IUBio Biosequences .. Software .. Molbio soft .. Network News .. FTP

Query problems: subtags and constructed subtypes

Fred Wobus fw at sanger.ac.uk
Fri Sep 17 05:20:41 EST 1999

I've found some suitable AQL queries for the problem that is
described here. I tested them on the Sanger Chr6 data using
AQL has seen a lot of development since version 4.7, so you
may encounter some problems if you start playing with
advanced queries. We're currently testing our development
version where most initial AQL problems have been fixed.

Note that tace doesn't understand multiline queries (you can
add a \ at the end of each line to wrap around). The textual
formatting across multiple lines just looks nice and shows
how the query is structured.
Both queries make use of the compound notation to access a
local tag. We dereference the object from the class by
following the map tag (the arrow is the notation for
following a tag). But once we've positioned mn (MapName) at
that position we want to stay there in the object and go to
the local tags position or left/right within the same
object. Therefore the [] notation.

On Thu, 16 Sep 1999 mummi at decode.is wrote:

>   I have been having problems with querying subtags within constructed
> subtypes. I am trying to extract all the contigs from a certain region
> on Chromosome 6 (the Sanger FTP-data), querying the Map tag and
> continuing to the Map_position. This works just fine with, say, the STS
> objects which only have one Position value
>   "find STS Map = Chr_6 & NEXT & NEXT > 70000 AND HERE <100000"

select s, pos 
from s in class STS, 
     mn in s->Map 
          where mn = "Chr_6", 
     pos in mn[Position] 
            where pos > 70000 and pos < 100000

I get 200 lines in the result.

> but the contigs are maps with two location valus, Left and Right. I
> imagined  that doing
>   "find Map Map = Chr_6 & NEXT # Ends # Left >70000"

select m, l, r 
from m in class map, 
     mn in m->Map
          where mn = "Chr_6", 
     l in mn[left] 
          where l > 70000, 
     r in mn[right]

The result table has 250 lines.

> would work or something similar, but none of the stuff I've tried
> worked.
>    I wound up just picking the relevant contigs out by hand and into a
> keyset, but this query-problem  keeps nagging me still. is it perhaps
> not possible to query such subtags (Ends split into Left and Right)
> inside a constructed subtype like Map_position?
>    Thanks in advance, greetings from Iceland to everybody on the list ;)
>                     Gudmundur
> Sent via Deja.com http://www.deja.com/
> Share what you know. Learn what you don't.

Fred Wobus                            Phone: +44-(0)7971-085893
                                      Web:   http://www.sanger.ac.uk/~fw


More information about the Acedb mailing list

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