Significance of a cluster of primary and metastatic cell line

To build metabolic models using transcriptomics expression data of a cancer type, is it advisable to choose the primary and metastatic cell line belonging to the single cluster?

We don’t have experience building such models and can’t comment on the best approaches for doing that.


Ok. Thanks.
Can you throw some light on the basis of clusters? What exactly does it mean if a group of cell lines is assigned a single cluster?

The clusters you’re referring to are those in the cellligner analysis, correct?

If that’s the case, I believe that they’re based on the distances in the N-dimensional space that the cell lines and tumors were co-embedded within. The embedding is derived from expression data, so I’d say an assignment to a cluster is generally telling you lines which are more transcriptionally similar to each other than the other lines that are not part of the cluster.

I don’t recall the specifics of which clustering method they used, and I could be mistaken about the fact the clustering being done in the lower dimensional embedding space.

Celligner is not my speciality, so I should refer you to the celligner paper for the most accurate information.


Thank you. I will check it out