RECORD
Probabilistic Count Matrix Factorization for Single Cell Expression Data Analysis
Methodology · Submitted 30 Oct 2017
abstract compiled · 2026-09-23 · no language-model review
The authors' main claim We show how this probabilistic framework induces a geometry that is suitable for single-cell data visualization, and produces a compression of the data that is very powerful for clustering purposes.
The compiled record
The paper's own sentences, arranged. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
Open the compiled record
The abstract, sentence by sentence
The development of high throughput single-cell sequencing technologies now allows the investigation of the population level diversity of cellular transcriptomes. This diversity has shown two faces. First, the expression dynamics (gene to gene variability) can be quantified more accurately, thanks to the measurement of lowly-expressed genes. Second, the cell-to-cell variability is high, with a low proportion of cells expressing the same gene at the same time/level. Those emerging patterns appear to be very challenging from the statistical point of view, especially to represent and to provide a summarized view of single-cell expression data. PCA is one of the most powerful framework to provide a suitable representation of high dimensional datasets, by searching for latent directions catching the most variability in the data. Unfortunately, classical PCA is based on Euclidean distances and projections that work poorly in presence of over-dispersed counts that show drop-out events (zero-inflation) like single-cell expression data We propose a probabilistic Count Matrix Factorization (pCMF) approach for single-cell expression data analysis, that relies on a sparse Gamma-Poisson factor model. This hierarchical model is inferred using a variational EM algorithm. It is able to jointly build a low dimensional representation of cells and genes. We show how this probabilistic framework induces a geometry that is suitable for single-cell data visualization, and produces a compression of the data that is very powerful for clustering purposes. Our method is competed against other standard representation methods like t-SNE, and we illustrate its performance for the representation of single-cell data. We especially focus on publicly available single-cell RNA-seq datasets.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | The development of high throughput single-cell sequencing technologies now allows the investigation of the population level diversity of cellular transcriptomes. | background 1.00 | 0.05 | 0.02 | asserted 0.98 |
| s1 | This diversity has shown two faces. | background 1.00 | 0.09 | 0.02 | asserted 0.95 |
| s2 | First, the expression dynamics (gene to gene variability) can be quantified more accurately, thanks to the measurement of lowly-expressed genes. | background 0.98 | 0.11 | 0.03 | asserted 0.97 |
| s3 | Second, the cell-to-cell variability is high, with a low proportion of cells expressing the same gene at the same time/level. | background 0.98 | 0.15 | 0.03 | asserted 1.00 |
| s4 | Those emerging patterns appear to be very challenging from the statistical point of view, especially to represent and to provide a summarized view of single-cell expression data. | background 0.74 | 0.09 | 0.03 | hedged possibility 0.64 |
| s5 | PCA is one of the most powerful framework to provide a suitable representation of high dimensional datasets, by searching for latent directions catching the most variability in the data. | background 1.00 | 0.04 | 0.03 | asserted 1.00 |
| s6 | Unfortunately, classical PCA is based on Euclidean distances and projections that work poorly in presence of over-dispersed counts that show drop-out events (zero-inflation) like single-cell expression data We propose a probabilistic Count Matrix Factorization (pCMF) approach for single-cell expression data analysis, that relies on a sparse Gamma-Poisson factor model. | contribution 0.87 | 0.51 | 0.11 | asserted 1.00 |
| s7 | This hierarchical model is inferred using a variational EM algorithm. | method 1.00 | 0.79 | 0.03 | asserted 1.00 |
| s8 | It is able to jointly build a low dimensional representation of cells and genes. | method 0.66 | 0.82 | 0.11 | asserted 1.00 |
| s9 | We show how this probabilistic framework induces a geometry that is suitable for single-cell data visualization, and produces a compression of the data that is very powerful for clustering purposes. | result 0.85 | 0.92 | 0.49 | asserted 1.00 |
| s10 | Our method is competed against other standard representation methods like t-SNE, and we illustrate its performance for the representation of single-cell data. | contribution 0.44 | 0.89 | 0.09 | asserted 0.94 |
| s11 | We especially focus on publicly available single-cell RNA-seq datasets. | method 0.70 | 0.79 | 0.03 | asserted 0.90 |
What was examined, and what was not
- arXiv listing abstract v5
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
The paper
Original abstract
The development of high throughput single-cell sequencing technologies now allows the investigation of the population level diversity of cellular transcriptomes. This diversity has shown two faces. First, the expression dynamics (gene to gene variability) can be quantified more accurately, thanks to the measurement of lowly-expressed genes. Second, the cell-to-cell variability is high, with a low proportion of cells expressing the same gene at the same time/level. Those emerging patterns appear to be very challenging from the statistical point of view, especially to represent and to provide a summarized view of single-cell expression data. PCA is one of the most powerful framework to provide a suitable representation of high dimensional datasets, by searching for latent directions catching the most variability in the data. Unfortunately, classical PCA is based on Euclidean distances and projections that work poorly in presence of over-dispersed counts that show drop-out events (zero-inflation) like single-cell expression data We propose a probabilistic Count Matrix Factorization (pCMF) approach for single-cell expression data analysis, that relies on a sparse Gamma-Poisson factor model. This hierarchical model is inferred using a variational EM algorithm. It is able to jointly build a low dimensional representation of cells and genes. We show how this probabilistic framework induces a geometry that is suitable for single-cell data visualization, and produces a compression of the data that is very powerful for clustering purposes. Our method is competed against other standard representation methods like t-SNE, and we illustrate its performance for the representation of single-cell data. We especially focus on publicly available single-cell RNA-seq datasets.
Human review
How the literature uses it
What this paper does with the arXiv papers it cites
Discussion
Know something about this paper? Leave a comment, or highlight any passage above to comment on it.