search

SofieVG / FlowSOM

Using self-organizing maps for visualization and interpretation of cytometry data
61 stars 26 forks source link

Can FlowSOM be used for scRNA-seq data? #32

Closed EDYAC closed 3 years ago

EDYAC commented 4 years ago

Hello everyone, I use FlowSOM in cytobank for CyTOF data, I was wondering if anyone have applied FlowSOM to scRNA-seq and compared it with other clustering methods?

Thank you!

SamGG commented 3 years ago

Hi, I don't see any reason against. SOM is a clustering technique, was used in transcriptome analysis with micro-arrays... I think it is available as clustering in SeqGeq (FlowJo). Best.

SofieVG commented 3 years ago

Hi,

One relevant adaptation is to include a PCA step upfront. In flow data typically the markers in the panel are chosen in such a way that they each contain relevant new information, while in RNA-seq you typically have many genes telling a very similar story, which would skew the clustering. With the Self-Organizing Map it is also important to make sure you first overcluster the data (potentially followed by a second metaclustering to come closer to the expected number of clusters, this is similar for flow data).

The Robinson lab did a benchmark study: https://f1000research.com/articles/7-1141/v3

All the best, Sofie