The 1000genomes.no database, containing the frequency of all germline variants in our Norwegian cancer patient cohort has been approved to be anonymous by the National Committee for Medical Research Ethics, but in the effort now funded by the Ministry of health to further develop this into an sustained national resource, the Directorate of Health has decided to follow the practise of Statistics Norway (The National Bureau of Statistics) to define any property present in less than 5 individuals in a cohort (i.e. not in the population) as non-anonymous. It is possible that this makes sense in their context, but for population genetics this appears to be ill-founded.

By this definition we have to remove the 900 000 least frequent variants form our 1000 patient cohort, which of course are the most valuable information the database provides.  Most frequent variants will already be present in other variant databases, in particular because a large fraction of Norwegians emigrated to the Americas, whereas we have shown that many infrequent variants are absent in international variant databases. Furthermore, a variant presnt 4 times in our cohort would be expected to be present in 20 000 Norwegians, thus in our mind does not identify a small group of individuals. Hopefully, this may be corrected.