Phiclust: a clusterability measure for single-cell transcriptomics reveals phenotypic subpopulations

Mircea, Maria; Hochane, Mazène; Fan, Xueying; Lopes, Susana M. Chuva de Sousa; Garlaschelli, Diego; Semrau, Stefan

doi:10.1186/s13059-021-02590-x

Cited by 10 publications

(6 citation statements)

References 65 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Input the data subset as the "target data" into ClusterDE. It is worth noting that ClusterDE does not provide an automatic decision about whether two clusters should be merged, unlike the methods that directly assess the quality of clusters [13][14][15][16][17]. Instead, ClusterDE focuses on identifying trustworthy post-clustering DE genes as potential cell-type marker genes, enabling researchers to gain biological insights into clusters by investigating the specific genes that distinguish the clusters.…”

Section: Practical Guidelines For Clusterde Usagementioning

confidence: 99%

“…In other words, the clusterfree DE genes and the post-clustering DE genes serve different purposes and are not conceptually comparable. Another stream of methods has been developed to assess the quality of clustering results, e.g., the "purity" of a cluster or if two clusters should be merged [13][14][15][16][17]. However, these methods do not provide a direct statistical test for identifying DE genes, and it remains difficult to determine the threshold for clustering quality above which double dipping is not a concern.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

ClusterDE: a post-clustering differential expression (DE) method robust to false-positive inflation caused by double dipping

Song

et al. 2023

Preprint

View full text Add to dashboard Cite

In typical single-cell RNA-seq (scRNA-seq) data analysis, a clustering algorithm is applied to find putative cell types as clusters, and then a statistical differential expression (DE) test is employed to identify the differentially expressed (DE) genes between the cell clusters. However, this common procedure uses the same data twice, an issue known as “double dipping”: the same data is used twice to define cell clusters as potential cell types and DE genes as potential cell-type marker genes, leading to false-positive cell-type marker genes even when the cell clusters are spurious. To overcome this challenge, we propose ClusterDE, a post-clustering DE method for controlling the false discovery rate (FDR) of identified DE genes regardless of clustering quality, which can work as an add-on to popular pipelines such as Seurat. The core idea of ClusterDE is to generate real-data-based synthetic null data containing only one cluster, as contrast to the real data, for evaluating the whole procedure of clustering followed by a DE test. Using comprehensive simulation and real data analysis, we show that ClusterDE has not only solid FDR control but also the ability to identify cell-type marker genes as top DE genes and distinguish them from housekeeping genes. ClusterDE is fast, transparent, and adaptive to a wide range of clustering algorithms and DE tests. Besides scRNA-seq data, ClusterDE is generally applicable to post-clustering DE analysis, including single-cell multi-omics data analysis.

show abstract

Section: Practical Guidelines For Clusterde Usagementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

ClusterDE: a post-clustering differential expression (DE) method robust to false-positive inflation caused by double dipping

Song

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Deep learning methods are frequently used however often struggle to cluster the high dimensional and increasing size of modern scRNA-seq datasets. [37][38][39] The most widely used linear dimensionality reduction algorithm is the PCA (Principal Component Analysis). Significant principal components can be used for nonlinear dimensionality reduction and to dependably indicate sources of heterogeneity in a dataset.…”

Section: Dimensionality Reductionmentioning

confidence: 99%

Bioinformatic Methods for Identifying Differentially Abundant Subpopulations in scRNA-seq Data

Kleinberg¹,

Shinde²,

Batchu³

et al. 2022

Preprint

View full text Add to dashboard Cite

Single-cell RNA sequencing data facilitates investigation of cell heterogeneity and subpopulations as well as differentially abundant states however modern single-cell RNA sequencing datasets are growing in size and complexity requiring advances in the bioinformatic methods that analyze them. Many methods exist for each step of analysis including read alignment, normalization, quality control, batch effect correction, imputation and dimensionality reduction. With so many options to choose from at each step of the analysis, benchmarking and a synthesis of the literature on the methods available is necessary to inform biological researchers on the most optimal workflow for their data. Here, recent key methods of analysis are highlighted with a focus on methods that facilitate identification of cell subpopulations and differentially abundant cell states. With a constantly expanding toolset for each step in single-cell RNA sequencing dataset analysis, biological researchers should stay informed to utilize the most applicable methods for their own analyses.

show abstract

“…To emphasize the general and important consequences of this form of EN for a diverse range of practical applications, we consider generic ensembles of random matrices with fixed margins. These ensembles, which include matrices with 0/1 (or equivalently ±1) and non-negative integer entries subject to global or local constraints, arise for instance in studies of multi-cell gene expression profiles [17], multiplex (online) social activity [18], multi-channel communication systems [19], complex networks [20], and multivariate time series in finance [16], neuroscience [21] or other disciplines. Our results imply that, in many practical situations, the assumption of EE is incorrect and leads to mathematically wrong conclusions.…”

Section: Introductionmentioning

confidence: 99%

Strong ensemble nonequivalence in systems with local constraints

Zhang

Garlaschelli

2022

New J. Phys.

Self Cite

View full text Add to dashboard Cite

The asymptotic equivalence of canonical and microcanonical ensembles is a central concept in statistical physics, with important consequences for both theoretical research and practical applications. However, this property breaks down under certain circumstances. The most studied violation of ensemble equivalence requires phase transitions, in which case it has a ‘restricted’ (i.e. confined to a certain region in parameter space) but ‘strong’ (i.e. characterized by a difference between the entropies of the two ensembles that is of the same order as the entropies themselves) form. However, recent research on networks has shown that the presence of an extensive number of local constraints can lead to ensemble nonequivalence even in the absence of phase transitions. This occurs in a ‘weak’(i.e. leading to a subleading entropy difference) but remarkably ‘unrestricted’ (i.e. valid in the entire parameter space) form. Here we look for more general manifestations of ensemble nonequivalence in arbitrary ensembles of matrices with given margins. These models have widespread applications in the study of spatially heterogeneous and/or temporally nonstationary systems, with consequences for the analysis of multivariate financial and neural time-series, multi-platform social activity, gene expression profiles and other Big Data. We confirm that ensemble nonequivalence appears in ‘unrestricted’ form throughout the entire parameter space due to the extensivity of local constraints. Surprisingly, at the same time it can also exhibit the ‘strong’ form.This novel, simultaneously‘strong and unrestricted’ form of nonequivalence is very robust and imposes a principled choice ofthe ensemble. We calculate the proper mathematical quantities to be used in real-world applications.

show abstract

Phiclust: a clusterability measure for single-cell transcriptomics reveals phenotypic subpopulations

Cited by 10 publications

References 65 publications

ClusterDE: a post-clustering differential expression (DE) method robust to false-positive inflation caused by double dipping

ClusterDE: a post-clustering differential expression (DE) method robust to false-positive inflation caused by double dipping

Bioinformatic Methods for Identifying Differentially Abundant Subpopulations in scRNA-seq Data

Strong ensemble nonequivalence in systems with local constraints

Contact Info

Product

Resources

About