Binomial models uncover biological variation during feature selection of droplet-based single-cell RNA sequencing

Single-cell RNA sequencing (scRNA-seq) aims to characterize how variation in gene expression is distributed across cells in tissues and organisms. Yet, effective comprehension of these extremely high-dimensional datasets remains a critical barrier to progress in biological research. In standard analyses of scRNA-seq data, feature selection steps aim to reduce the dimensionality of the data by focusing on a subset of genes that are the most biologically variable across a set of cells. Ideally, these features provide the genes that are the most informative for partitioning groups of transcriptionally distinct cells, each representing a different cell type or identity. In this work, we propose a simple feature selection model where a binomial sampling process for each mRNA species produces a null model of technical variation. To compare our model to existing methods, we use scRNA-seq data where cell identities have been established a priori for each cell, and characterize whether different feature sets retain biologically varying genes, distort neighborhood structures, and allow popular clustering algorithms to partition groups of cells into their established classes. We find that our model of biological variation, which we term "Differentially Distributed Genes" or DDGs, outperforms existing methods, and enables dimensionality reduction without loss of critical structure within the data set.

Download Full-text

Information Theoretic Feature Selection Methods for Single Cell RNA-Sequencing

10.1101/646919 ◽

2019 ◽

Author(s):

Umang Varma ◽

Justin Colacino ◽

Anna Gilbert

Keyword(s):

Feature Selection ◽

Single Cell ◽

Rna Sequencing ◽

Complex Mixture ◽

Cell Types ◽

Marker Genes ◽

Selection Methods ◽

Information Theoretic ◽

Single Cell Rna Sequencing ◽

Information Theoretic Methods

AbstractSingle cell RNA-sequencing (scRNA-seq) technologies have generated an expansive amount of new biological information, revealing new cellular populations and hierarchical relationships. A number of technologies complementary to scRNA-seq rely on the selection of a smaller number of marker genes (or features) to accurately differentiate cell types within a complex mixture of cells. In this paper, we benchmark differential expression methods against information-theoretic feature selection methods to evaluate the ability of these algorithms to identify small and efficient sets of genes that are informative about cell types. Unlike differential methods, that are strictly binary and univariate, information-theoretic methods can be used as any combination of binary or multiclass and univariate or multivariate. We show for some datasets, information theoretic methods can reveal genes that are both distinct from those selected by traditional algorithms and that are as informative, if not more, of the class labels. We also present detailed and principled theoretical analyses of these algorithms. All information theoretic methods in this paper are implemented in our PicturedRocks Python package that is compatible with the widely used scanpy package.

Download Full-text

A Comprehensive Multi-Center Cross-platform Benchmarking Study of Single-cell RNA Sequencing Using Reference Samples

10.1101/2020.03.27.010249 ◽

2020 ◽

Author(s):

Wanqiu Chen ◽

Yongmei Zhao ◽

Xin Chen ◽

Xiaojiang Xu ◽

Zhaowei Yang ◽

...

Keyword(s):

Single Cell ◽

Rna Sequencing ◽

Cell Types ◽

Batch Effect ◽

Batch Effects ◽

Data Set ◽

Batch Correction ◽

Single Cell Rna Sequencing ◽

Cross Platform ◽

Reference Samples

AbstractSingle-cell RNA sequencing (scRNA-seq) has become a very powerful technology for biomedical research and is becoming much more affordable as methods continue to evolve, but it is unknown how reproducible different platforms are using different bioinformatics pipelines, particularly the recently developed scRNA-seq batch correction algorithms. We carried out a comprehensive multi-center cross-platform comparison on different scRNA-seq platforms using standard reference samples. We compared six pre-processing pipelines, seven bioinformatics normalization procedures, and seven batch effect correction methods including CCA, MNN, Scanorama, BBKNN, Harmony, limma and ComBat to evaluate the performance and reproducibility of 20 scRNA-seq data sets derived from four different platforms and centers. We benchmarked scRNA-seq performance across different platforms and testing sites using global gene expression profiles as well as some cell-type specific marker genes. We showed that there were large batch effects; and the reproducibility of scRNA-seq across platforms was dictated both by the expression level of genes selected and the batch correction methods used. We found that CCA, MNN, and BBKNN all corrected the batch variations fairly well for the scRNA-seq data derived from biologically similar samples across platforms/sites. However, for the scRNA-seq data derived from or consisting of biologically distinct samples, limma and ComBat failed to correct batch effects, whereas CCA over-corrected the batch effect and misclassified the cell types and samples. In contrast, MNN, Harmony and BBKNN separated biologically different samples/cell types into correspondingly distinct dimensional subspaces; however, consistent with this algorithm’s logic, MNN required that the samples evaluated each contain a shared portion of highly similar cells. In summary, we found a great cross-platform consistency in separating two distinct samples when an appropriate batch correction method was used. We hope this large cross-platform/site scRNA-seq data set will provide a valuable resource, and that our findings will offer useful advice for the single-cell sequencing community.

Download Full-text

SMaSH: A scalable, general marker gene identification framework for single-cell RNA sequencing and Spatial Transcriptomics

10.1101/2021.04.08.438978 ◽

2021 ◽

Author(s):

Michael E Nelson ◽

Simone G Riva ◽

Ann Cvejic

Keyword(s):

Single Cell ◽

Rna Sequencing ◽

Marker Gene ◽

Marker Genes ◽

Sequencing Data ◽

Computational Framework ◽

Data Set ◽

Spatially Resolved ◽

Single Cell Rna Sequencing ◽

The Given

Spatial transcriptomics is revolutionising the study of single-cell RNA and tissue-wide cell heterogeneity, but few robust methods connecting spatially resolved cells to so-called marker genes from single-cell RNA sequencing, which generate significant insight gleaned from spatial methods, exist. Here we present SMaSH, a general computational framework for extracting key marker genes from single-cell RNA sequencing data for spatial transcriptomics approaches. SMaSH extracts robust and biologically well-motivated marker genes, which characterise the given data-set better than existing and limited computational approaches for global marker gene calculation.

Download Full-text

Differential dropout analysis captures biological variation in single-cell RNA sequencing data

10.1101/2021.02.01.429187 ◽

2021 ◽

Author(s):

Gerard A. Bouland ◽

Ahmed Mahfouz ◽

Marcel J.T. Reinders

Keyword(s):

Single Cell ◽

Differential Expression ◽

Rna Sequencing ◽

Relative Abundance ◽

Differential Expression Analysis ◽

Biological Variation ◽

Sequencing Data ◽

Single Cell Rna Sequencing ◽

Technical Artifacts ◽

Zero Counts

AbstractSingle-cell RNA sequencing data is characterized by a large number of zero counts, yet there is growing evidence that these zeros reflect biological rather than technical artifacts. We propose differential dropout analysis (DDA), as an alternative to differential expression analysis (DEA), to identify the effects of biological variation in single-cell RNA sequencing data. Using 16 publicly available datasets, we show that dropout patterns are biological in nature and can assess the relative abundance of transcripts more robustly than counts.

Download Full-text

Factorization-based Imputation of Expression in Single-cell Transciptomic Analysis (FIESTA) recovers Gene-Cell-State relationships

10.1101/2021.04.29.441691 ◽

2021 ◽

Author(s):

Elnaz Mirzaei Mehrabad ◽

Aditya Bhaskara ◽

Benjamin T. Spike

Keyword(s):

Gene Expression ◽

Single Cell ◽

Rna Sequencing ◽

Missing Values ◽

Expression Patterns ◽

Cellular Systems ◽

False Negatives ◽

Data Set ◽

Link Type ◽

Single Cell Rna Sequencing

AbstractMotivationSingle cell RNA sequencing (scRNA-seq) is a powerful gene expression profiling technique that is presently revolutionizing the study of complex cellular systems in the biological sciences. Existing single-cell RNA-sequencing methods suffer from sub-optimal target recovery leading to inaccurate measurements including many false negatives. The resulting ‘zero-inflated’ data may confound data interpretation and visualization.ResultsSince cells have coherent phenotypes defined by conserved molecular circuitries (i.e. multiple gene products working together) and since similar cells utilize similar circuits, information about each each expression value or ‘node’ in a multi-cell, multi-gene scRNA-Seq data set is expected to also be predictable from other nodes in the data set. Based on this logic, several approaches have been proposed to impute missing values by extracting information from non-zero measurements in a data set. In this study, we applied non-negative matrix factorization approaches to a selection of published scRNASeq data sets to recommend new values where original measurements are likely to be inaccurate and where ‘zero’ measurements are predicted to be false negatives. The resulting imputed data model predicts novel cell type markers and expression patterns more closely matching gene expression values from orthogonal measurements and/or predicted literature than the values obtained from other previously published imputation [email protected] and implementationFIESTA is written in R and is available at https://github.com/elnazmirzaei/FIESTA and https://github.com/TheSpikeLab/FIESTA.

Download Full-text

Differential analysis of binarized single-cell RNA sequencing data captures biological variation

NAR Genomics and Bioinformatics ◽

10.1093/nargab/lqab118 ◽

2021 ◽

Vol 3 (4) ◽

Author(s):

Gerard A Bouland ◽

Ahmed Mahfouz ◽

Marcel J T Reinders

Keyword(s):

Single Cell ◽

Rna Sequencing ◽

Expression Profiles ◽

Biological Variation ◽

Expression Data ◽

Differential Analysis ◽

Sequencing Data ◽

Single Cell Rna Sequencing ◽

Cell Expression ◽

Zero Counts

Abstract Single-cell RNA sequencing data is characterized by a large number of zero counts, yet there is growing evidence that these zeros reflect biological variation rather than technical artifacts. We propose to use binarized expression profiles to identify the effects of biological variation in single-cell RNA sequencing data. Using 16 publicly available and simulated datasets, we show that a binarized representation of single-cell expression data accurately represents biological variation and reveals the relative abundance of transcripts more robustly than counts.

Download Full-text

Out-of-distribution prediction with disentangled representations for single-cell RNA sequencing data

10.1101/2021.09.01.458535 ◽

2021 ◽

Author(s):

Mohammad Lotfollahi ◽

leander Dony ◽

Harshita Agarwala ◽

Fabian J Theis

Keyword(s):

Gene Expression ◽

Single Cell ◽

Rna Sequencing ◽

State Of The Art ◽

Representation Learning ◽

Cell Types ◽

Biological Variation ◽

Sequencing Data ◽

Variational Autoencoder ◽

Single Cell Rna Sequencing

Learning robust representations can help uncover underlying biological variation in scRNA-seq data. Disentangled representation learning is one approach to obtain such informative as well interpretable representations. Here, we learn disentangled representations of scRNA-seq data using β-variational autoencoder (β-VAE) and apply the model for out-of-distribution (OOD) prediction. We demonstrate accurate gene expression predictions for cell types absent from training in a perturbation and a developmental dataset. We further show that β-VAE outperforms a state-of-the-art disentanglement method for scRNA-seq in OOD prediction while achieving better disentanglement performance.

Download Full-text

A monotonicity-based gene clustering algorithm for enhancing clarity in single-cell RNA sequencing data

10.1101/2020.12.20.423308 ◽

2020 ◽

Author(s):

Victor Wang ◽

Pietro Antonio Cicalese ◽

Chandra Mohan

Keyword(s):

Single Cell ◽

Rna Sequencing ◽

Clustering Algorithm ◽

Clustering Algorithms ◽

Expression Patterns ◽

Gene Clustering ◽

Sequencing Data ◽

Technical Noise ◽

Cell Clustering ◽

Single Cell Rna Sequencing

AbstractSingle-cell RNA sequencing (scRNA-seq) technologies and analysis tools have allowed for meaningful insight into the roles and relationships of cells. However, high dimensionality, frequent dropout values, and technical noise remain prevalent challenges for scRNA-seq data, obscuring the already complex expression patterns. To address several shortcomings in commonly used distance metrics, we present a monotonicity-based distance metric designed to enhance the clarity of scRNA-seq data. We apply our metric in a gene clustering algorithm, which we run on several biological datasets. We compare our results to those generated by popular clustering algorithms to demonstrate that our algorithm has substantial ability to improve the accuracy of subsequent cell clustering.

Download Full-text

Application of HDBSСAN Method for Clustering scRNA-seq Data

Proceedings of the Institute for System Programming of RAS ◽

10.15514/ispras-2020-32(5)-8 ◽

2020 ◽

Vol 32 (5) ◽

pp. 111-120

Author(s):

Maria Andreevna Akimenkova ◽

Anna Anatolyevna Maznina ◽

Anton Yurievich Naumov ◽

Evgeny Andreevich Karpulevich

Keyword(s):

Feature Selection ◽

Dimensionality Reduction ◽

Single Cell ◽

Rna Sequencing ◽

Cell Types ◽

Adjusted Rand Index ◽

Clustering Method ◽

Clustering Problem ◽

Single Cell Rna Sequencing

One of the main tasks in the analysis of single cell RNA sequencing (scRNA-seq) data is the identification of cell types and subtypes, which is usually based on some method of clustering. There is a number of generally accepted approaches to solving the clustering problem, one of which is implemented in the Seurat package. In addition, the quality of clustering is influenced by the use of preprocessing algorithms, such as imputation, dimensionality reduction, feature selection, etc. In the article, the HDBSCAN hierarchical clustering method is used to cluster scRNA-seq data. For a more complete comparison Experiments and comparisons were made on two labeled datasets: Zeisel (3005 cells) and Romanov (2881 cells). To compare the quality of clustering, two external metrics were used: Adjusted Rand index and V-measure. The experiments demonstrated a higher quality of clustering by the HDBSCAN method on the Zeisel dataset and a poorer quality on the Romanov dataset.

Download Full-text

A Single-cell Transcriptomic Atlas of the Developing Chicken Limb

10.1101/598227 ◽

2019 ◽

Cited By ~ 1

Author(s):

Christian Feregrino ◽

Fabio Sacher ◽

Oren Parnas ◽

Patrick Tschopp

Keyword(s):

Pattern Formation ◽

Single Cell ◽

Rna Sequencing ◽

Cell Types ◽

Cell Populations ◽

Cell Type ◽

Data Set ◽

Cellular Resolution ◽

Single Cell Rna Sequencing ◽

Type Specification

AbstractBackgroundThrough precise implementation of distinct cell type specification programs, differentially regulated in both space and time, complex patterns emerge during organogenesis. Thanks to its easy experimental accessibility, the developing chicken limb has long served as a paradigm to study vertebrate pattern formation. Through decades’ worth of research, we now have a firm grasp on the molecular mechanisms driving limb formation at the tissue-level. However, to elucidate the dynamic interplay between transcriptional cell type specification programs and pattern formation at its relevant cellular scale, we lack appropriately resolved molecular data at the genome-wide level. Here, making use of droplet-based single-cell RNA-sequencing, we catalogue the developmental emergence of distinct tissue types and their transcriptome dynamics in the distal chicken limb, the so-called autopod, at cellular resolution.ResultsUsing single-cell RNA-sequencing technology, we sequenced a total of 17,628 cells coming from three key developmental stages of chicken autopod patterning. Overall, we identified 23 cell populations with distinct transcriptional profiles. Amongst them were small, albeit essential populations like the apical ectodermal ridge, demonstrating the ability to detect even rare cell types. Moreover, we uncovered the existence of molecularly distinct sub-populations within previously defined compartments of the developing limb, some of which have important signaling functions during autopod pattern formation. Finally, we inferred gene co-expression modules that coincide with distinct tissue types across developmental time, and used them to track patterning-relevant cell populations of the forming digits.ConclusionsWe provide a comprehensive functional genomics resource to study the molecular effectors of chicken limb patterning at cellular resolution. Our single-cell transcriptomic atlas captures all major cell populations of the developing autopod, and highlights the transcriptional complexity in many of its components. Finally, integrating our data-set with other single-cell transcriptomics resources will enable researchers to assess molecular similarities in orthologous cell types across the major tetrapod clades, and provide an extensive candidate gene list to functionally test cell-type-specific drivers of limb morphological diversification.

Download Full-text