Earlier this spring we announced that Meta-Analysis was coming to SVS very soon. Now, I am pleased to announce that it is available in the latest release of SVS (version 8.4.0).
Meta-Analysis takes the results of two or more GWAS studies for multiple SNPs or markers, and standard meta-analysis statistics are then performed on each SNP and the results compiled into one spreadsheet. SVS can perform meta-analysis on results created within the SVS software or from third-party software programs or a combination of the two. Results for a fixed-effects model, random-effects model and tests for heterogeneity between studies are automatically computed for every meta-analysis performed.
To run Meta-Analysis, open a SVS project. If you are performing a meta-analysis on results reported by other researchers or provided from a data repository, import these files into SVS using the numerous data import utilities available. If you are performing meta-analysis on results that you generated with SVS, open the project containing the data.
With a project open, from the project navigator, go to Tools > Perform Meta-Analysis. Select the results spreadsheets you wish to perform meta-analysis on. Each study’s results should be in a separate spreadsheet, and each should have results from the same type of study. In other words, all have p-values, effect sizes and direction of effect (called sample-size based inputs), or all have either effect sizes with standard errors or odds ratios with confidence intervals (called inverse-variance based inputs). The initial meta-analysis window looks like Figure 1 below.
Figure 1: Meta-Analysis Initial Window
There are more options for parameters for the first study, as this is how it is determined if the meta-analysis will use sample-size based or inverse-variance based inputs. Subsequent studies will have inputs restricted by the options set for the first study. If there is a marker map applied to the first spreadsheet, the options dialog for Study #1 will look like Figure 2 below.
Figure 2: Meta-Analysis Window for the First Study
Since Sample-Size-Based inputs were selected for Study #1, only that tab will be in the dialog for the other studies. Figure 3 shows what the dialog for Study #2 would look like.
Figure 3: Meta-Analysis Window for Subsequent Studies (Sample-Size-Based)
For more information about the statistics computed, the input or output data, please see the documentation in the SVS Manual.
Also in this release is a drastic change in our license system that will streamline the user experience across all of Golden Helix’s products. Because of this change, we ask that customers contact us and let us know that they would like to upgrade to the latest version. Just email firstname.lastname@example.org and let us know your existing license ID or key and we can help you move over to the latest version.
Figure 4: The new SVS Login Screen
If you have any questions or would like assistance, please feel free to contact email@example.com.