In SydneyBioX/scMerge: scMerge: Merging multiple batches of scRNA-seq data

Introduction

This is a mouse embryonic stem cell (mESC) single-cell data comprising of 3 different batches from a single experiment.

Integration challenge

Prior to integration, there is a strong separation effect by batches.
This data is relatively simple with only the usual batch effects within an experiment. Hence, in the scMerge article it was used as an illustrative example for various aspects of the methodology and also included as an example dataset in our package.

Data description

Data source:

| Type of merge | | Name | | ID | Author | DOI or URL | Protocol | Organism | Tissue | # of cell types | # of cells | # of batches | |-------------------|---|------|---|-------------|---------------|----------------------------|------------|----------|-----------|-----------------|------------|--------------| | Within experiment | | mESC | | E-MTAB-2600 | Kolodziejczyk | 10.1016/j.stem.2015.09.011 | SMARTer/C1 | Mouse | Mouse ESC | 3 | 704 | 5 |

Relation to the scMerge article: Main Figure 1, Supplementary Figure 6 and 17.

Data visualisation

tSNE plots by cell types and batch

{width=100%}

Illustrating the robustness of the parameters $k_{RUVIII}$ and $k_{Cluster}$

{width=100%}

Integrated `scMerge` data

Data availability: Unmerged data (in RData format) and mouse ESC Data (in RData format)
scMerge parameters for integration:
- Unsupervised scMerge
- kmeans K = (1,3,3,1,1)
- Negative controls are mouse scSEG