Fast, scalable and accurate differential expression analysis for single cells

Sengupta, Debarka; Rayan, Nirmala Arul; Lim, Michelle Gek Liang; Lim, Bing; Prabhakar, Shyam

doi:10.1101/049734

Cited by 36 publications

(47 citation statements)

References 25 publications

(11 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2H). The slowest tool was SCDE (684 min), as reported in previous studies (Sengupta et al 2016;Jaakkola et al 2017). We next compared the scalability of bigSCale to MAST with respect to samples sizes.…”

Section: Identification Of Differentially Expressed Genesmentioning

confidence: 99%

bigSCale: an analytical framework for big-scale single-cell data

et al. 2018

View full text Add to dashboard Cite

Single-cell RNA sequencing (scRNA-seq) has significantly deepened our insights into complex tissues, with the latest techniques capable of processing tens of thousands of cells simultaneously. Analyzing increasing numbers of cells, however, generates extremely large data sets, extending processing time and challenging computing resources. Current scRNA-seq analysis tools are not designed to interrogate large data sets and often lack sensitivity to identify marker genes. With bigSCale, we provide a scalable analytical framework to analyze millions of cells, which addresses the challenges associated with large data sets. To handle the noise and sparsity of scRNA-seq data, bigSCale uses large sample sizes to estimate an accurate numerical model of noise. The framework further includes modules for differential expression analysis, cell clustering, and marker identification. A directed convolution strategy allows processing of extremely large data sets, while preserving transcript information from individual cells. We evaluated the performance of bigSCale using both a biological model of aberrant gene expression in patient-derived neuronal progenitor cells and simulated data sets, which underlines the speed and accuracy in differential expression analysis. To test its applicability for large data sets, we applied bigSCale to assess 1.3 million cells from the mouse developing forebrain. Its directed down-sampling strategy accumulates information from single cells into index cell transcriptomes, thereby defining cellular clusters with improved resolution. Accordingly, index cell clusters identified rare populations, such as reelin (Reln)-positive Cajal-Retzius neurons, for which we report previously unrecognized heterogeneity associated with distinct differentiation stages, spatial organization, and cellular function. Together, bigSCale presents a solution to address future challenges of large single-cell data sets.[Supplemental material is available for this article.]Single-cell RNA sequencing (scRNA-seq) is at the forefront of techniques to chart molecular properties of individual cells. Recent microfluidic-based methods are scalable to tens of thousands of cells, enabling an unbiased sampling and in-depth characterization without prior knowledge Macosko et al. 2015;Zheng et al. 2017). Consequently, studies are less confined by the number of cells and aim to produce comprehensive cellular atlases of entire tissues, organs, and organisms (Regev et al. 2017). Increasing cell numbers, however, generate extremely large data sets, which extend processing time and challenge computing resources. Current scRNA-seq analysis tools are not designed to analyze data sets larger than thousands of cells and often lack sensitivity and specificity to identify marker genes for cell populations or experimental conditions.To address the challenges of large scRNA-seq data sets, we developed bigSCale, an analytical framework for the sensitive detection of population markers and differentially expressed genes, being scalable to analyze mil...

show abstract

Section: Identification Of Differentially Expressed Genesmentioning

confidence: 99%

bigSCale: an analytical framework for big-scale single-cell data

et al. 2018

View full text Add to dashboard Cite

show abstract

“…Overall performance of 29 scImpute has been shown to to be superior to MAGIC. Parametric modeling of single 30 cell expression is challenging due to our lack of knowledge about possible sources of 31 technical noise and biases [30]. Moreover, there is clear lack of consensus about the 32 choice of probability density function.…”

mentioning

confidence: 99%

“…On the 43 same dataset, we assess the impact of imputation on differential genes prediction. We 44 further investigate mcImpute's ability to recover artificially planted missing values in a 45 matrix of mouse brain cells [30]. Accurate imputation should enhance cell type identity 46 i.e., transcriptomic similarity between cells of identical type.…”

mentioning

confidence: 99%

“…The experiments are carried out on the processed Mousebrain dataset [30]. We 93 artificially removed some counts at random (sub-sampling) in the data to mimic 94 dropout cases and used our algorithms (MF and NNM) to impute the missing values.…”

mentioning

confidence: 99%

“…developmental time points[30]. Profiling of 333 single-cells isolated from the193 mouse brain was done.…”

mentioning

confidence: 99%

See 2 more Smart Citations

McImpute: Matrix completion based imputation for single cell RNA-seq data

Mongia

Sengupta

Majumdar

2018

Preprint

Self Cite

View full text Add to dashboard Cite

Motivation: Single cell RNA sequencing has been proved to be revolutionary for its potential of zooming into complex biological systems. Genome wide expression analysis at single cell resolution, provides a window into dynamics of cellular phenotypes. This facilitates characterization of transcriptional heterogeneity in normal and diseased tissues under various conditions. It also sheds light on development or emergence of specific cell populations and phenotypes. However, owing to the paucity of input RNA, a typical single cell RNA sequencing data features a high number of dropout events where transcripts fail to get amplified. Results:We introduce mcImpute, a low-rank matrix completion based technique to impute dropouts in single cell expression data. On a number of real datasets, application of mcImpute yields significant improvements in separation of true zeros from dropouts, cell-clustering, differential expression analysis, cell type separability and performance of dimensionality reduction techniques for cell visualization. Availability and Implementation: https://github.com/aanchalMongia/McImpute_scRNAseq 7 cell type and understanding the dynamics of transcriptional changes during 8 development [1, 33, 40].9Despite all the goodness, scRNA-seq technologies suffer from a number of sources of 10 technical noise. Most important of these is insufficient input RNA. Due to small 11 quantities transcripts are frequently missed during the reverse transcription step. As a 12 direct consequence, these transcripts are not detected during the sequencing step [12]. 13Often times the lowly expressed genes are the worst hit. Excluding these genes from 14 analysis may not be the best solution as many of the transcription factors and cell 15 surface markers are sacrificed in this process [38]. Added to that, variability in dropout 16 rate across individual cells or cell types, works as a confounding factor for a number of 17 July 5, 2018 1/15 downstream analyses [18, 30]. Hicks and colleagues [8] showed, on a number of 18 scRNA-seq datasets, that the first principal components highly correlate with 19 proportion of dropouts across individual transcriptomes. In summary, there is a 20 standing need for efficient methods to impute scRNA-seq datasets. 21 Very recently, efforts have been made to devise imputation techniques for scRNA-seq 22 data. Most notable of among these are MAGIC [38], scImpute [19] and drImpute [15]. 23 MAGIC uses a neighborhood based heuristic to infer the missing values based on the 24 idea of heat diffusion, altering all gene expression levels including the ones not affected 25 by dropouts. On the other hand, scImpute first estimates which values are affected by 26dropouts based on Gamma-Normal mixture model and then fills the dropout values in a 27 cell by borrowing information of the same gene in other similar cells, which are selected 28 based on the genes unlikely affected by dropout events. Overall performance of 29 scImpute has been shown to to be superior to MAGIC. Parametric modeling of single 30...

show abstract