search

CancelDeath / chibe

Automatically exported from code.google.com/p/chibe
0 stars 0 forks source link

Case sensitive gene symbols #8

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago 
When performing querying symbols seem to be case sensitive. For instance, I can 
query neighborhood of "TP53" but "tp53" says "unknown symbol". If all symbols 
are upper case, can't we just convert all input to uppercase?

Original issue reported on code.google.com by ugurdogr...@gmail.com on 28 Jun 2012 at 12:02

GoogleCodeExporter commented 8 years ago 
Human gene symbols are written in all uppercase letters but for instance, only 
the first letter is capitalized for mouse genes. If we accept all kinds of 
input and convert them into uppercase letters, this may result in misleading 
query results. For example, user can query "Per1" believing that the result 
will be based on the mouse gene whereas we will be returning human gene's 
result. To prevent confusions like these, in my opininon it is safer to 
continue with the current state.

Original comment by mervecak...@gmail.com on 3 Jul 2012 at 9:56

GoogleCodeExporter commented 8 years ago 
Can others think of a way to improve the current way we take input? I know 
Arman implemented some validation mechanism for gene sets for the portal, 
perhaps he has something to say?

Original comment by ugurdogr...@gmail.com on 4 Jul 2012 at 7:31

GoogleCodeExporter commented 8 years ago

Original comment by ugurdogr...@gmail.com on 4 Jul 2012 at 7:32

GoogleCodeExporter commented 8 years ago 
Not all human gene symbols are upper case. For instance 
[http://www.genenames.org/data/hgnc_data.php?hgnc_id=28678 C10orf107].

Original comment by ozgunba...@gmail.com on 5 Jul 2012 at 8:00

GoogleCodeExporter commented 8 years ago 
We are somewhat lucky on the portal side, since it is only and only human 
centric. So the gene name matching does not have to deal with the upper/lower 
case issues at all -- we convert all names to upper case, both on the query and 
the database side and match accordingly. Maybe we can make the same assumption 
since the gene->id mapping comes already from HGNC?

Original comment by arm...@gmail.com on 5 Jul 2012 at 8:26

GoogleCodeExporter commented 8 years ago

Original comment by mervecak...@gmail.com on 30 Jul 2012 at 2:12