A Bayesian Framework to Account for Complex Non-Genetic Factors in Gene Expression Levels Greatly Increases Power in eQTL Studies
Open accessThe DOI matches its Crossref record; 34 of the 34 references it cites were found in a registry or an index.
Abstract
Gene expression measurements are influenced by a wide range of factors, such as the state of the cell, experimental conditions and variants in the sequence of regulatory regions. To understand the effect of a variable of interest, such as the genotype of a locus, it is important to account for variation that is due to confounding causes. Here, we present VBQTL, a probabilistic approach for mapping expression quantitative trait loci (eQTLs) that jointly models contributions from genotype as well as known and hidden confounding factors. VBQTL is implemented within an efficient and flexible inference framework, making it fast and tractable on large-scale problems. We compare the performance of VBQTL with alternative methods for dealing with confounding variability on eQTL mapping datasets from simulations, yeast, mouse, and human. Employing Bayesian complexity control and joint modelling is shown to result in more precise estimates of the contribution of different confounding factors resulting in additional associations to measured transcript levels compared to alternative approaches. We present a threefold larger collection of cis eQTLs than previously found in a whole-genome eQTL scan of an outbred human population. Altogether, 27% of the tested probes show a significant genetic association in cis, and we validate that the additional eQTLs are likely to be real by replicating them in different sets of individuals. Our method is the next step in the analysis of high-dimensional phenotype data, and its application has revealed insights into genetic regulation of gene expression by demonstrating more abundant cis-acting eQTLs in human than previously shown. Our software is freely available online at http://www.sanger.ac.uk/resources/software/peer/.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures






Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Elastic-net models were trained using nested cross-validation from genotype and normalized gene expression data adjusted for age, sex, smoking status (current or not), the first 60 probabilistic estimation of expression residuals50 factors, and the first five ancestry-based principal components.”
From Genomic and transcriptomic analyses of aortic stenosis enhance therapeutic target discovery and disease prediction (2026) · cites this paper as [50]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Age, sex, smoking status (current or not), the first 60 Probabilistic Estimation of Expression Residuals (PEER)61 factors and the first five ancestry-based principal components were included as covariates.”
From Integrative genomic analyses identify candidate causal genes for calcific aortic valve stenosis involving tissue-specific regulation (2024) · cites this paper as [61]
The check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“Some methods (specifically sva and PEER) that leveraged all data across all sites were quite successful at ameliorating the high rate of false positives (Fig. 1, Supplementary Fig. 2), removing 85.1% to 87.7% of the original total false positive DEGs.”
From Detecting and correcting systematic variation in large-scale RNA sequencing data (2014) · cites this paper as [30]
Experimental — accuracy not yet measuredIn the citing paper
“However, the majority of RNA-seq false positives (>85%) could successfully be removed by subtracting the effects of latent variables identified by either sva26,27 or PEER30, which could be achieved by jointly analyzing the set of measurements of all genes across multiple sites, without a decrease in the sensitivity or specificity of DEG detection at each site or across sites.”
From Detecting and correcting systematic variation in large-scale RNA sequencing data (2014) · cites this paper as [30]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Information theory, inference, and learning algorithms
(2004) · Choice Reviews Online · doi:10.5860/choice.41-5949 · Cited by 6,362
A haplotype map of the human genome
The International HapMap Consortium (2005) · Nature · doi:10.1038/nature04226 · Cited by 5,858
Mapping mendelian factors underlying quantitative traits using RFLP linkage maps.
Lander, E. S., & Botstein, D. (1989) · Genetics · doi:10.1093/genetics/121.1.185 · Cited by 5,235
Information Theory, Inference, and Learning Algorithms
(2004) · Kybernetes · doi:10.1108/03684920410534506 · Cited by 4,937
Empirical threshold values for quantitative trait mapping.
Churchill, G. A., & Doerge, R. W. (1994) · Genetics · doi:10.1093/genetics/138.3.963 · Cited by 4,935
Bayesian Learning for Neural Networks
Neal, R. M. (1996) · Lecture Notes in Statistics · doi:10.1007/978-1-4612-0745-0 · Cited by 4,276 · cited 3×
“[21] Neal RM (1996) Bayesian Learning for Neural Networks.”
BackgroundMethodAn Introduction to Variational Methods for Graphical Models
Jordan, M. I., Ghahramani, Z., Jaakkola, T. S., & Saul, L. K. (1999) · Machine Learning · doi:10.1023/a:1007665907178 · Cited by 3,791 · cited 2×
“Parameter inference in VBQTL is implemented using variational Bayesian learning [18], a generalisation of the expectation maximisation algorithm.”
MethodCapturing Heterogeneity in Gene Expression Studies by Surrogate Variable Analysis
Leek, J. T., & Storey, J. D. (2007) · PLoS Genetics · doi:10.1371/journal.pgen.0030161 · Cited by 1,930 · cited 1×
“often have an influence on large numbers of genes. We and others have proposed methods to detect and correct for such effects”
MethodRelative Impact of Nucleotide and Copy Number Variation on Gene Expression Phenotypes
Stranger, B. E., Forrest, M. S., Dunning, M., Ingle, C. E., Beazley, C., Thorne, N., Redon, R., Bird, C. P., de Grassi, A., Lee, C., et al. (2007) · Science · doi:10.1126/science.1136678 · Cited by 1,723 · cited 3×
“The additional VBeQTLs have very similar characteristics to the standard eQTLs, being concentrated around the gene start (Figure 6c, 6f), in line with previous results [24].”
BackgroundResultg:Profiler—a web-based toolset for functional profiling of gene lists from large-scale experiments
Reimand, J., Kull, M., Peterson, H., Hansen, J., & Vilo, J. (2007) · Nucleic Acids Research · doi:10.1093/nar/gkm226 · Cited by 1,545 · cited 1×
“We used the ordered GO profiling method [27] to find significantly enriched GO categories for 30 genes most affected by each factor.”
Method
Cited by
Items 101–200 of 517
Brynedal, B., Choi, J., Raj, T., Bjornson, R. D., Stranger, B. E., Neale, B. M., Voight, B. F., & Cotsapas, C. J. (2017) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2017.02.004 · Cited by 101
Gong, J., Wang, F., Xiao, B., Panjwani, N., Fan, L., Keenan, K., Avolio, J., Esmaeili, M., Zhang, L., He, G., et al. (2019) · PLoS Genetics · doi:10.1371/journal.pgen.1008007 · Cited by 84
Zhong, J., Jermusyk, A. A., Wu, L., Hoskins, J. W., Collins, I. J., Mocci, E., Zhang, M., Song, L., Chung, C. C., Zhang, T., et al. (2019) · JNCI Journal of the National Cancer Institute · doi:10.1093/jnci/djz246 · Cited by 87
“Gene expression values were adjusted for five principal components, probabilistic estimation of expression residuals factors (31), and gender.”
Epigenomics and human adaptation to high altitude
Julian, C. G. (2017) · Journal of Applied Physiology · doi:10.1152/japplphysiol.00351.2017 · Cited by 92
“Strategies to account for such factors include subject matching and statistical approaches, such as principal components analysis to identify observable batch effects or covariates within the data set that explain variation (52) and probabilistic estimation of expression residuals (90) to account for unknown or “hidde…”
MethodSoltis, A. R., Bateman, N. W., Liu, J., Nguyen, T. M., Franks, T. J., Zhang, X., Dalgard, C. L., Viollet, C., Somiari, S., Yan, C., et al. (2022) · Cell Reports Medicine · doi:10.1016/j.xcrm.2022.100819 · Cited by 91
Nuclear genetic regulation of the human mitochondrial transcriptome
Ali, A. T., Boehme, L., Carbajosa, G., Seitan, V. C., Small, K. S., & Hodgkinson, A. (2019) · eLife · doi:10.7554/elife.41927 · Cited by 83
Covariate Powered Cross-Weighted Multiple Testing
Ignatiadis, N., & Huber, W. K. (2021) · Journal of the Royal Statistical Society Series B (Statistical Methodology) · doi:10.1111/rssb.12411 · Cited by 100
“Once they are a matter of fact, it is sometimes possible to remove them by mapping the data to a new set of properly “normalized” and “batch-corrected” variables (Leek and Storey, 2008; Stegle et al., 2010; Wang et al., 2017).”
BackgroundNormalizing RNA-Sequencing Data by Modeling Hidden Covariates with Prior Knowledge
Mostafavi, S., Battle, A., Zhu, X., Urban, A. E., Levinson, D., Montgomery, S. B., & Koller, D. (2013) · PLoS ONE · doi:10.1371/journal.pone.0068141 · Cited by 77
“For example, PEER [7] is a special case of this scenario.”
BackgroundMethodResultInfluentialHeinig, M., Adriaens, M. E., Schäfer, S., van Deutekom, H. W. M., Lodder, E. M., Ware, J. S., Schneider-Lunitz, V., Felkin, L. E., Creemers, E. E., Meder, B., et al. (2017) · Genome biology · doi:10.1186/s13059-017-1286-z · Cited by 81
“Within the eQTL analyses we addressed any residual confounding by including latent confounder estimates through the PEER procedure [62].”
MethodUncovering genetic mechanisms of hypertension through multi-omic analysis of the kidney
Eales, J., Jiang, X., Xu, X., Saluja, S., Akbarov, A., Cano-Gamez, E., McNulty, M. T., Finan, C., Guo, H., Wystrychowski, W., et al. (2021) · Nature Genetics · doi:10.1038/s41588-021-00835-w · Cited by 76
Runcie, D. E., & Mukherjee, S. (2013) · Genetics · doi:10.1534/genetics.113.151217 · Cited by 76
“…economics, finance, and social sciences ( Fan et al. , 2011), as well as with genetic associ-24 ation studies (e.g., Engelhardt and Stephens 2010; Stegle et al. 2010; Parts et al. 2011) demonstrates that sparsity (or shrinkage) is necessary to perform robust inference on high-dimensional data…”
BackgroundMethodThe homotopic connectivity of the functional brain: a meta-analytic approach
Mancuso, L., Costa, T., Nani, A., Manuello, J., Liloia, D., Gelmini, G., Panero, M., Duca, S., & Cauda, F. (2019) · Scientific Reports · doi:10.1038/s41598-019-40188-3 · Cited by 73
Tissue-specific sex differences in human gene expression
Kassam, I., Wu, Y., Yang, J., Visscher, P. M., & McRae, A. F. (2019) · Human Molecular Genetics · doi:10.1093/hmg/ddz090 · Cited by 69
Efficient inference in matrix-variate Gaussian models with iid observation noise
Stegle, O., Lippert, C., Mooij, J. M., Lawrence, N. D., Borgwardt, K. M., & Shawe-Taylor, J. (2011) · Radboud Repository (Radboud University) · Cited by 67
Self-citationUBD modifies APOL1 -induced kidney disease risk
Zhang, J.-Y., Wang, M., Tian, L., Genovese, G., Yan, P., Wilson, J. G., Thadhani, R., Mottl, A. K., Appel, G. B., Bick, A. G., et al. (2018) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1716113115 · Cited by 71
“By using the data from the Genetic European Variation in Health and Disease, A European Medical Sequencing Consortium (GEUVADIS) Project (21) and normalizing the RNA sequencing data to control for confounding factors (22, 23), we first checked for genes with ciseQTL peaks (FDR ≤ 0.”
MethodContext Specific and Differential Gene Co-expression Networks via Bayesian Biclustering
Gao, C., McDowell, I. C., Zhao, S., Brown, C. D., & Engelhardt, B. E. (2016) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1004791 · Cited by 76
“In gene expression data, observed covariates or unobserved confounding effects may systematically influence variation in the observation [38, 40, 41].”
BackgroundCritical reasoning on causal inference in genome-wide linkage and association studies
Li, Y., Tesson, B. M., Churchill, G. A., & Jansen, R. C. (2010) · Trends in Genetics · doi:10.1016/j.tig.2010.09.002 · Cited by 64
“The effects of hidden factors affecting larger numbers of traits can be detected and corrected for by dimension-reduction methods ([28-32]).”
MethodSex bias in MHC I-associated shaping of the adaptive immune system
Schneider‐Hohendorf, T., Görlich, D., Savola, P., Kelkka, T., Mustjoki, S. M., Groß, C. C., Owens, G. C., Klotz, L., Dornmair, K., Wiendl, H., et al. (2018) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1716146115 · Cited by 73
“To determine possible hidden factors in addition to known covariates, we used the PEER package (33, 34).”
MethodFeng, H., Zhang, X., & Zhang, C. (2015) · Nature Communications · doi:10.1038/ncomms8816 · Cited by 64
Hoffman, G. E., Bendl, J. M., Girdhar, K., Schadt, E. E., & Roussos, P. (2019) · Nucleic Acids Research · doi:10.1093/nar/gkz808 · Cited by 78
“Covariates also included 10 PEER components (51) gender, and GC content of reads for each peak for each sample.”
BackgroundDecreased STARD10 Expression Is Associated with Defective Insulin Secretion in Humans and Mice
Carrat, G. R., Hu, M., Nguyen-Tu, M.-S., Chabosseau, P., Gaulton, K. J., van de Bunt, M., Siddiq, A., Falchi, M., Thurner, M., Canouil, M., et al. (2017) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2017.01.011 · Cited by 62
“This was followed by rank normalization per exon, after which 30 hidden factors (accounting for non-genic variability in the samples) were inferred from the count matrix using PEER.”
MethodGlobal Properties and Functional Complexity of Human Gene Regulatory Variation
Gaffney, D. J. (2013) · PLoS Genetics · doi:10.1371/journal.pgen.1003501 · Cited by 58
“Computational methods to remove batch technical artefacts may also inadvertently remove trans eQTLs that affect large numbers of loci [35].”
MethodIdentification of trans-eQTLs using mediation analysis with multiple mediators
Shan, N., Wang, Z., & Hou, L. (2019) · BMC Bioinformatics · doi:10.1186/s12859-019-2651-6 · Cited by 81
“…to improve trans-eQTL detection, such as reducing the multiple-testing burden based on pairwise partial correlations from the gene expression data to increase power [3], and constructing or selecting variables to control for unmeasured confounders that may lead to spurious association [4 – 6].”
BackgroundPignon, C. P., Fernandes, S. B., Valluru, R., Bandillo, N. B., Lozano, R. D., Buckler, E. S., Gore, M. A., Long, S. P., Brown, P. J., & Leakey, A. D. B. (2021) · PLANT PHYSIOLOGY · doi:10.1093/plphys/kiab395 · Cited by 64
“, 2019), 29 Peer factors (Stegle et al., 2010) and 5 multidimensional scaling factors were used as covariates.”
MethodHulur, I., Gamazon, E. R., Skol, A. D., Xicola, R. M., Llor, X., Onel, K., Ellis, N. A., & Kupfer, S. S. (2015) · BMC Genomics · doi:10.1186/s12864-015-1292-z · Cited by 68
“A PCA-based approach was used to correct for confounding variability due to hidden factors in the gene expression data [85] as outlined in the Supplementary Methods (Additional file 8).”
MethodRoman, T. S., Marvelle, A. F., Fogarty, M. P., Vadlamudi, S., Gonzalez, A. J., Buchkovich, M. L., Huyghe, J. R., Fuchsberger, C., Jackson, A., Wu, Y., et al. (2015) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2015.10.016 · Cited by 63
Jamkhande, P. G., Ghante, M. H., & Ajgunde, B. R. (2017) · Bulletin of Faculty of Pharmacy Cairo University · doi:10.1016/j.bfopcu.2017.10.001 · Cited by 59
High-resolution mapping of cis -regulatory variation in budding yeast
Kita, R., Venkataram, S., Zhou, Y., & Fraser, H. B. (2017) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1717421114 · Cited by 63
“We thenusedPEER to discover hidden covariates (k = 15), and these covariates were regressed out (76).”
BackgroundMethodInfluentialTranscriptomic Insight Into the Polygenic Mechanisms Underlying Psychiatric Disorders
Hernandez, L. M., Kim, M., Hoftman, G. D., Haney, J. R., de la Torre-Ubieta, L., Paşaniuc, B., & Gandal, M. J. (2020) · Biological Psychiatry · doi:10.1016/j.biopsych.2020.06.005 · Cited by 69
“Recent developments in single-cell and long-read sequencing enable interrogation of individual cells and full-length isoforms, respectively (21).”
Walsh, A. M., Whitaker, J. W., Huang, C., Cherkas, Y., Lamberth, S. L., Brodmerkel, C. M., Curran, M. E., & Dobrin, R. (2016) · Genome biology · doi:10.1186/s13059-016-0948-6 · Cited by 64
“Prior to eQTL detection, the gene expression data were adjusted by the first ten principal components (PCs) from principal component analysis and transformed with inverse normal transformation as in previous eQTL studies [14, 18, 53].”
MethodZhong, Y., Perera, M. A., & Gamazon, E. R. (2019) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2019.04.009 · Cited by 65
Gene regulation underlies environmental adaptation in house mice
Mack, K. L., Ballinger, M. A., Phifer‐Rixey, M., & Nachman, M. W. (2018) · Genome Research · doi:10.1101/gr.238998.118 · Cited by 68
“S19, S20; Stegle et al. 2010, 2012).”
Proteins Associated with Risk of Kidney Function Decline in the General Population
Grams, M. E., Surapaneni, A., Chen, J., Zhou, L., Yu, Z., Dutta, D., Welling, P. A., Chatterjee, N., Zhang, J., Arking, D. E., et al. (2021) · Journal of the American Society of Nephrology · doi:10.1681/asn.2020111607 · Cited by 62
“GWAS were performed using the inverserank normalized residuals of the log-transformed protein regressed on age, sex, the first ten genetic principal components, and the first ten protein principal components, with the latter included to remove variation due to latent factors, such as technical variance in the proteomi…”
MethodSajuthi, S. P., Everman, J. L., Jackson, N. D., Saef, B. A., Rios, C. L., Moore, C. M., Mak, A. C. Y., Eng, C. S., Fairbanks-Mahnke, A., Salazar, S., et al. (2022) · Nature Communications · doi:10.1038/s41467-022-28973-7 · Cited by 61
Lona-Durazo, F., Hernandez-Pacheco, N., Fan, S., Zhang, T., Choi, J., Kovacs, M. A., Loftus, S. K., Le, P., Edwards, M., Fortes-Lima, C. A., et al. (2019) · BMC Genetics · doi:10.1186/s12863-019-0765-5 · Cited by 53
“To account for hidden factors driving expression variabilities, a set of covariates were further identified using the PEER method [55] and applied to calculate the normalized expression matrices.”
MethodExtent, Causes, and Consequences of Small RNA Expression Variation in Human Adipose Tissue
Parts, L., Hedman, Å. K., Keildson, S., Knights, A. J., Abreu-Goodger, C., van de Bunt, M., Guerra-Assunção, J. A., Bartonicek, N., van Dongen, S., Mägi, R., et al. (2012) · PLoS Genetics · doi:10.1371/journal.pgen.1002704 · Cited by 50
“Since it has been demonstrated that unmeasured confounders similarly have an influence on expression levels [52,53], we tested whether applying the Bayesian factor analysis package PEER [54], to account for these confounders, increases the number of discoveries.”
MethodSelf-citationSrinivasan, S., Treacy, R., Herrero, T. C., Olsen, R. N., Leonardo, T. R., Zhang, X., DeHoff, P., To, C. C., Poling, L., Fernando, A., et al. (2020) · Cell Reports Medicine · doi:10.1016/j.xcrm.2020.100013 · Cited by 63
Barbeira, A., Melia, O., Liang, Y., Bonazzola, R., Wang, G., Wheeler, H. E., Aguet, F., Ardlie, K., Wen, X., & Im, H. K. (2020) · Genetic Epidemiology · doi:10.1002/gepi.22346 · Cited by 63
“For eQTL analyses, latent factor covariates were calculated using probabilistic estimation of expression re-siduals (PEER) (Stegle, Parts, Durbin, & Winn, 2010) as (Ongen, Buil, Brown, Dermitzakis, & Dela-neau, 2016) was used for cis ‐ eQTL mapping in each tissue.”
MethodOn Negative Heritability and Negative Estimates of Heritability
Steinsaltz, D., Dahl, A. W., & Wachter, K. W. (2020) · Genetics · doi:10.1534/genetics.120.303161 · Cited by 56
Grasping nettles: cellular heterogeneity and other confounders in epigenome-wide association studies
Liang, L., & Cookson, W. (2014) · Human Molecular Genetics · doi:10.1093/hmg/ddu284 · Cited by 54
“High dimension methods such as principal component analyses (PCA) and multidimensional scaling (MDS) can be used effectively to estimate and adjust for such factors when they are sufficiently strong (12,13).”
Kolberg, L., Kerimov, N., Peterson, H., & Alasoo, K. (2020) · eLife · doi:10.7554/elife.58705 · Cited by 55
The Relationship of Common Risk Variants and Polygenic Risk for Schizophrenia to Sensorimotor Gating
Roussos, P., Giakoumaki, S. G., Zouraraki, C., Fullard, J. F., Karagiorga, V. E., Tsapakis, E.-M., Petraki, Z., Siever, L. J., Lencz, T., Malhotra, A. K., et al. (2015) · Biological Psychiatry · doi:10.1016/j.biopsych.2015.06.019 · Cited by 52
MODEM: multi-omics data envelopment and mining in maize
Liu, H., Wang, F., Xiao, Y., Tian, Z., Wen, W., Zhang, X., Chen, X., Liu, N., Li, W., Liu, L., et al. (2016) · Database · doi:10.1093/database/baw117 · Cited by 51
“The top six hidden confounding factors determined to be contributing to expression variability by Bayesian factor analysis (implemented in PEER) (17) were additionally included in the mixed model, in addition to population structure, to examine the validity of association significance for eQTL mapping.”
MethodGenetic mapping across autoimmune diseases reveals shared associations and mechanisms
Lincoln, M. R., Connally, N., Axisa, P., Gasperi, C., Mitrovič, M., van Heel, D. A., Wijmenga, C., Withoff, S., Jonkers, I. H., Padyukov, L., et al. (2024) · Nature Genetics · doi:10.1038/s41588-024-01732-8 · Cited by 58
“We used covariates to adjust for age, sex, the first five principal components and 30 PEER factors 76 .”
D’Antonio, M., Reyna, J., D’Antonio‐Chronowska, A., Bonder, M. J., Jakubosky, D. A., Matsui, H., Smith, E. N., Stegle, O., Nariai, N., & Frazer, K. A. (2019) · eLife · doi:10.7554/elife.48476 · Cited by 55
Self-citationThériault, S., Li, Z., Abner, E., Luan, J., Manikpurage, H. D., Houessou, U., Zamani, P., Briend, M., Estonian Biobank Research Team, Boudreau, D. K., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-46639-4 · Cited by 49
An, J., Wang, L., Li, K., Zeng, Y., Su, Y., Jin, Z., Yu, X., & Si, T. (2017) · Scientific Reports · doi:10.1038/s41598-017-10575-9 · Cited by 48
C-reactive protein upregulates the whole blood expression of CD59 - an integrative analysis
Lepik, K., Annilo, T., Kukuškina, V., eQTLGen Consortium, Kisand, K., Kutalik, Z., Peterson, P., & Peterson, H. (2017) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1005766 · Cited by 48
The landscape of hervRNAs transcribed from human endogenous retroviruses across human body sites
She, J., Du, M., Xu, Z., Jin, Y., Li, Y., Zhang, D., Tao, C., Chen, J., Wang, J., & Yang, E. (2022) · Genome biology · doi:10.1186/s13059-022-02804-w · Cited by 49
“To detect hidden batch effects in the HERV expression data, we applied the probabilistic estimation of expression residuals (PEER) method for each body site [77].”
MethodGenetics of Gene Expression in the Aging Human Brain Reveal TDP-43 Proteinopathy Pathophysiology
Yang, H.-S., White, C. C., Klein, H.-U., Yu, L., Gaiteri, C., Ma, Y., Felsky, D., Mostafavi, S., Petyuk, V. A., Sperling, R. A., et al. (2020) · Neuron · doi:10.1016/j.neuron.2020.05.010 · Cited by 47
Effect of Normalization on Statistical and Biological Interpretation of Gene Expression Profiles
Qin, S., Kim, J., Arafat, D., & Gibson, G. C. (2012) · Frontiers in Genetics · doi:10.3389/fgene.2012.00160 · Cited by 53
“This is an intuitively appealing approach that is just beginning to gain traction following the development of open source algorithms, including Supervised Normalization of Microarrays (SNM; see also Stegle et al., 2010) that facilitate the complex calculations involved.”
MethodResultInfluentialHost genetics and microbiome associations through the lens of genome wide association studies
Weissbrod, O., Rothschild, D., Barkan, E., & Segal, E. (2018) · Current Opinion in Microbiology · doi:10.1016/j.mib.2018.05.003 · Cited by 53
Pang, J., Fu, J., Zong, N., Wang, J., Song, D., Zhang, X., He, C., Fang, T., Zhang, H., Fan, Y., et al. (2019) · The Plant Journal · doi:10.1111/tpj.14193 · Cited by 58
“The hidden factors of gene expression variability were analyzed using PEER (Stegle et al., 2010), which uses a Bayesian framework to account for complex non-genetic factors.”
MethodLocal Adaptation of Sun-Exposure-Dependent Gene Expression Regulation in Human Skin
Kita, R., & Fraser, H. B. (2016) · PLoS Genetics · doi:10.1371/journal.pgen.1006382 · Cited by 46
Richard, A. C., Peters, J. E., Yarkoni, N. S., Lee, J. C., Hawley, E., Meylan, F., Siegel, R. M., Lyons, P. A., & Smith, K. G. C. (2018) · PLoS Genetics · doi:10.1371/journal.pgen.1007458 · Cited by 46
“Microarray mRNA expression was normalized by robust multiarray averaging using the oligo package [71] and adjusted with PEER [72].”
MethodData-based RNA-seq simulations by binomial thinning
Gerard, D. (2020) · BMC Bioinformatics · doi:10.1186/s12859-020-3450-9 · Cited by 57
“Principle component analysis (PCA) [Hotelling, 1933], Sparse singular value decomposition (SSVD) [Yang et al., 2014], Independent component analysis (ICA) [Hyvärinen and Oja, 2000], Factors and loadings by adaptive shrinkage (flash), an empirical Bayes matrix factorization approach proposed in [Wang and Stephens, 2018…”
MethodInfluentialBenaglio, P., D’Antonio‐Chronowska, A., Ma, W., Yang, F., Greenwald, W. W. Y., Donovan, M. K. R., DeBoever, C., Li, H., Drees, F., Singhal, S., et al. (2019) · Nature Genetics · doi:10.1038/s41588-019-0499-3 · Cited by 45
Accounting for non-genetic factors by low-rank representation and sparse regression for eQTL mapping
Yang, C., Wang, L., Zhang, S., & Zhao, H. (2013) · Bioinformatics · doi:10.1093/bioinformatics/btt075 · Cited by 42
Cubillos, F. A., Yansouni, J., Khalili, H., Balzergue, S., Elftieh, S., Martin‐Magniette, M., Serrand, Y., Lepiniec, L., Baud, S., Dubreucq, B., et al. (2012) · BMC Genomics · doi:10.1186/1471-2164-13-117 · Cited by 41
“We applied the VBQTL (Variational Bayesian QTL Mapper) approach as previously described [24] on the dataset obtained in the Cvi-0 × Col-0 population, using one known factor to model hidden confounding sources.”
BackgroundMethodInfluentialEvolutionarily conserved regulation of sleep by epidermal growth factor receptor signaling
Lee, D. A., Liu, J., Hong, Y. K., Lane, J. M., Hill, A. J., Hou, S., Wang, H., Oikonomou, G., Pham, U., Engle, J., et al. (2019) · Science Advances · doi:10.1126/sciadv.aax4249 · Cited by 43
Yan, Y., Liu, H., Abedini, A., Sheng, X., Palmer, M., Li, H., & Suszták, K. (2024) · Nature Communications · doi:10.1038/s41467-024-45295-y · Cited by 41
“…these CpG sites (constituting 673,614 SNP – CpG pairs) were tested with an additive linear model fi tted with covariates including general variables (sample collection site, age, sex, top fi ve genotype PCs, degree of bisul fi te conversion, sample plate and sentrix position) and PEER factors 6,70 .”
MethodBuilding, benchmarking, and exploring perturbative maps of transcriptional and morphological data
Çelik, S., Hütter, J.-C., Carlos, S. M., Lazar, N. H., Mohan, R., Tillinghast, C. A., Biancalani, T., Fay, M. M., Earnshaw, B. A., & Haque, I. S. (2024) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1012463 · Cited by 48
Predicted loss and gain of function mutations in ACO1 are associated with erythropoiesis
Oskarsson, G. R., Oddsson, Á., Magnússon, M. K., Kristjansson, R. P., Halldorsson, G. H., Ferkingstad, E., Zink, F., Helgadóttir, A., Ivarsdottir, E. V., Arnadottir, G. A., et al. (2020) · Communications Biology · doi:10.1038/s42003-020-0921-5 · Cited by 42
Reverse GWAS: Using genetics to identify and model phenotypic subtypes
Dahl, A. W., Cai, N., Ko, A., Laakso, M., Pajukanta, P., Flint, J., & Zaitlen, N. (2019) · PLoS Genetics · doi:10.1371/journal.pgen.1008009 · Cited by 42
“These applications resemble methods that infer heterogeneity 26 from, for example, population structure [35] or technical artifacts [36,37].”
MethodInsight into genetic predisposition to chronic lymphocytic leukemia from integrative epigenomics
Speedy, H. E., Beekman, R., Chapaprieta, V., Orlando, G., Law, P., Martin‐García, D., Gutiérrez‐Abril, J., Catovsky, D., Beà, S., Clot, G., et al. (2019) · Nature Communications · doi:10.1038/s41467-019-11582-2 · Cited by 41
Ellinghaus, D., Uellendahl-Werth, F., Maj, C., Borisov, O., Juzėnas, S., Wacker, E. M., Jørgensen, I. F., Steiert, T., Bej, S., International PSC Study Group, et al. (2022) · Communications Biology · doi:10.1038/s42003-022-03031-6 · Cited by 41
Chen, M., & Zhou, X. (2017) · Scientific Reports · doi:10.1038/s41598-017-13665-w · Cited by 37
“For example, principal component analysis (PCA) or factor models extract the principal components or factors from the set of control genes as surrogates for the confounding factors [25, 27-29].”
MethodThe Dissection of Expression Quantitative Trait Locus Hotspots
Tian, J., Keller, M. P., Broman, A. T., Kendziorski, C. M., Yandell, B. S., Attie, A., & Broman, K. W. (2016) · Genetics · doi:10.1534/genetics.115.183624 · Cited by 52
Kasela, S., Ortega, V. E., Martorella, M., Garudadri, S., Nguyen, J., Ampleford, E., Pasanen, A., Nerella, S., Buschur, K. L., Barjaktarevic, I. Z., et al. (2021) · Genome Medicine · doi:10.1186/s13073-021-00866-2 · Cited by 40
“As covariates in the model, we used 15 PEER factors [36], 4 genotype principal components and sex imputed from genotype data.”
MethodUnique DNA Methylation Patterns in Offspring of Hypertensive Pregnancy
Julian, C. G., Pedersen, B. S., Salmón, C. S., Yang, I. V., Gonzales, M., Vargas, E., MOORE, L. G., & Schwartz, D. A. (2015) · Clinical and Translational Science · doi:10.1111/cts.12346 · Cited by 37
Prediction of gene expression with cis-SNPs using mixed models and regularization methods
Zeng, P., Zhou, X., & Huang, S. (2017) · BMC Genomics · doi:10.1186/s12864-017-3759-6 · Cited by 37
“Since the original gene expression measurements were read counts, the PEER normalization [82 – 84] was employed to remove technical variations and batch effects.”
MethodGuertin, J., Kaiser, Y., Manikpurage, H. D., Perrot, N., Bourgeois, R., Couture, C., Wareham, N. J., Bossé, Y., Pîbarot, P., Stroes, E. S. G., et al. (2021) · Circulation Genomic and Precision Medicine · doi:10.1161/circgen.120.003271 · Cited by 37
Chen, Y., Kassam, I., Lau, S. H., Kooner, J. S., Wilson, R., Peters, A., Winkelmann, J., Chambers, J. C., Chow, V. T. K., Khor, C. C., et al. (2021) · Clinical Epigenetics · doi:10.1186/s13148-021-01162-x · Cited by 42
“These included (1) exclusion of lowly expressed genes, defined as genes with expression levels less than the mean expression level of control probes in more than 90% of the samples; (2) variance stabilization [58] and quantile normalization to standardize the distribution of expression levels across samples; (3) accou…”
MethodInterferon-stimulated genes: new platforms and computational approaches
Green, R. R., Ireton, R. C., & Gale, M. (2018) · Mammalian Genome · doi:10.1007/s00335-018-9755-6 · Cited by 41
Li, W., Cai, X. M., Li, H., Song, M., Zhang, C., Yang, Y., Zhang, L., Zhao, L., Liu, W., Wang, L., et al. (2020) · Neuropsychopharmacology · doi:10.1038/s41386-020-00788-4 · Cited by 37
Making sense of genomes of parasitic worms: Tackling bioinformatic challenges
Korhonen, P. K., Young, N. D., & Gasser, R. B. (2016) · Biotechnology Advances · doi:10.1016/j.biotechadv.2016.03.001 · Cited by 39
Adiposity-Dependent Regulatory Effects on Multi-tissue Transcriptomes
Glastonbury, C. A., Viñuela, A., Buil, A., Halldorsson, G. H., Þorleifsson, G., Helgason, H., Þorsteinsdóttir, U., Stefánsson, K., Dermitzakis, E. T., Spector, T. D., et al. (2016) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2016.07.001 · Cited by 39
Integrating GWAS and TWAS to elucidate the genetic architecture of maize leaf cuticular conductance
Lin, M., Qiao, P., Matschi, S., Vasquez, M. F., Ramstein, G. P., Bourgault, R., Mohammadi, M., Scanlon, M. J., Molina, I., Smith, L. G., et al. (2022) · PLANT PHYSIOLOGY · doi:10.1093/plphys/kiac198 · Cited by 37
“To account for inferred confounders, the probabilistic estimation of expression residuals (PEERs; Stegle et al., 2010) approach was applied to the 310 line 20,018 gene matrix of BLUP expression values.”
Han, S. K., McNulty, M. T., Benway, C. J., Wen, P., Greenberg, A., Onuchic-Whitford, A. C., Nephrotic Syndrome Study Network (NEPTUNE), Jang, D., Flannick, J. A., Burtt, N. P., et al. (2023) · Nature Communications · doi:10.1038/s41467-023-37691-7 · Cited by 42
“Single-SNP eQTL analysis Single-SNP cis-EQTL (± 1Mb) analysis was performed with MatrixEQTL (Shabalin, 2012) adjusting for age, sex, batch, 4 genotype PCs, and PEER factors (Parts et al., 2011; Stegle et al., 2010).”
Jermusyk, A. A., Zhong, J., Connelly, K. E., Gordon, N., Perera, S., Abdolalizadeh, E., Zhang, T., O’Brien, A., Hoskins, J. W., Collins, I. J., et al. (2021) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2021.09.002 · Cited by 36
Natri, H. M., Hudjashov, G., Jacobs, G., Kusuma, P., Saag, L., Darusallam, C. C., Metspalu, M., Sudoyo, H., Cox, M. P., Gallego Romero, I., et al. (2022) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2021.11.017 · Cited by 33
Exploiting Gene Expression Variation to Capture Gene-Environment Interactions for Disease
Idaghdour, Y., & Awadalla, P. (2013) · Frontiers in Genetics · doi:10.3389/fgene.2012.00228 · Cited by 34
Genetic risk mechanisms of posttraumatic stress disorder in the human brain
Bharadwaj, R. A., Jaffe, A. E., Chen, Q., Deep‐Soboslay, A., Goldman, A. L., Mighdoll, M. I., Cotoia, J. A., Brandtjen, A. C., Shin, J. H., Hyde, T. M., et al. (2016) · Journal of Neuroscience Research · doi:10.1002/jnr.23957 · Cited by 36
“Including these genome-wide factors in eQTL analyses has been shown to increase statistical power (Leek and Storey, 2008; Stegle et al., 2010).”
BackgroundAn improved and explicit surrogate variable analysis procedure by coefficient adjustment
Lee, S., Sun, W., Wright, F. A., & Zou, F. (2017) · Biometrika · doi:10.1093/biomet/asx018 · Cited by 44
Etheridge, A. S., Gallins, P. J., Jima, D. D., Broadaway, K. A., Ratain, M. J., Schuetz, E. G., Schadt, E. E., Schröder, A., Molony, C. M., Zhou, Y., et al. (2019) · Clinical Pharmacology & Therapeutics · doi:10.1002/cpt.1751 · Cited by 35
“uk/scien ce/tools/ PEER).(39) Following identification of cis-eQTLs in each individual data set, ciseQTLs identified in at least two data sets were included in the combined analysis.”
BackgroundSensitivity, specificity, and reproducibility of RNA-Seq differential expression calls
Łabaj, P. P., & Kreil, D. P. (2016) · Biology Direct · doi:10.1186/s13062-016-0169-7 · Cited by 39
“The number of false positives can be reduced when unwanted variation is removed [16, 17].”
Empirical Bayes Matrix Factorization
Wang, W., & Stephens, M. (2018) · Journal of Machine Learning Research · doi:10.48550/arxiv.1802.06931 · Cited by 62
“Both EB and variational methods have been previously used for this problem (Bishop, 1999; Lim and Teh, 2007; Raiko et al., 2007; Stegle et al., 2010).”
MethodExpression reflects population structure
Brown, B. C., Bray, N., & Pachter, L. S. (2018) · PLoS Genetics · doi:10.1371/journal.pgen.1007841 · Cited by 34
Chen, L., & Zheng, S. (2020) · Informatics in Medicine Unlocked · doi:10.1016/j.imu.2020.100443 · Cited by 33
“PEER factors represent hidden determinants of expression variability including batch effects, experimental confounders, and systemic gene expression effect (e.g., simultaneous activation of many genes by a major pathway or transcription factor).”
InfluentialSimon, L. M., Chen, E. S., Edelstein, L. C., Kong, X., Bhatlekar, S., Rigoutsos, I., Bray, P. F., & Shaw, C. A. (2016) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2016.03.007 · Cited by 32
“Accounting for these trans factors when modeling cis eQTLs improves the power to detect true cis eQTLs by reducing spurious false-positive associations.(32) Therefore, we included these 15 PEER factors—in addition to age, gender, and the first two genotype principal components, which completely account for self-identi-”
BackgroundHan, Z., Xue, W., Tao, L., Lou, Y., Qiu, Y., & Zhu, F. (2019) · Briefings in Bioinformatics · doi:10.1093/bib/bbz036 · Cited by 31
“Among these parameters, PEER factors were obtained by the R package ‘PEER’ based on Bayesian method [48], and population stratification was divided by source of the individuals (Amsterdam or Utrecht).”
MethodInfluentialRegulatory sites for splicing in human basal ganglia are enriched for disease-relevant information
Guelfi, S., D’Sa, K., Botía, J. A., Vandrovcová, J., Reynolds, R. H., Zhang, D., Trabzuni, D., Collado‐Torres, L., Thomason, A., Leyton, P. Q., et al. (2020) · Nature Communications · doi:10.1038/s41467-020-14483-x · Cited by 32
RNA-Seq optimization with eQTL gold standards
Ellis, S. E., Gupta, S., Ashar, F. N., Bader, J. S., West, A. B., & Arking, D. E. (2013) · BMC Genomics · doi:10.1186/1471-2164-14-892 · Cited by 30
“We utilized four data decomposition methods – independent surrogate variable analysis (ISVA), surrogate variable analysis (SVA), principal component analysis (PCA) and PEER [13,22] – to account for unknown covariates.”
MethodVariation-preserving normalization unveils blind spots in gene expression profiling
Roca, C. P., Gomes, S. I. L., Amorim, M. J. B., & Scott-Fordsmand, J. J. (2017) · Scientific Reports · doi:10.1038/srep42460 · Cited by 32
“analyses, such as SVA (Leek and Storey, 2007) or PEER (Stegle et al., 2010), should be applied when necessary, after normalizing.”
Li, S. S., Gilbert, P. B., Carpp, L. N., Pyo, C., Janes, H., Fong, Y., Shen, X., Neidich, S. D., Goodman, D. D., deCamp, A. C., et al. (2019) · Journal of Virology · doi:10.1128/jvi.02041-18 · Cited by 34
“Before eQTL analysis, expression quantifications were normalized by PEER correction (73) and transformed into standard normal distributions.”
MethodKasela, S., Aguet, F., Kim-Hellmuth, S., Brown, B. C., Nachun, D. C., Tracy, R. P., Durda, P., Liu, Y., Taylor, K. D., Johnson, W. C., et al. (2024) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2023.11.013 · Cited by 31
Jang, B., Kailash, B. P., Tokolyi, A., Dredge, W. H., Ravi, A., Jung, S., Naito, T., Kim, B., Kim, M. S., Cho, M., et al. (2026) · Nature Genetics · doi:10.1038/s41588-026-02541-x · Cited by 37
Interpretable generative deep learning: an illustration with single cell gene expression data
Treppner, M., Binder, H., & Hess, M. (2022) · Human Genetics · doi:10.1007/s00439-021-02417-6 · Cited by 33
“Deep generative models can also be employed as an extension to shallow latent variable models, which have been proposed to account for gene expression heterogeneity related to nuisance factors (Yang et al. 2013; Stegle et al. 2010).”
BackgroundAn organism-wide ATAC-seq peak catalog for the bovine and its use to identify regulatory variants
Yuan, C., Tang, L., Lopdell, T., Петров, В. А., Oget, C., Moreira, G. C. M., Duarte, J. L. G., Sartelet, A., Cheng, Z., Salavati, M., et al. (2023) · Genome Research · doi:10.1101/gr.277947.123 · Cited by 31
“Cis eQTL analyses (variantswithin1Mbfromgene ’ sTSS)wereconductedusingresid-uals corrected for hidden PEER factors (Stegle et al. 2010), country (of origin of the samples), and polygenic effects estimated with GenABEL (Aulchenko et al. 2007), under an additive model using QTLtools (Delaneau et al.…”
Mahama, T. (2023) · International Journal of Science and Research Archive · doi:10.30574/ijsra.2023.10.2.0998 · Cited by 29
Metadata
- Authors
- Stegle, O., Parts, L., Durbin, R., & Winn, J.
- Affiliations
- University of Cambridge (GB)
- Wellcome Sanger Institute (GB)
- Microsoft Research (United Kingdom) (GB)
- Corresponding author
- Stegle, Oliver
- Winn, John
- Venue
- PLoS Computational Biology
- Publisher
- Public Library of Science (PLoS)
- Published
- 2010-05-06
- Volume
- 6
- Issue
- 5
- Pages
- e1000770
- ISSN
- 1553-734X (ISSN-L)
- 1553-7358
- PMC
- PMC2865505
- OpenAlex
- W2048612325
- Cited by
- 497
- References
- 34
- Citations by year
- 2026: 24 · 2025: 25 · 2024: 48 · 2023: 31 · 2022: 47 · 2021: 46 · 2020: 42 · 2019: 45 · 2018: 30 · 2017: 39 · 2016: 34 · 2015: 28 · 2014: 13 · 2013: 20 · 2012: 15
- FWCI
- 7.91
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- PLoS Computational Biology (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Max Planck Institute for Plasma Physics (submittedVersion)
- Figshare (submittedVersion, cc-by)
- MPG.PuRe (Max Planck Society) (submittedVersion)
- Funders
- Microsoft Research
- Wellcome Trust (WT077192 ⁄ Z ⁄ 05 ⁄ Z, WT077192)
- Engineering and Physical Sciences Research Council
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Mapping and Diversity in Plants and Animals
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic and phenotypic traits in livestock
- Keywords
- expression quantitative trait loci, eQTL mapping, confounding factors, non-genetic variation, cis-eQTL, gene expression regulation, hidden confounders, statistical power
- MeSH terms
- Animals, Bayes Theorem, Humans, Markov Chains, Models, Genetic, Phenotype, Software, Yeasts, Reproducibility of Results, Models, Statistical, Gene Expression, Internet, Databases, Genetic, Quantitative Trait Loci, Mice
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.