Forum:#saveNCBI - it's a joke, right?

Question

[5heikki]

This happened yesterday:

Surely this is not going to happen, right?

[Adamc]

Although some of the roles that NCBI serves are effectively "duplicated" through other similar resources such as ENSEMBL, as a one-stop-shop well-curated (although sometimes chaotically organized) information that's essential to research in nearly any known organism, NCBI is most certainly critical. Any efforts to "save money" by scaling it down would pretty much set us back 20 years. I certainly hope no one in a position of power would be fool enough to put such a plan into action.

NCBI is not like NASA- bioscience research is critical for human health in the short term, and open sharing of such resources as provided by NCBI are cornerstones of that research.

Preaching to the choir here though.

[John]

Until we know exactly what is being proposed, it's difficult to pass any judgement... I mean, as useful as the NCBI is, I don't know if I've ever met anyone who has said the way the data/departments are laid out is... optimal. Don't get me wrong, there are plenty of examples of business people "not getting it", particularly when it's some business man who did a great job at Manufacturing Company X and is now heading up some Scientific initiative thinking Engineering == Technology == Science -- but we should at least wait and see dontcha think :)

[Devon Ryan]

I have yet to see any source for this and every time someone asks Michael Eisen then don't get a reply.

[5heikki]

In the comments he writes "have heard from three separate people at NIH"

[Devon Ryan]

Missed that one, thanks! I'll wait to see if other people can independently mention hearing the same things from folks before I get overly concerned.

Answer 1

[Istvan Albert]

I for one believe that NCBI should get out of the business of providing user interfaces. Right now they are in the business of both: data storage and interface building with that monopolize each and suffocate independent efforts to improve on either.

They should provide fast and efficient APIs to their data and let the community build tools around those.

Answer 2

[GenoMax]

This has already happened to TAIR (Arabidopsis) database which now requires a subscription for full access. Since it is a niche database many have not felt the pain but a few labs that I know of certainly have.

[Istvan Albert]

I got curious and looked it up. The Academic Individual License comes to $10 per month per person. That is extremely cheap, practically a non-issue.

[5heikki]

KEGG is $2,000 for single academic end user or $5,000 for institutional license. Non-profit licenses are of course much more costly..

[Chris Fields]

Keep in mind that universities typically budget huge amounts for journal subscriptions and web-based resources. If database resources were treated similarly (as perpetually reviewed/curated resources of scientific data, available via a web interface) then there is an obvious model in place for them to tap into.

Personally I have no problem suggesting our library pay for a subscription to TAIR, KEGG, etc. The individual cost may be high, but it's minuscule in comparison to what big publishers like Elsevier make.

[GenoMax]

You must be a generous PI for people in your group :-) Not every PI is so inclined. But just like netflix/spotify etc. these subscriptions will add up.

Going to a subscription based model would invariably mean that the PI's would charge these costs to their grants so ultimately the federal government/foundations would be paying for these. With a case like NCBI (if it ever comes to that) it would be easy to sell the institutions on acquiring a site license (since it would be the cost effective way).

[John]

I would pay for NCBI. Much better than the system that negates my purchasing power by taking the money out of my grant before I even get it.

[Istvan Albert]

Exactly, the actual costs of running these services are hidden and very likely extraordinary high. There is a pure monopoly in place where competition is not actually possible. We can't make a better NCBI if no one can compete with NCBI.

[GenoMax]

NCBI was conceived during a very different era when when entities were actively seeking to patent sequences. Even though it is not explicitly noted in NCBI's mission statement the "intent" was to keep the information open and freely available for all.

[Istvan Albert]

Well, it is not about generosity, a graduate student's cumulative costs will be around $50K per year. $100 is a rounding error on that scale.

I think the only hope for these services to actually become useful and user oriented is to compete on usability (as in providing what users need) rather than compete on lofty (exagerated) claims used to convince a tiny subset of reviewers.

[5heikki]

Something similar happened to KEGG, i.e. public funding ended and they had to find new ways to support the database.