GWA Test Driver


Genome-wide association studies in which the association of clinically relevant phenotypes (e.g., obesity, diabetes, lupus, pulmonary hypertension, etc.) with many (e.g., 1 million) single-nucleotide polymorphisms (SNPs) is perhaps the largest current wave in human genetic research. For investigators to be part of this wave, they need to answer key questions to their satisfaction and that of grant reviewers. Such questions include:

  • How many subjects do I need to study?
  • Am I better off studying more SNPs and fewer subjects or the reverse?
  • If I conduct a two-stage procedure where I replicate the topmost significant findings at a first stage with a fresh set of subjects in the second stage, what is the optimal number of SNPs to bring forward to stage 2?
  • For any design and analysis plan, what are my expected false positive and negative rates?

Often, these questions cannot be answered by simple calculations. Instead, they require highly computer intensive simulations and analyses of real data to provide information on likely distributions of effects . We have built this interactive web-based system to allow investigators to 'test-drive' combinations of prospective design and analysis plans, evaluate their performance, and automatically generate graphs characterizing that performance for grant applications.

Five ways GWA Test Driver can work for you
  1. We have downloaded a number of representative GWA datasets, incorporated the published analysis results where available (or analyzed the dataset ourselves), and processed each of them with our power analysis software. Investigators may search among the datasets for the experiment that most closely resembles their proposed project and view sample size and power estimates.

    Click here to search these real datasets.

  2. We have also simulated a number of GWA datasets with various characteristics that may be of interest to the investigator planning a new study.

    Click here to search these simulated datasets.

  3. Investigators may upload their own preliminary data and the program will extrapolate power from this dataset.

    Click here to use your own dataset.

  4. Analytic Power Calculation

    Click here to perform an analytic power calculation.

  5. Other Genetic Power Software

    Click here for links to other genetic power software.