Difference between revisions of "HLab:Gedi"
| (6 intermediate revisions by the same user not shown) | |||
| Line 4: | Line 4: | ||
This must be run on a local machine (not remotely). Input is tab delimited text file with expression levels that have been normalized and log2 based. Add headers to the file as described [https://apps.childrenshospital.org/clinical/research/ingber/GEDI/GEDI_InputDataFormat.htm here]. | This must be run on a local machine (not remotely). Input is tab delimited text file with expression levels that have been normalized and log2 based. Add headers to the file as described [https://apps.childrenshospital.org/clinical/research/ingber/GEDI/GEDI_InputDataFormat.htm here]. | ||
| − | Open the .jar file to run the program. Load the input file and run with default parameters. If the count of genes per box is high (many greater than 20) then rerun with a larger grid size parameters. Ideally most boxes will be between 1 and 20 genes. | + | Open the .jar file to run the program. Load the input file and run with default parameters. Gedi was written for an older Java and is not able to save a file unless the file is created first. From an unix prompt type "touch filename" then when you save it will ask if you want to replace the file, say yes. |
| + | |||
| + | If the count of genes per box is high (many greater than 20) then rerun with a larger grid size parameters. Ideally most boxes will be between 1 and 20 genes. To get the average count and quartile counts of genes per box export the Density Map Value and run the numbers through the Perl script [[Media:BasicStats.txt|basicStats.pl]]. | ||
| + | |||
| + | When you are happy with the parameters a higher resolution map can be drawn with the R script gedi.r. Export the Map Centroids, and note the grid size for input into [[Media:Gedi.r.txt|gedi.r]]. | ||
| + | |||
| + | Return to [[Hlab:Main|Main]] | ||
Latest revision as of 13:29, 7 January 2015
The GEDI home page.
To download the program you must register.
This must be run on a local machine (not remotely). Input is tab delimited text file with expression levels that have been normalized and log2 based. Add headers to the file as described here.
Open the .jar file to run the program. Load the input file and run with default parameters. Gedi was written for an older Java and is not able to save a file unless the file is created first. From an unix prompt type "touch filename" then when you save it will ask if you want to replace the file, say yes.
If the count of genes per box is high (many greater than 20) then rerun with a larger grid size parameters. Ideally most boxes will be between 1 and 20 genes. To get the average count and quartile counts of genes per box export the Density Map Value and run the numbers through the Perl script basicStats.pl.
When you are happy with the parameters a higher resolution map can be drawn with the R script gedi.r. Export the Map Centroids, and note the grid size for input into gedi.r.
Return to Main
