Following various suggestions i think we all converge on the following:
1)
a) A column of float is better because flaot are more general
columns are easier to query than lines
the treedisp will automatically come Closed showing the number of lines.
b) however the lenght of the ace file may seem awkward
2) The second tag system is more versatile, and best used as a
qualifier for the set of band values
3) Additional info may be requested
Therefore i implement the following model
1) A magic tag: Gel (3 letters only because in ace files it may have to be repeasted
millions of times
2) the folowing model:
Gel #Gel_Pattern
#Gel_Pattern Bands UNIQUE Float REPEAT
Band Float Text // why Unique ?
Motif ?Motif
Probe ?Probe
.... // this etc is left open to the curators
my display code will use again a second tag approach here
so you can use what you want
The display code will be called DtGelPattern
you can attach it to any class in options.wrm
YThe dispaly will accept any object conatining Gel#Band (or Bands)
the code is allready running, it has the following capabilitioes:
zoom/center
flip lanes
import new lanes via blocksubs
import keysets
sort the lanes automatically (in test)
it is invoked by the boxMenu in pmapdisp and by a button in
dnaanalysis
I will try to fus it with the artificial gel system which exists
separatelly
the only problem with Band in the the model is the ace format:
MyClass myObj
Gel Probe THE
Gel Band 23.1
Gel Band 28.8
Gel Band 43
Gel Band 87
i.e. Gel Bands is repeated million times
So i suggest to have an alternative Band or Bands up to the curator.
The following selectors are implemented:
In the window there are text box entries
if filled only the patterns fitting these will be displayed
other wise you may have 3 or more pattern per obj
Actually i wonder if the ?XXX is useful since the same sort of info
can be included in the #model
for the genethon i would use
Gel #Gel_Pattern
#Gel_Pattern Bands Float REPEAT
Enzyme ?Motif // one of ecor1 etc
Probe ?Probe // THE etc