https://github.com/kieranrcampbell/ouija

Descriptive probabilistic marker gene approach to single-cell pseudotime inference
https://github.com/kieranrcampbell/ouija

bayesian-methods factor-analysis gene-expression single-cell

Last synced: 24 days ago
JSON representation

Descriptive probabilistic marker gene approach to single-cell pseudotime inference

• Host: GitHub
• URL: https://github.com/kieranrcampbell/ouija
• Owner: kieranrcampbell
• Created: 2016-03-30T12:44:36.000Z (over 8 years ago)
• Default Branch: master
• Last Pushed: 2019-11-29T16:11:10.000Z (over 4 years ago)
• Last Synced: 2024-02-24T13:33:56.985Z (4 months ago)
• Topics: bayesian-methods, factor-analysis, gene-expression, single-cell
• Language: R
• Homepage: http://kieranrcampbell.github.io/ouija
• Size: 9.05 MB
• Stars: 28
• Watchers: 6
• Forks: 3
• Open Issues: 9
• Metadata Files:
  • Readme: README.md

Lists

• awesome_single_cell - ouija - [R] - [A descriptive marker gene approach to single-cell pseudotime inference](https://doi.org/10.1101/060442) (Software packages / RNA-seq)
• awesome-single-cell - ouija - [R] - [A descriptive marker gene approach to single-cell pseudotime inference](https://doi.org/10.1101/060442) (Software packages / Pseudotime and trajectory inference)
• awesome-single-cell - ouija - [R] - [A descriptive marker gene approach to single-cell pseudotime inference](https://doi.org/10.1101/060442) (Software packages / RNA-seq)

README

        
# Ouija

Ouija is a probabilistic pseudotime framework. Ouija 

* infers pseudotimes from a **small number of marker genes** letting you understand **why** the pseudotimes have been learned in terms of those genes (**A**)

* provides parameter estimates (with uncertainty) for **interpretable gene regulation behaviour** (such as the peak time or the upregulation time) (**B**)

* has a Bayesian hypothesis test to **find genes regulated before others** along the trajectory (**C**)

* identifies **metastable states**, ie discrete cell types along the continuous trajectory (**D**)



## Getting started

### Installation

```r

# install.packages("devtools")

devtools::install_github("kieranrcampbell/ouija")

```

To build the Ouija vignette install using

```r

devtools::install_github("kieranrcampbell/ouija", local = FALSE, 

                          args = "--preclean", build_vignettes = TRUE)

```

### Model fitting

Input is a cell-by-gene expression matrices that is non-negative and represents logged gene expression values. We recommend using `log2(TPM + 1)`. This can either take the form of a matrix or a [`SingleCellExperiment`](https://bioconductor.org/packages/release/bioc/html/SingleCellExperiment.html) (use of the `SingleCellExperiment` infrastructure is highly encouraged for single-cell analyses). By default the `logcounts` assay of a `SingleCellExperiment` will be used.

To fit the pseudotimes, pass the input data to the `ouija` function:

```r

library(ouija)

data(example_gex) # synthetic gene expression data bundled

oui <- ouija(example_gex)

pseudotimes <- map_pseudotime(oui)

```

The `map_pseudotimes` function extracts the maximum-a-posteriori (MAP) estimates of the pseudotimes.

For further usage options see the vignette. A prebuilt vignette can be found [here](http://kieranrcampbell.github.io/ouija).

## Authors

Kieran Campbell & Christopher Yau  

Wellcome Trust Centre for Human Genetics, University of Oxford

## Artwork



Artwork by `cwcyau`, the mysterious banksy-esque artist of the statistical genomics world.