Entering edit mode
7.8 years ago
cdsouthan
★
1.9k
I just warned (in friendly way of course) a helpful commenter who recommended quite a high profile protein database that had not been updated since 2010 (or if it had they still left this date on the download site)
We all know this is a chronic problem, especially for the less experienced we try to help here
As we also know its a sustainable funding issue in most cases
I actually don't think much can be done :( e.g. a name and shame list would be too disingenuous?
but does anyone have any better ideas?
Private philanthropists who support biomedical research could help with these infrastructure resources. Unfortunately this is not "risky/cutting edge science" so would not help generate any publicity.
Could you contact the 'high profile protein db' and find out what has happened? Is it really updated up to 2010? Or is the data wrong? Imho, I'd not name and shame the dbs, rather continue warning people who comment or reply of the pitfalls of relying on out-of-date data...
I think cdsouthan was not interested in the database mentioned directly. cdsouthan was only referring to an answer someone else had provided.
While it is not clear from the website of the said database the site may only be for a static snapshot of the associated publication. The underlying database itself may have changed to a commercial/subscription only model (though there is no clear indication if that is the case or the database has gone dormant).
I did understand the the OP was not interested in the 'high profile protein db'. It was a mere example. However I still think it is worth contacting the aforementioned db to find out whether the data has been updated, but the website has not; or the data was updated in 2010 and no sooner than that. My 2 cents.
I take the point and I do sometimes contact databases re updates that are important to us (e.g. for reciprocal connectivity as described https://blog.guidetopharmacology.org/2016/03/30/collation-and-assessment-of-gtopdb-in-links/)
Notwithstanding;
a) the db in question was not in this category
b) neither I, nor anyone else I'm guessing, is prepared to be an update policeman (life being not only too short but it also diminishes the chances of being brought drinks at conferences)
This having been said I'm wondering how far web crawling could reliably go as an automated way of harvesting dates from, say, front pages, stats lists and ftp sites?