[MOBY-l] The idea of namespaces !!!
markw at illuminae.com
Tue Sep 11 21:06:31 EDT 2007
Unfortunately the idea of "Namespace" is not only the most confusing part
of Moby for newcomers, it is also a part of Moby that is (at least
partially) out of our control.
Our Namespace list is actually synchronized with the Gene Ontology Xref
Abbreviations List. In the Moby community we use these in specific ways,
however they aren't necessarily used *precisely* the same way in GO. This
is likely why there are some somewhat "ambiguous" namespaces in the Moby
namespace list (there is a non-Moby API into the Namespace Ontology that
is used by the GO curators!).
The rule-of-thumb is: If you need to identify it, then it needs a
namespace. i.e. if you need to be able to find a "thing" based on its
identifier, then that identifier needs a namespace that is sufficiently
specific that you don't have identifier collisions. If your splice
variants are all named as sub-species of your transcripts (e.g. transcript
100, splice variant 100.1 splice variant 100.2, etc.) then you really
don't need two namespaces for them...
On Tue, 11 Sep 2007 17:47:07 -0700, maliha aziz <maliha.aziz at yahoo.com>
> I am a little confused about namespaces ... would really appreciate it
> if someone could help me out...
> In biomoby namespaces are actually Identifier spaces... For my
> webservice that I am going to register in BIOMOBY I am thinking of
> declaring a namespace cluster ID ... this would be basically the
> namespace of one of my variables i.e. clusterID .Now I also have a
> variable called Splice Variant. This variable is a boolean and tells us
> whether the cluster contains a splice variant or not. Now what to do
> with this variable. Does it have it own namespace or should I change the
> clusterID namespace into something general that would be good enough for
> both.These were some of the thoughts running through my mind.
> I have opened and gone through the namspaces file contained in biomoby
> (through dashboard) quite a number of times since the past 6 months.What
> confuses me is that i see namespaces (in biomoby) for example
> Name: EcoliCommunity
> Auth: www.ecolicommunity.org
> Desc: A comprehensive K-12 information resource for experimentation
> and modeling through interoperation with exisitng datasources and
> warehousing other key data.
> which would be providing a very abstract source of anything they would
> On the other hand I see namespaces (in biomoby) like
> \u003cbr style\u003d\"color:rgb(204, 51, 204)\"\>\u003cspan
> style\u003d\"color:rgb(204, 51, 204)\"\>Auth: \u003ca
> href\u003d\"http://www.arabidopsis.org/\" target\u003d\"_blank\"
> style\u003d\"color:rgb(204, 51, 204)\"\>\n\u003cspan
> style\u003d\"color:rgb(204, 51, 204)\"\>Desc: Splice variant
> identifier from the Arabidopsis Genome
> Initiative.\u003c/span\>\u003cbr\>\u003cbr\>that have specifically made
> a namespace for their splice variant.\u003cbr\>\u003cbr\>Therefore\nI
> ama bit undecided about whether we should have a single namespace\nfor
> both or that a BIPASS_CLUSTERID would be the source of clusters
> and\nBIPASS_SPLICEVARIANT would be the source of whether a cluster
> possess\na variant or not.\u003cbr\>\u003cbr\>Once again your guidance
> and explanation is needed in this confusion
> Aziz \u003cbr\>\n\u003c/span\>\u003c/span\>\n",0] ); D(["ce"]); //-->
> Name: AGI_SpliceVariant
> Auth: www.arabidopsis.org
> Desc: Splice variant identifier from the Arabidopsis Genome
> that have specifically made a namespace for their splice variant.
> Therefore I ama bit undecided about whether we should have a single
> namespace for both or that a BIPASS_CLUSTERID would be the source of
> clusters and BIPASS_SPLICEVARIANT would be the source of whether a
> cluster possess a variant or not.
> Maliha Aziz
> Need a vacation? Get great deals to amazing places on Yahoo! Travel.
> moby-l mailing list
> moby-l at lists.open-bio.org
More information about the moby-l