In article <00972869.51070240.8005 at BSCR.UGA.EDU> weise at BSCR.UGA.EDU ("Michael J. Weise") writes:
>> Anybody know the relative value of GeneSeq, now that both EMBL and
>GenBank have categories for patented sequences?
>> MJW
>>>
GeneSeq is produced by Derwent who specialise in patent work. They have cleared
the patent backlog to the early 80's in GeneSeq. The last quoted figures I read
though these were form Derwent was that all sequences from Derwent weeks 8401 to
9308 were contained in Release 11.0 of Geneseq.
This contained 16,009,476 bases in 30,843 nucleic acid sequences and
4,048,030 residues in 30,847 protein and peptide sequences
Apparently a couple of studies have been done to show the "uniqueness" of
GeneSeq one a sample based study against thr public databanks showed the
percentage unique in Geneseq was as follows:
Nucleotide Sequences ; 56% unique
Amino Acid Sequences ; 63% unique the sample was around 100 sequences of each
type which it was claimed at the time to be a statistically representative
sample.
A later study using BLAZE to search the whole of GeneSeq against Swiss-Prot 21
showed that of the 17,273 sequences then in Geneseq 9,200 were unique (53%)
which was in rough accord with the earlier figures.
I am not associated with either Derwent or Ig other than as a customer. If you
want to check out the details contact:
ig-consultant at presto.ig.com
Any feedback you get i would be interested to hear.
Rgds,
Tony Parsons (mbpcr at seqnet.dl.ac.uk MORE frequently parsons_a at snd01.pcr.co.uk)
