In fact, in searching databases like Genbank, I find this one of the more frustrating aspects, finding a sequence or data and then spending an inordinate amount of time confirming that sequence (or not doing it and getting misleading data). I can see this could be very helpful.

I do see also the issue of moderation, even among scientific research communities, to insure the quality of the attached data.

It will be interesting to see how this develops over time.

If someone does an analysis (like builds a tree) on existing that shows that a) the species/genus is wrong b) the “uncultured soil clone” is actually a known species then it should be possible to attach this information to a record so that the next time someone does a BLAST search, etc they have a way of building on that knowledge. The Letter to Science was primarily focused on attaching proper or more detailed species identification while Steven’s opinion more on fixing gene structure problems in genbank (which is also a huge problem that also needs help).

The problem right now is there is no way to update the actual record (or attach a pointer to a corrected one) except if the original author submits the request OR they have turned that permission over to a proxy like all the yeast chromosome sequences have been been proxied to SGD to make the updates.