DAPPLE Updates

Click here to return to the dapple homepage

-CI scores are now bonferroni corrected for the number of CIs. This is a rather harsh correction, since the number of CIs can be in the 1000s and the best p-value DAPPLE can achieve is 1/10001, corresponding to 10,000 permutations. Thus we suggest rank-ordering CIs to prioritize interesting ones, or relaxing your p-value cutoff. To remove this correction, divide the p-values by the number of CIs.
-SeedScore files are now sorted by p-value, there was a bug that prevented this.
-Iterate files now show up in the results section of the status update page, there was a bug that prevented this.
-You can now return to the main DAPPLE page from the status update page.

-The plot now has a legend to help match colors to region IDs.

-The common interactor file (*_CI) now outputs the number of disease proteins that the common interactor binds to. This is useful if you would like to prioritize common interactors by this value.

-The "nearest gene" option. This may be of interest to people who are inputing a list of SNPs and would like to use only the gene closest to each SNP, rather than all genes in the LD wingspan. This option is not relevant to other types of input, including genes, regions and genes with a region tag.

-Fixed a small bug. If the user inputs overlapping regions, they weren't being merged properly in some instances. This is now fixed.

-Optimized the code to be a bit faster, which required re-writing of a substantial portion. Please report any bugs to dapple@broadinstitute.org.

-We have implemented a new plotting function which will automatically appear on the results page. The benefit to this new function is that it is interactive. You can click on nodes to drag and drop them to a desired location. We are still fixing this up, but we hope that it drastically improves some of the visualization problems with large networks.

-A regulatory region flag has been put in place. This applies only to SNPs or regions as inputs. When DAPPLE looks for genes that overlap an input region (either defined from a SNP or explicitly as a region), it adds a default 50kb upstream and 50kb downstream to the gene. The reason is that causal variation can affect regulatory regions, so we'd like to capture that. Now, you can choose how big you'd like that regulatory region to be!
-We also spruced up the site a bit - changed a few pull downs to check boxes, added descriptions where needed, and added an option to receive emails about DAPPLE updates.
-We now allow input of virtually all description types. DAPPLE will automatically strip any non-alphanumeric characters or spaces, however. For example, "test_genes 2" would become "testgenes2".

-Indirect connections are now available in the results as a text file.

-We have implemented a new parallelization technique in oerder to improve the speed dramatically. The current limit is now 2000 inputs and it should finish with a run time of 4 hours or less.
-Inputs can be mixed: SNPs, genes and regions can be input together.
-The program now accepts hg19 and 1KG SNPs
-The iterate funtion has been removed. This can be manually accomplished by taking the "genes to prioritize" output and using it in the "genes to specify" field.

Click here to return to the dapple homepage