In article <CvJ1B6.Hxq at ncifcrf.gov> toms at fcsparc6.ncifcrf.gov (Tom Schneider) writes:
>In article <RSMITH.94Aug31204529 at dot.bcm.tmc.edu> rsmith at dot.bcm.tmc.edu>(Randall Smith) writes:
>>| : Announcing a new WWW Server: The BCM Sequence Annotation Server
>| : URL: http://dot.imgen.bcm.tmc.edu:9331/seq-annot/home.html>>This idea has both good and bad points and I think that they should be
>considered extremely carefully.
>>First, the idea that individual researchers should be able to update a database
>directly is wonderful. If everybody were to do that to the data they are
>experts on, GenBank would quickly become a clean database.
>>However, there are several problems:
>>1. Corruption of the data either intentionally or inadvertently. The authors
>already are aware of this possibility.
>>2. Inconsistency between researchers. The NCBI attempts to make entries
>uniform by certain standards, but individual researchers cannot know what these
>are in detail. The only solutions I see are to force the kinds of annotations
>to follow certain forms or to pass all the changes to experts for editing. The
>trouble with passing the data to experts is that the number of experts has to
>grow exponentially along with the growth of the database. Maybe we just have
>to accept that to avoid a worse mess!
>>2. Do the data flow into GenBank? There is no indication that it will. This
>means that if GenBank makes a correction it won't go into this database and
>vice versa. Since the data are not maintained in one place, it is duplicated
>and will eventually become inconsistent. Who should or will a researcher
>believe? A randomly annotated database or the "official" one? How will
>inconsistencies be resolved?
>>I see this as an experiment. Once the experiment has been shown to work
>reasonably well, all the data should be passed to NCBI for careful processing
>along with all new data. That is, the server should be run by NCBI or closely
>in conjunction with them.
>>For my reasoning behind this posting, see the philosophy paper available from:
>http://fconvx.ncifcrf.gov:2001/~toms/onlinepapers.html>> Tom Schneider
> National Cancer Institute
> Laboratory of Mathematical Biology
> Frederick, Maryland 21702-1201
>toms at ncifcrf.gov
