IUBio

Gel patterns

Richard Durbin rd at mrc-lmb.cam.ac.uk
Mon Apr 11 08:45:24 EST 1994


I like the option of Bands or Band.  I agree that Text following float
does not have to be UNIQUE -- I just thought that by default only the
first would be used as an explicit label.

But removing the ?XXX makes an important difference.  As you have it, wiuth
	Gel #Gel_Pattern
you can only attach one pattern to an object.  This is because the #Gel_pattern
is a unique subobject.  You could have two Probe entries in it and two Bands, 
but there would be nothing to attach one set of bands to a particular probe.
The alternative
	Gel ?XXX #Gel_Pattern
lets you attach a label to each gel_pattern, so you can differentiate them.  It
is very important.  Of course in any one model you would have, e.g.
	Gel ?Text #Gel_Pattern
but I thought that the DtGelPattern code might as well not care about the
class of this first key -- just use it as a label.  You could even have a tag:
	Gel Line1_fp #Gel_Pattern
	    Line2_fp #Gel_Pattern
	    Alu_pcr #Gel_Pattern
In fact, I imagine that this could well be the most common usage.  The .ace
file would then say:

Clone F37C3
Line1_fp Bands 2.3 4.8 4.9 15.6
Alu_pcr Bands 1.3 17.6 26.1

So I argue strongly that you are not using your marvellous invention of the
second tag system in your proposal.

Everything else looked good.

Richard




More information about the Acedb mailing list

Send comments to us at biosci-help [At] net.bio.net