Manifold learning-based methods for analyzing single-cell RNA-sequencing data

Moon, Kevin R.; Stanley, Jay S.; Burkhardt, Daniel; Dijk, David van; Wolf, Guy; Krishnaswamy, Smita

doi:10.1016/j.coisb.2017.12.008

Cited by 120 publications

(126 citation statements)

References 77 publications

Supporting

Mentioning

122

Contrasting

Order By: Relevance

“…Data with such properties can thus be modeled geometrically by a collection of smoothly varying data patches defined by local neighborhoods [62]. This collection essentially fits the manifold learning paradigm, which relies on a mathematical manifold model for the geometry of a progression track, together with analysis tools for characterizing it.…”

Section: Visualization: To Enable Visualization Phate Captures Variamentioning

confidence: 99%

Visualizing Structure and Transitions for Biological Data Exploration

Dijk

Wang

et al. 2017

Preprint

Self Cite

109

View full text Add to dashboard Cite

In the era of 'Big Data' there is a pressing need for tools that provide human interpretable visualizations of emergent patterns in high-throughput high-dimensional data. Further, to enable insightful data exploration, such visualizations should faithfully capture and emphasize emergent structures and patterns without enforcing prior assumptions on the shape or form of the data. In this paper, we present PHATE (Potential of Heat-diffusion for Affinity-based Transition Embedding) -an unsupervised low-dimensional embedding for visualization of data that is aimed at solving these issues. Unlike previous methods that are commonly used for visualization, such as PCA and tSNE, PHATE is able to capture and highlight both local and global structure in the data. In particular, in addition to clustering patterns, PHATE also uncovers and emphasizes progression and transitions (when they exist) in the data, which are often missed in other visualization-capable methods. Such 24, 2017; patterns are especially important in biological data that contain, for example, single-cell phenotypes at different phases of differentiation, patients at different stages of disease progression, and gut microbial compositions that vary gradually between individuals, even of the same enterotype.International license peer-reviewed) is the author/funder. It is made available under a The copyright holder for this preprint (which was not . http://dx.doi.org/10.1101/120378 doi: bioRxiv preprint first posted online Mar.The embedding provided by PHATE is based on a novel informational distance that captures long-range nonlinear relations in the data by computing energy potentials of dataadaptive diffusion processes. We demonstrate the effectiveness of the produced visualization in revealing insights on a wide variety of biomedical data, including single-cell RNA-sequencing, mass cytometry, gut microbiome sequencing, human SNP data, Hi-C data, as well as non-biomedical data, such as facebook network and facial image data. In order to validate the capability of PHATE to enable exploratory analysis, we generate a new dataset of 31,000 single-cells from a human embryoid body differentiation system. Here, PHATE provides a comprehensive picture of the differentiation process, while visualizing major and minor branching trajectories in the data. We validate that all known cell types are recapitulated in the PHATE embedding in proper organization. Furthermore, the global picture of the system offered by PHATE allows us to connect parts of the developmental progression and characterize novel regulators associated with developmental lineages.

show abstract

Section: Visualization: To Enable Visualization Phate Captures Variamentioning

confidence: 99%

Visualizing Structure and Transitions for Biological Data Exploration

Dijk

Wang

et al. 2017

Preprint

Self Cite

109

View full text Add to dashboard Cite

show abstract

“…The cell-state embedding identified a total of 11 cell subtypes with gene expression (Figure 5b, Supplementary Table 6). While it is well-understood that a set of individual cells, such as those undergoing differentiation, may demonstrate manifold structure [22,23], our PhEMD embedding suggested that a set of tumors from different patients with a shared phenotype (e.g., melanoma) may also lie on a continuous manifold.…”

Section: Phemd Highlights Manifold Structure Of Tumor Samples In Cytomentioning

confidence: 73%

“…C-6 and C-7 had roughly the opposite expression profile with respect to the markers described above (Figure 2c). E-cadherin is the hallmark cell adhesion marker of epithelial cells [6], and vimentin and CD44 are known mesenchymal markers involved in cell migration [6][7][8][9]. Moreover, recent studies found high CD44:CD24 expression to be indicative of breast cancer cell invasiveness and an as an EMT endpoint, suggestive of mesenchymal properties [10][11][12].…”

Section: Effect Of Drug Perturbations On the Emt Landscape In Breast mentioning

confidence: 99%

“…The analysis of high-dimensional data, especially in unsupervised or exploratory settings, often introduces various challenges that are collectively referred to as the curse of dimensionality [6,7]. A popular approach to analyzing such data is to use manifold learning methods that assume the intrinsic geometry of the data can conceptually be modeled as a low dimensional manifold (i.e., a collection of smoothly varying locally low dimensional data patches), which is immersed in the high dimensional ambient space of collected features [8]. Such methods often aim to uncover this intrinsic geometry by first capturing local neighborhoods, then using them to form a rigid structure of nonlinear relations in the data, and finally embedding this structure in low dimensions via a new set of features that preserve those relations (e.g., as distances).…”

Section: Supplementary Note 2: Motivation For "Cell Subtype" Featuresmentioning

confidence: 99%

See 1 more Smart Citation

Embedding single-cell experimental conditions to reveal manifold structure of cancer drug perturbation effects

Chen

Zivanovic

Dijk

et al. 2018

Preprint

Self Cite

View full text Add to dashboard Cite

Previously, the effect of a drug on a cell population was measured based on simple metrics such as cell viability. However, as single-cell technologies are becoming more advanced, drug screen experiments can now be conducted with more complex readouts such as gene expression profiles of individual cells. The increasing complexity of measurements from these multi-sample experiments calls for more sophisticated analytical approaches than are currently available. We develop a novel method called PhEMD (Phenotypic Earth Mover's Distance) and show that it can be used to embed the space of drug perturbations on the basis of the drugs' effects on cell populations. When testing PhEMD on a newly-generated, 300-sample CyTOF kinase inhibition screen experiment, we find that the state space of the perturbation conditions is surprisingly low-dimensional and that the network of drugs demonstrates manifold structure. We show that because of the fairly simple manifold geometry of the 300 samples, we can accurately capture the full range of drug effects using a dictionary of only 30 experimental conditions. We also show that new drugs can be added to our PhEMD embedding using similarities inferred from other characterizations of drugs using a technique called Nystrom extension. Our findings suggest that large-scale drug screens can be conducted by measuring only a small fraction of the drugs using the most expensive high-throughput single-cell technologies-the effects of other drugs may be inferred by mapping and extending the perturbation space. We additionally show that PhEMD can be useful for analyzing other types of single-cell samples, such as patient tumor biopsies, by mapping the patient state space in a similar way as the drug state space. We demonstrate that PhEMD is highly scalable, compatible with leading batch effect correction techniques, and generalizable to multiple experimental designs. Altogether, our analyses suggest that PhEMD may facilitate drug discovery efforts and help uncover the network geometry of a collection of single-cell samples.

show abstract

“…Furthermore, most of these 13 features had only a very small contribution to the overall explanatory power suggesting that cell state distribution can be represented by few latent dimensions. An observation that emboldens efforts to learn the cell state manifold (Moon et al, 2018) . 2.…”

Section: Discussionmentioning

confidence: 99%

Mammalian gene expression variability is explained by underlying cell state

Foreman

Wollman

2019

Preprint

View full text Add to dashboard Cite

Gene expression variability in mammalian systems plays an important role in physiological and pathophysiological conditions. This variability can come from differential regulation related to cell state (extrinsic) and allele-specific transcriptional bursting (intrinsic). Yet, the relative contribution of these two distinct sources is unknown. Here we exploit the qualitative difference in the patterns of covariance between these two sources to quantify their relative contributions to expression variance in mammalian cells. Using multiplexed error robust RNA fluorescent in situ hybridization (MERFISH) we measured the multivariate gene expression distribution of 150 genes related to Ca 2+ signaling coupled with the dynamic Ca 2+ response of live cells to ATP. We show that after controlling for cellular phenotypic states such as size, cell cycle stage, and Ca 2+ response to ATP, the remaining variability is effectively at the Poisson limit for most genes. These findings demonstrate that the majority of expression variability results from cell state differences and that the contribution of transcriptional bursting is relatively minimal.

show abstract

Manifold learning-based methods for analyzing single-cell RNA-sequencing data

Cited by 120 publications

References 77 publications

Visualizing Structure and Transitions for Biological Data Exploration

Visualizing Structure and Transitions for Biological Data Exploration

Embedding single-cell experimental conditions to reveal manifold structure of cancer drug perturbation effects

Mammalian gene expression variability is explained by underlying cell state

Contact Info

Product

Resources

About