Microarray similarity estimation

Question

[sosal]

Hello, I am cell biologist.

I succeeded in differentiating stem cells.

I want to test whether the cells are differentiated correctly or not by using microarray. (There is gold standard cell line.)

And now, I have three groups of microarray data.

Group1: Before differentiation Group2: After differentiation Group3: Gold standard cell line.

Is there a good way to evaluate whether cells are differentiated correctly (gold standard) ? (MDS plot shows that group2 and group3 are similar.)

thank you.

Answer 1

[Jean-Karim Heriche]

If you've done multidimensional scaling (MDS) then you already have a measure of similarity between your groups. If you need to quantify the similarity, either use the similarity/distance measure you used as input to MDS or use the distance in the MDS projection.