Aaron’s single-cell thoughts

Overview

A collection of thoughts related to single-cell data analysis, mostly for RNA sequencing.

General

Broad thoughts about practices in single-cell data analysis, not specific to my own software.

Comments:

• Quality control
• Linearity
• Transformation
• Standardization
• Clustering
• Cellular detection rate
• Bootstrapping
• Visualization

Links:

• Effects of log-transformation at https://github.com/LTLA/PseudoCount2018
• Other imputation pitfalls at https://github.com/LTLA/Counterspell2018

Software

Justifications for some of the programming choices used in my own scRNA-seq data analysis software.

Comments:

• Correlations (scran)
• Marker detection (scran)
• Doublet detection by cluster or per cell (scran)

Links:

• Scaling normalization in https://github.com/MarioniLab/FurtherNorm2018
• HVG selection in https://github.com/LTLA/HVGDetection2018
• PC selection in https://github.com/LTLA/PCSelection2018
• MNN correction in https://github.com/MarioniLab/FurtherMNN2018

Workflow

Justifications for some of the parameter choices and strategies in my scRNA-seq analysis workflows.

Comments:

• Quality control (simpleSingleCell)
• Cell cycle detection (simpleSingleCell)

Other

Some collected miscellaneous thoughts:

• Research integrity