[Genome] tissue distribution of genes

[Archana Thakkapallayil]

Hello Dr. Khuri,

You can get this information using the Table Browser by connecting 
information from the gnfHumanAtlas2Median and gnfHumanAtlas2MedianExps 
tables in the hgFixed database, and the knownToGnfAtlas2 table in the 
hg18 database.

Click the "Tables" link at the top of the page and make the following 
selections:
set the clade, genome, and assembly
group: All Tables
database: hgFixed

1. The table hgFixed.gnfHumanAtlas2MedianExps connects an id with the 
type of tissue (there are 79 tissue types in this data set). For example:

#id    name
58    heart
60    skeletal muscle
71    liver
etc.


2. The hgFixed.gnfHumanAtlas2Median table has a column for the name of 
the probe and a comma separated column of values for the experimental 
scores.

 Liver has an id=71 , so you will need to write your own program to 
parse the gnfHumanAtlas2Median table and grab the 72nd comma separated 
value (tissue ids are indexed starting with 0) from the expScore column, 
to get data for only the liver tissue type.

3. You can then relate the probes to gene name by using the 
hg18.knownToGnfAtlas2 table.

Note that this only gives you access to the gnfAtlas2 data which contain 
values for "genes" that were expressed in each tissue type.  But a 
cutoff to decide whether a gene was actually expressed is up to you.  
Here is information on how this is determined via affymetrix since the 
gnf experiments were done on affymetrix arrays.
 
"Statistical Algorithms Reference Guide",  
http://www.affymetrix.com/support/technical/technotesmain.affx

You can get the original data here:
http://wombat.gnf.org/index.html

Also, please see this previously-answered mail list question on GNF 
expression data for mouse.

https://www.soe.ucsc.edu/pipermail/genome/2006-September/011672.html

I hope this was helpful to you. Please let us know if you have further 
questions.

Regards,

--
Archana
UCSC Genome Bioinformatics Group


Khuri, Sawsan wrote:
> Hello there,
>
>  
>
> I wonder if you can please help:  I'd like to end up with a dataset of
> human genes that are expressed in liver, and a second dataset of genes
> that are expressed in heart, another for those that are expressed in
> muscle, etc for all the major human tissue groups.  I do not necessarily
> want only the tissue specific genes in those datasets, and I don't mind
> the fact that housekeeping genes that will occur in every one of my
> datasets.  Is there a way to do this using the UCSC genome browser?  I
> tried a look at your Tables, but since there isn't a 'tissue
> distribution' track (or is there???), I wasn't sure how to do it.
>
>  
>
> Many thanks!
>
> Sawsan
>
>  
>
>  
>
> ----------------------//
>
> Sawsan Khuri, Ph.D.
>
> Assistant Research Professor
>
> The Dr. John T. Macdonald Foundation Center for Medical Genetics
>
> University of Miami Miller School of Medicine
>
> 1580 NW 10th Ave (BCRI R-189) Rm 606
>
> Miami, FL 33136
>
> Tel. + 1 (305) 243-6069
>
> Fax + 1 (305) 243-3919
>
> Center URL http://medgen.med.miami.edu <http://medgen.med.miami.edu> 
>
> Homepage
> http://www.medgen.med.miami.edu/research/faculty_profiles/khuri.asp
> <http://www.medgen.med.miami.edu/research/faculty_profiles/khuri.asp> 
>
>  
>
>  
>
> _______________________________________________
> Genome maillist  -  Genome at soe.ucsc.edu
> http://www.soe.ucsc.edu/mailman/listinfo/genome
>