Talk:GeneDatabaseLayout
From PathVisio Wiki
This is similar to an idea we talked about a while ago.
You can implement an object-oriented scheme using tables like this:
Gene
| column | type | key | description |
|---|---|---|---|
| id | varchar(50) | PRI | datanode identifier |
| code | varchar(50) | PRI | identifier database code |
| bpText | varchar | backpage text (html formatted) |
GeneWithDescription
| column | type | key | description |
|---|---|---|---|
| id | varchar(50) | PRI | datanode identifier |
| code | varchar(50) | PRI | identifier database code |
| description | varchar | description |
GeneWithChromosomeLocation
| column | type | key | description |
|---|---|---|---|
| id | varchar(50) | PRI | datanode identifier |
| code | varchar(50) | PRI | identifier database code |
| chromosome | varchar(5) | Chromosome number (roman) | |
| location | number | Position on reference sequence |
But then, your idea may be much simpler.
Alex 20:35, 25 July 2007 (CEST)
I don't think the WHERE - IN clause will work here. As far as I understand it is for 1-to-many matches, i.e., WHERE one IN (many). I would use the following instead:
SELECT * FROM Gene LEFT JOIN Attributes USING (id, code) WHERE attName = 'symbol' AND attValueE = 'TNF1'
As long as id and code exist in both tables this works and there is no need for a unique key.
