Capturing Heterogeneity in Gene Expression Studies by "Surrogate Variable Analysis"

Leek, Jeffrey T.; Storey, J. Douglas

doi:10.1371/journal.pgen.0030161.eor

Cited by 175 publications

(273 citation statements)

References 33 publications

(73 reference statements)

Supporting

Mentioning

270

Contrasting

Order By: Relevance

“…To fully exploit single-cell RNA-seq data, we have to account for the random noise inherent to such data sets 20 and, equally important, to account for different hidden factors that might result in gene expression heterogeneity. Although the importance of accounting for unobserved factors is well established in bulk RNA-seq studies [21][22][23] , robust approaches to detect and account for confounding factors in single-cell RNA-seq studies remain to be developed. Here, we describe a computational approach that uses latent variable models to reconstruct such hidden factors from the observed data.…”

Section: A N a Ly S I Smentioning

confidence: 99%

“…This approach can also be used to create a 'corrected' gene expression data set, in which the effect of the identified factor(s) is removed, which can be used as the input for existing analysis methods. scLVM is related to approaches for modeling variability in bulk mRNA expression studies 21,22 and to methods used in genome-wide association studies in which the relatedness between individuals is inferred from genotype 29 and/or expression levels 30 and then accounted for in downstream analyses using linear mixed models.…”

Section: Development Of Sclvm To Account For Effects Of the Cell Cyclementioning

confidence: 99%

“…Additionally, the hidden factors can be accounted for when performing pairwise gene-gene correlation analyses, and further allow 'corrected' residual gene expression data sets to be generated. scLVM is closely related to previous approaches that correct for hidden confounding factors in gene expression data 21,48 and the inference employed to fit hidden factors builds on the PANAMA model 30 .…”

Section: Supplementary Data 1)mentioning

confidence: 99%

“…22 and Choosing the rank of the cell-cycle factor). In general, the P-value distribution of a test statistic on the residual data set 48 , heuristic selection approaches 21 or hierarchical modeling to regularize the effective dimensionality of the hidden factor 22,30 can also be employed (Supplementary Notes).…”

Section: Supplementary Data 1)mentioning

confidence: 99%

See 3 more Smart Citations

Computational analysis of cell-to-cell heterogeneity in single-cell RNA-sequencing data reveals hidden subpopulations of cells

et al. 2015

View full text Add to dashboard Cite

Recent technical developments have enabled the transcriptomes of hundreds of cells to be assayed in an unbiased manner, opening up the possibility that new subpopulations of cells can be found. However, the effects of potential confounding factors, such as the cell cycle, on the heterogeneity of gene expression and therefore on the ability to robustly identify subpopulations remain unclear. We present and validate a computational approach that uses latent variable models to account for such hidden factors. We show that our single-cell latent variable model (scLVM) allows the identification of otherwise undetectable subpopulations of cells that correspond to different stages during the differentiation of naive T cells into T helper 2 cells. Our approach can be used not only to identify cellular subpopulations but also to tease apart different sources of gene expression heterogeneity in single-cell transcriptomes.Single-cell measurements of gene expression, using imaging techniques such as RNA-FiSH (fluorescence in situ hybridization), have provided important insights into the kinetics of transcription and cell-to-cell variation in gene expression [1][2][3] . However, such approaches can examine the expression of only a small number of genes in each experiment, thus restricting our ability to examine co-expression patterns and to robustly identify subpopulations of cells. Protocols have been developed to overcome these limitations by amplifying small quantities of mRNA 4,5 , which, in combination with microfluidics approaches for isolating individual cells 6,7 , have been used to analyze the co-expression of tens to hundreds of genes in single cells 8,9 . These protocols also allow the entire transcriptome of large numbers of single cells to be assayed in an unbiased way. This was initially done using microarrays 10,11 but is more often now done using next-generation sequencing [12][13][14][15] . Such approaches have been used to model early embryogenesis in the mouse 16 and to investigate bimodality in gene expression patterns of differentiating immune cell types 17 .After the generation of single-cell RNA-sequencing (RNA-seq) profiles from hundreds of cells, one goal to identify subpopulations that share a common gene-expression profile. Some of these subpopulations may represent previously unidentified cell types. Additionally, by studying patterns of gene expression in different single cells, insights into the regulatory landscape of each cell population can be obtained.However, methods for identifying subpopulations of cells and modeling their gene regulatory landscapes are only now beginning to emerge 18,19 . To fully exploit single-cell RNA-seq data, we have to account for the random noise inherent to such data sets 20 and, equally important, to account for different hidden factors that might result in gene expression heterogeneity. Although the importance of accounting for unobserved factors is well established in bulk RNA-seq studies [21][22][23] , robust approaches to detect and account for confounding f...

show abstract

Section: A N a Ly S I Smentioning

confidence: 99%

Section: Development Of Sclvm To Account For Effects Of the Cell Cyclementioning

confidence: 99%

Section: Supplementary Data 1)mentioning

confidence: 99%

Section: Supplementary Data 1)mentioning

confidence: 99%

See 2 more Smart Citations

Computational analysis of cell-to-cell heterogeneity in single-cell RNA-sequencing data reveals hidden subpopulations of cells

et al. 2015

View full text Add to dashboard Cite

show abstract

“…We used the residuals of the model as batch-corrected gene expression traits. To additionally account for unknown unwanted sources of variation, we computed surrogate variables of the gene expression matrix by using the R package Surrogate Variable Analysis (SVA) (Leek and Storey 2007;Leek et al 2012). We used the two surrogate variables reported by the software as covariates for expression QTL mapping.…”

Section: Gene Expression Preprocessingmentioning

confidence: 99%

Genetic Influences on Brain Gene Expression in Rats Selected for Tameness and Aggression

Heyne

Lautenschlaeger²,

Nelson

et al. 2014

Preprint

View full text Add to dashboard Cite

Interindividual differences in many behaviors are partly due to genetic differences, but the identification of the genes and variants that influence behavior remains challenging. Here, we studied an F 2 intercross of two outbred lines of rats selected for tame and aggressive behavior toward humans for .64 generations. By using a mapping approach that is able to identify genetic loci segregating within the lines, we identified four times more loci influencing tameness and aggression than by an approach that assumes fixation of causative alleles, suggesting that many causative loci were not driven to fixation by the selection. We used RNA sequencing in 150 F 2 animals to identify hundreds of loci that influence brain gene expression. Several of these loci colocalize with tameness loci and may reflect the same genetic variants. Through analyses of correlations between allele effects on behavior and gene expression, differential expression between the tame and aggressive rat selection lines, and correlations between gene expression and tameness in F 2 animals, we identify the genes Gltscr2, Lgi4, Zfp40, and Slc17a7 as candidate contributors to the strikingly different behavior of the tame and aggressive animals. B EHAVIORAL differences among members of a species are in part due to genetic variation. The identification of the genes and variants that influence behavior remains challenging. In human genome-wide association studies of psychiatric and cognitive traits, the identified loci typically explain only a small fraction of the heritability, i.e., the additive genetic contribution to the trait (Watanabe et al. 2007;Hovatta and Barlow 2008;Deary et al. 2009;Otowa et al. 2009;Calboli et al. 2010;Terracciano et al. 2010;Flint and Munafo 2013;Rietveld et al. 2013;Sokolowska and Hovatta 2013). In experimental crosses in model species, especially between inbred lines, the identified loci (termed quantitative trait loci, QTL), often explain much more of the heritability (Lynch and Walsh 1998). However, in this design the spatial resolution is limited-the QTL are wide and contain many, sometimes hundreds of genes. With the exception of a handful of identified genes with quantitative effects on behavioral traits (Yalcin et al. 2004;Watanabe et al. 2007;McGrath et al. 2009;Bendesky et al. 2012;Goodson et al. 2012), gene identification is particularly difficult for behavioral QTL that often have modest effect sizes (Flint 2003;Willis-Owen and Flint 2006;Wright et al. 2006;Hovatta and Barlow 2008).The causal variants within a QTL can alter the protein sequence encoded by a gene or affect gene expression. Such Musunuru et al. 2010). A few eQTL studies have also recently been carried out to identify candidate genes for behavioral QTL (Hitzemann et al. 2004;De Jong et al. 2011;Saba et al. 2011;Kelly et al. 2012).Here, we studied two lines of rats (Rattus norvegicus) that, starting from one wild population, have been selected for increased tameness and increased aggression toward humans, respectively. These experimental...

show abstract

Identification of Three Rheumatoid Arthritis Disease Subtypes by Machine Learning Integration of Synovial Histologic Features and RNA Sequencing Data

Orange

Agius

DiCarlo

et al. 2018

Arthritis & Rheumatology

162

159

View full text Add to dashboard Cite

Objective In this study, we sought to refine histologic scoring of rheumatoid arthritis (RA) synovial tissue by training with gene expression data and machine learning. Methods Twenty histologic features were assessed in 129 synovial tissue samples (n = 123 RA patients and n = 6 osteoarthritis [OA] patients). Consensus clustering was performed on gene expression data from a subset of 45 synovial samples. Support vector machine learning was used to predict gene expression subtypes, using histologic data as the input. Corresponding clinical data were compared across subtypes. Results Consensus clustering of gene expression data revealed 3 distinct synovial subtypes, including a high inflammatory subtype characterized by extensive infiltration of leukocytes, a low inflammatory subtype characterized by enrichment in pathways including transforming growth factor β, glycoproteins, and neuronal genes, and a mixed subtype. Machine learning applied to histologic features, with gene expression subtypes serving as labels, generated an algorithm for the scoring of histologic features. Patients with the high inflammatory synovial subtype exhibited higher levels of markers of systemic inflammation and autoantibodies. C‐reactive protein (CRP) levels were significantly correlated with the severity of pain in the high inflammatory subgroup but not in the others. Conclusion Gene expression analysis of RA and OA synovial tissue revealed 3 distinct synovial subtypes. These labels were used to generate a histologic scoring algorithm in which the histologic scores were found to be associated with parameters of systemic inflammation, including the erythrocyte sedimentation rate, CRP level, and autoantibody levels. Comparison of gene expression patterns to clinical features revealed a potentially clinically important distinction: mechanisms of pain may differ in patients with different synovial subtypes.

show abstract

Capturing Heterogeneity in Gene Expression Studies by "Surrogate Variable Analysis"

Cited by 175 publications

References 33 publications

Computational analysis of cell-to-cell heterogeneity in single-cell RNA-sequencing data reveals hidden subpopulations of cells

Computational analysis of cell-to-cell heterogeneity in single-cell RNA-sequencing data reveals hidden subpopulations of cells

Genetic Influences on Brain Gene Expression in Rats Selected for Tameness and Aggression

Identification of Three Rheumatoid Arthritis Disease Subtypes by Machine Learning Integration of Synovial Histologic Features and RNA Sequencing Data

Contact Info

Product

Resources

About