Yeah I agree that 23andme name recognition and brand is valuable and maybe best in class for engagement of patients with genomic data. So for clinical trial recruitment and marketing that would help.
I'm less convinced of the value of 23 and me data in drug discovery. It could be quite useful but is not best in class. The regeneron genetics center is a much better approach imo -- sequencing data vs genotyping, access to robust medical data, partnerships with leading medical institutions to get patients and data, and a targeted biology driven strategy that reduces the cost of getting a signal. As far as cost per good potential target, 23and me is probably an order of magnitude or more behind
Not sure how 23 and me would help GSK analyze genetic subpopulations -- GSK could sequence patients just fine on their own without 23and me and get more robust data than 23 and me genotype data. For biomarker discovery once you have a candidate and are moving into clinical studies there are probably much better targeted solutions than 23 andme
Value of 23 and me is 1) brand and 2) data on large number of pts. 2) helps with finding genetic markers with low effect size which isn't really helpful for drug discovery and also finding rare variants with large effect, which is important, but there are better / cheaper ways of doing that (i.e. regeneron)