zhang et al.
Summary: MAT(2) aligns cells in the manifold space using a deep neural network with a contrastive learning strategy, defining cell triplets based on known cell type annotations to produce a more robust consensus manifold and reconstructing batch-effect-free gene expression, which helps to annotate cell types more effectively.
Article
Biochemical Research Methods
Xiangtao Li et al.
Summary: This study proposes an evolutionary multiobjective DEC method to simultaneously evolve the hyperparameters and architectures of DEC for better clustering analysis of single-cell RNA sequencing data. Experimental results demonstrate that the algorithm outperforms other state-of-the-art clustering methods under different metrics.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Cell Biology
Winston Koh et al.
Summary: This study explores the use of deep metric learning methods to achieve transfer of cell type annotation across different scRNA-seq platforms, batches, and species, using only a small training set, achieving high accuracy.
FRONTIERS IN CELL AND DEVELOPMENTAL BIOLOGY
(2021)
Article
Multidisciplinary Sciences
Juexin Wang et al.
Summary: Single-cell RNA-Seq faces challenges such as sparsity in sequencing and complex patterns in gene expression. The introduction of a graph neural network based on a hypothesis-free deep learning framework provides an effective representation of gene expression and cell-cell relationships.
NATURE COMMUNICATIONS
(2021)
Article
Multidisciplinary Sciences
Jiarui Ding et al.
Summary: scPhere is a scalable deep generative model that accurately embeds cells into low-dimensional hyperspherical or hyperbolic spaces to represent scRNA-seq data. By addressing multi-level technical and biological variability, resolving cell crowding, and uncovering temporal trajectories, scPhere facilitates the interpretation and analysis of biological data.
NATURE COMMUNICATIONS
(2021)
Article
Genetics & Heredity
Wenbo Yu et al.
Summary: The power of single-cell RNA sequencing in detecting cell heterogeneity or developmental process is increasing, and combining two batches of scRNA-seq data requires solving technical differences, which can be further constrained by matching cells and cell types. In this study, an auto-encoder was utilized to achieve this goal, and the performance was evaluated against other alignment methods by preserving cluster separation and identifying biologically meaningful differential gene expressions.
FRONTIERS IN GENETICS
(2021)
Article
Multidisciplinary Sciences
Jiahua Rao et al.
Summary: Single-cell RNA sequencing technology enables analysis of single-cell transcriptomes with unprecedented throughput and resolution, but faces the challenge of dropout problem. The developed method GraphSCI, based on graph convolution networks, outperforms other state-of-the-art methods in imputation, accurately inferring gene-to-gene relationships and providing powerful assistance during training.
Article
Biochemical Research Methods
Chunman Zuo et al.
Summary: DCCA is a computational tool for joint analysis of single-cell multi-omics data, capable of dissecting cellular heterogeneity, denoising and aggregating data, and constructing links between multi-omics data. By fine-tuning networks and inferring new transcriptional regulatory relations, DCCA demonstrates superior capability in analyzing and understanding complex biological processes.
Review
Biotechnology & Applied Microbiology
Parminder S. Reel et al.
Summary: With the advancement of high-throughput omics technologies, it is crucial for biomedical research to adopt integrative approaches to analyze diverse omics data using machine learning algorithms. This can lead to the discovery of novel biomarkers and improve disease prediction and precision medicine delivery.-Methods in machine learning have enabled researchers to gain a deeper insight into biological systems and provide recommendations for interdisciplinary professionals looking to incorporate machine learning skills in multi-omics studies.
BIOTECHNOLOGY ADVANCES
(2021)
Article
Biochemical Research Methods
Madalina Ciortan et al.
Summary: This study introduces contrastive-sc, an unsupervised learning method for scRNA-seq data that outperforms state-of-the-art techniques in clustering performance. Through extensive experimental analysis on simulated and real-world datasets, contrastive-sc demonstrates efficiency and robustness in clustering analysis.
BMC BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Tianyu Wang et al.
Summary: The study proposes a multimodal end-to-end deep learning model named sigGCN for cell classification, which combines a graph convolutional network (GCN) and a neural network to exploit gene interaction networks. Results show that the proposed method outperforms other commonly used methods in terms of classification accuracy and F1 scores, demonstrating that integrating prior knowledge about gene interactions with gene expressions using GCN methodologies can extract effective features improving the performance of cell classification.
BMC BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Andrea Tangherloni et al.
Summary: Single-cell RNA sequencing experiments are increasingly used to study normal development and pathologies, but the low-dimensional representation and dataset integration remain challenging. The newly introduced tool, scAEspy, utilizes advanced autoencoders and loss functions to enhance downstream analysis of scRNA-Seq data, surpassing existing solutions.
BMC BIOINFORMATICS
(2021)
Article
Biochemistry & Molecular Biology
Justin Lakkis et al.
Summary: Recent advances in single-cell RNA-seq technologies have led to significant biological discoveries, but batch effects remain a major challenge in studies involving human tissues. Existing methods for batch effect correction focus on low-dimensional spaces, leaving gene expression spaces vulnerable to batch effects. CarDEC, a joint deep learning model, has demonstrated superior performance in denoising and batch effect correction for scRNA-seq data, making it a valuable tool for large-scale studies.
Review
Biochemical Research Methods
Zoe A. Clarke et al.
Summary: This tutorial provides guidelines for interpreting single-cell transcriptomic maps to identify cell types, states and other biologically relevant patterns, with a recommended three-step workflow including automatic cell annotation, manual cell annotation, and verification. It also discusses frequently encountered challenges, strategies to address them, as well as guiding principles and specific recommendations for software tools and resources.
Article
Biochemistry & Molecular Biology
Xin Shao et al.
Summary: scDeepSort is a pre-trained tool for cell-type annotation in single-cell transcriptomics using deep learning and a weighted graph neural network. It demonstrates high performance and robustness across multiple datasets, achieving an accuracy of 83.79%.
NUCLEIC ACIDS RESEARCH
(2021)
Article
Biochemistry & Molecular Biology
Bin Duan et al.
Summary: Efficient single-cell assignment is crucial for analyzing single-cell sequencing data, and integrating multiple references can further improve single-cell assignment. The mtSC framework proposed in this study integrates multiple references based on multitask deep metric learning, demonstrating state-of-the-art effectiveness for integrative single-cell assignment with multiple references through evaluation on publicly available benchmark datasets.
NUCLEIC ACIDS RESEARCH
(2021)
Article
Multidisciplinary Sciences
Shobana Stassen et al.
Summary: Scalable trajectory inference for multi-omic single cell datasets is challenging due to the complexity of topologies. The VIA method presented in this study overcomes these limitations by using lazy-teleporting random walks to accurately reconstruct complex cellular trajectories.
NATURE COMMUNICATIONS
(2021)
Article
Multidisciplinary Sciences
Qianqian Song et al.
Summary: Single-cell omics is a rapidly growing area in genomics, but leveraging disparate datasets for analysis is challenging. The scGCN, a graph convolutional network, allows for effective knowledge transfer across different omics datasets.
NATURE COMMUNICATIONS
(2021)
Article
Multidisciplinary Sciences
Yifan Zhao et al.
Summary: The new model scETM addresses challenges in single-cell RNA-seq analysis by optimizing gene signatures and cell functions through meaningful embeddings, utilizing both neural network and linear decoder for scalability and interpretability. It demonstrates remarkable cross-tissue and cross-species zero-shot transfer-learning performance, enriching biologically relevant pathways and disease-related topics.
NATURE COMMUNICATIONS
(2021)
Article
Biochemical Research Methods
Andreas Kopf et al.
Summary: The study introduces a novel generative clustering model MoE-Sim-VAE, which efficiently learns multi-modal distributions of high-dimensional data and shows superior performance in clustering tasks.
PLOS COMPUTATIONAL BIOLOGY
(2021)
Article
Genetics & Heredity
Deeksha Doddahonnaiah et al.
Summary: By utilizing natural language processing methods, this study established gene-cell type associations (GCAs) to assist in more accurate classification and annotation of single cell RNA sequencing (scRNA-seq) datasets. Through the development of an enhanced annotation algorithm (scALE) that leverages GCAs for cell cluster categorization, its effectiveness in predicting cellular identity was demonstrated.
Article
Genetics & Heredity
Jianping Zhao et al.
Summary: This paper proposes a scRNA-seq data dimensionality reduction algorithm SCDRHA based on a hierarchical autoencoder, which consists of two core modules for denoising and projecting data into a low-dimensional space. Experimental results show that SCDRHA outperforms existing algorithms in dimension reduction and noise reduction on five real scRNA-seq datasets, and significantly improves data visualization and cell clustering performance.
FRONTIERS IN GENETICS
(2021)
Article
Genetics & Heredity
Bin Zou et al.
Summary: DeepMNN is a novel deep learning-based method for batch effect correction in scRNA-seq data, which has shown better or comparable performance compared to other state-of-the-art batch correction methods in various scenarios.
FRONTIERS IN GENETICS
(2021)
Article
Biotechnology & Applied Microbiology
Boying Gong et al.
Summary: Cobolt is a novel method designed for analyzing data from joint-modality platforms and integrating multiple datasets across different modalities. It demonstrates its integration capabilities by jointly analyzing multi-modality data of gene expression and chromatin accessibility with single-cell RNA-seq and ATAC-seq datasets.
Article
Biochemical Research Methods
Xiang Zhou et al.
Summary: The study developed a virtual adversarial domain adaptation network called scAdapt to transfer cell labels between datasets with batch effects. scAdapt utilized both the labeled source and unlabeled target data to train an enhanced classifier, aligning centroids to create a joint embedding. The scAdapt outperformed existing methods for classification in various datasets and showed superior capabilities in preserving cluster structure.
BRIEFINGS IN BIOINFORMATICS
(2021)
Review
Biochemical Research Methods
Lucrezia Patruno et al.
Summary: The advancements in single-cell sequencing methods have allowed for unprecedented characterization of cellular states, but technical issues such as data noise remain a challenge. Various data science methods have been proposed to recover lost or corrupted information from single-cell sequencing data. A comprehensive analysis comparing 19 denoising and imputation methods revealed their performance in different experimental scenarios.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Biotechnology & Applied Microbiology
David S. Fischer et al.
Summary: sfaira is a single-cell data zoo paired with executable pre-trained models for public datasets, addressing issues with independent analysis and contextualization in single-cell RNA-seq datasets. We propose an adaptation of cross-entropy loss for cell type classification tailored to datasets annotated at different levels of coarseness.
Article
Biotechnology & Applied Microbiology
Rohit Singh et al.
Summary: Schema is a tool that synthesizes multiple biological information modalities using a metric learning strategy, which can be used for inferring cell types, data visualization, performing differential gene expression analysis, and estimating evolutionary pressure on peptide sequences.
Article
Biotechnology & Applied Microbiology
Hengshi Yu et al.
Summary: This study systematically assesses the strengths and weaknesses of deep generative models such as VAEs and GANs on single-cell gene expression data, and introduces MichiGAN, a neural network that combines their strengths to sample from disentangled representations. MichiGAN allows for sampling from disentangled representations and manipulating distinct aspects of cellular identity, as well as predicting single-cell gene expression responses to drug treatment.
Article
Biotechnology & Applied Microbiology
Dongfang Wang et al.
Summary: The integration of single-cell RNA-sequencing datasets from multiple sources is crucial for deciphering cell-to-cell heterogeneities and interactions. The novel unsupervised batch effect removal framework, iMAP, shows superior, robust, and scalable performance in reliably detecting batch-specific cells and effectively mixing distributions of batch-shared cell types. Applying iMAP to tumor microenvironment datasets from different platforms can leverage the advantages of both platforms to discover novel cell-cell interactions.
Article
Biochemical Research Methods
Songwei Ge et al.
Summary: Dimensionality reduction is crucial in analyzing single-cell RNA-sequencing (scRNA-seq) data, with downstream methods relying on reduced representations for tasks like clustering and alignment. Existing methods are sensitive to batch effects, which can lead to biased representations. A new domain adversarial neural network approach has been developed to address this issue, allowing for more accurate cell-type assignment and mitigating the impact of batch effects.
JOURNAL OF COMPUTATIONAL BIOLOGY
(2021)
Article
Biochemical Research Methods
Eugene Lin et al.
BMC BIOINFORMATICS
(2020)
Letter
Biotechnology & Applied Microbiology
Valentine Svensson
NATURE BIOTECHNOLOGY
(2020)
Article
Biochemical Research Methods
Robert A. Amezquita et al.
Article
Biochemical Research Methods
Valentine Svensson et al.
Article
Biochemical Research Methods
Christopher Heje Gronbech et al.
Article
Biochemistry & Molecular Biology
Charles Blatti et al.
Article
Multidisciplinary Sciences
Xiangjie Li et al.
NATURE COMMUNICATIONS
(2020)
Article
Genetics & Heredity
Weilai Chi et al.
Article
Genetics & Heredity
Liang Chen et al.
FRONTIERS IN GENETICS
(2020)
Article
Biochemistry & Molecular Biology
Shixiong Zhang et al.
NUCLEIC ACIDS RESEARCH
(2020)
Review
Psychiatry
Chang Su et al.
TRANSLATIONAL PSYCHIATRY
(2020)
Article
Genetics & Heredity
Liang Chen et al.
Article
Biochemistry & Molecular Biology
Nicholas J. Bernstein et al.
Article
Biochemistry & Molecular Biology
Yungang Xu et al.
NUCLEIC ACIDS RESEARCH
(2020)
Article
Biochemistry & Molecular Biology
Suhas Srinivasan et al.
Review
Biochemistry & Molecular Biology
Jeongwoo Lee et al.
EXPERIMENTAL AND MOLECULAR MEDICINE
(2020)
Article
Multidisciplinary Sciences
Zhi-Jie Cao et al.
NATURE COMMUNICATIONS
(2020)
Article
Multidisciplinary Sciences
Ioana Bica et al.
SCIENTIFIC REPORTS
(2020)
Article
Biochemical Research Methods
Maria Brbic et al.
Article
Biochemical Research Methods
Ayse B. Dincer et al.
Article
Biochemical Research Methods
Mohammad Lotfollahi et al.
Article
Genetics & Heredity
Kaikun Xie et al.
NAR GENOMICS AND BIOINFORMATICS
(2020)
Article
Genetics & Heredity
Liang Chen et al.
NAR GENOMICS AND BIOINFORMATICS
(2020)
Review
Biochemistry & Molecular Biology
David N. Nicholson et al.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2020)
Article
Computer Science, Artificial Intelligence
Soeren Lukassen et al.
NATURE MACHINE INTELLIGENCE
(2020)
Article
Computer Science, Artificial Intelligence
Jian Hu et al.
NATURE MACHINE INTELLIGENCE
(2020)
Article
Biotechnology & Applied Microbiology
Nikolaus Fortelny et al.
Article
Mathematical & Computational Biology
Md Bahadur Badsha et al.
QUANTITATIVE BIOLOGY
(2020)
Article
Biotechnology & Applied Microbiology
Manu Setty et al.
NATURE BIOTECHNOLOGY
(2019)
Article
Biochemical Research Methods
Yue Deng et al.
Review
Endocrinology & Metabolism
Klaus H. Kaestner et al.
Article
Biotechnology & Applied Microbiology
Wouter Saelens et al.
NATURE BIOTECHNOLOGY
(2019)
Article
Biochemistry & Molecular Biology
Samuel L. Wolock et al.
Article
Biochemistry & Molecular Biology
Christopher S. McGinnis et al.
Article
Biochemical Research Methods
Jiajie Peng et al.
BMC BIOINFORMATICS
(2019)
Article
Biochemistry & Molecular Biology
Tim Stuart et al.
Article
Biochemical Research Methods
Jingshu Wang et al.
Article
Biochemical Research Methods
Matthew Amodio et al.
Article
Biochemical Research Methods
Thomas A. Geddes et al.
BMC BIOINFORMATICS
(2019)
Article
Multidisciplinary Sciences
Goekcen Eraslan et al.
NATURE COMMUNICATIONS
(2019)
Article
Biotechnology & Applied Microbiology
Cedric Arisdakessian et al.
Article
Biotechnology & Applied Microbiology
Christoph Hafemeister et al.
Article
Multidisciplinary Sciences
V. A. Traag et al.
SCIENTIFIC REPORTS
(2019)
Article
Biotechnology & Applied Microbiology
Tongxin Wang et al.
Article
Biochemistry & Molecular Biology
David van Dijk et al.
Article
Biochemical Research Methods
Mo Huang et al.
Article
Multidisciplinary Sciences
Jiarui Ding et al.
NATURE COMMUNICATIONS
(2018)
Article
Biochemistry & Molecular Biology
Hyunghoon Cho et al.
Article
Biochemical Research Methods
Romain Lopez et al.
Article
Genetics & Heredity
Li Zhang et al.
FRONTIERS IN GENETICS
(2018)
Article
Genetics & Heredity
Dongfang Wang et al.
GENOMICS PROTEOMICS & BIOINFORMATICS
(2018)
Article
Oncology
Kumardeep Chaudharyl et al.
CLINICAL CANCER RESEARCH
(2018)
Article
Biotechnology & Applied Microbiology
F. Alexander Wolf et al.
Article
Biochemical Research Methods
Marlon Stoeckius et al.
Article
Biochemical Research Methods
Todd M. Gierahn et al.
Article
Multidisciplinary Sciences
Grace X. Y. Zheng et al.
NATURE COMMUNICATIONS
(2017)
Article
Biology
Daniel Scott Himmelstein et al.
Review
Biotechnology & Applied Microbiology
Yehudit Hasin et al.
Review
Multidisciplinary Sciences
Yann LeCun et al.
Review
Genetics & Heredity
Maxwell W. Libbrecht et al.
NATURE REVIEWS GENETICS
(2015)
Article
Multidisciplinary Sciences
Heonjong Han et al.
SCIENTIFIC REPORTS
(2015)
Article
Biotechnology & Applied Microbiology
Cole Trapnell et al.
NATURE BIOTECHNOLOGY
(2014)
Article
Computer Science, Artificial Intelligence
Cheng-Yuan Liou et al.
Article
Biochemical Research Methods
Fuchou Tang et al.
Article
Multidisciplinary Sciences
Amit Tzur et al.
Article
Mechanics
Vincent D. Blondel et al.
JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT
(2008)
Article
Biotechnology & Applied Microbiology
Chuong B. Do et al.
NATURE BIOTECHNOLOGY
(2008)
Article
Biochemistry & Molecular Biology
MA Harris et al.
NUCLEIC ACIDS RESEARCH
(2004)
