Single cell classification via cell-type hierarchies based on ensemble learning and sample size estimation.
Install Bioconductor packages S4Vectors, hopach and limma packages using BiocManager:
# install.packages("BiocManager")
BiocManager::install(c("S4Vectors", "hopach", "limma"))Then install the latest scClassify using devtools (For R >= 4.0):
library(devtools)
devtools::install_github("SydneyBioX/scClassify")For R >= 3.6, install scClassify(v0.2.3) via
devtools::install_github("SydneyBioX/scClassify@085c72f")
You can find the vignette at this website (https://sydneybiox.github.io/scClassify/index.html):
- scClassify Model Building and Prediction: https://sydneybiox.github.io/scClassify/articles/scClassify.html
- Sample size calculation: https://sydneybiox.github.io/scClassify/articles/webOnly/sampleSizeCal.html
- Performing scClassify using pretrained models: https://sydneybiox.github.io/scClassify/articles/pretrainedModel.html
Also, you can find our interactive shiny application (beta) at this website: http://shiny.maths.usyd.edu.au/scClassify.
Currently available pre-trained scClassify models (in scClassifyTrainModel class)
| Tissue | Organism | Training Data | Accession | Summary | Download .rds |
Gene Name Format |
|---|---|---|---|---|---|---|
| Primary visual cortex | mouse | Tasic (2018) | GSE115746 | link | link | Mm Gene Symbol |
| Primary visual cortex | mouse | Tasic (2016) | GSE71585 | link | link | Mm Gene Symbol |
| Visual cortex | mouse | Hrvatin | GSE102827 | link | link | Mm Gene Symbol |
| Lung | mouse | Cohen | GSE119228 | link | link | Mm Gene Symbol |
| Kidney | mouse | Park | GSE107585 | link | link | Mm Gene Symbol |
| Liver | human | MacParland | GSE115469 | link | link | Hs Gene Symbol |
| Liver | human | Aizarani | GSE124395 | link | link | Hs Gene Symbol |
| Pancreas | human | Xin | GSE81608 | link | link | Hs Gene Symbol |
| Pancreas | human | Wang | GSE83139 | link | link | Hs Gene Symbol |
| Pancreas | human | Lawlor | GSE86469 | link | link | Hs Gene Symbol |
| Pancreas | human | Segerstolpe | E-MTAB-5061 | link | link | Hs Gene Symbol |
| Pancreas | human | Muraro | GSE85241 | link | link | Hs Gene Symbol |
| Pancreas | human | Baron | GSE84133 | link | link | Hs Gene Symbol |
| Pancreas | human | joint | - | link | link | Hs Gene Symbol |
| Melanoma | human | Li | GSE123139 | link | link | Hs Gene Symbol |
| PBMC | human | Ding (joint) | - | link | link | Mm EMSEMBL ID |
| Tabula Muris | mouse | Tabula Muris | GSE109774 | link | link | Mm Gene Symbol |
If you have any enquiries, especially about performing scClassify to classify your cells or to build your own models, please contact yingxin.lin@sydney.edu.au or bioinformatics@maths.usyd.edu.au.
scClassify: sample size estimation and multiscale classification of cells using single and multiple reference
Yingxin Lin, Yue Cao, Hani Kim, Agus Salim, Terence Speed, David Lin, Pengyi Yang† & Jean Yang†. Molecular Systems Biology, 2020, 16, e9389. Full Text; BioC R package