Processes | Workflows | P-Value Adjustment

P -Value Adjustment
Smoothing methods can be applied that take into account p-values from neighboring, and possibly correlated, markers. That is, the peak length can be used to indicate significance in addition to the peak height.
This smoothing can be performed using the Fisher, Simes, or truncated product method ( TPM ) on sliding windows of markers, as described in the table below:
Replace the p -value at the center of the sliding window with the p -value of the t-statistic .
Replace the p -value at the center of the sliding window with the p -value computed using Simes’ method 2 .
Replace the p -value at the center of the sliding window with the p -value computed using Zaykin’s method 3 .
The TPM is a variation of Fisher’s method that leads to a different alternative hypothesis . With the TPM, rejection of the null hypothesis implies that there is at least one false null hypothesis among those with p -values less than the specified tau value .
Note : This method is equivalent to Fisher’s method when tau =1.

1
Selected methods are not available in all processes.

2
Simes, R.J. (1986). An Improved Bonferroni Procedure for Multiple Tests of Significance. Biometrika 73: 751–754.

3
Zaykin, D.V., Zhivotovsky, L.A., Westfall, P.H., and Weir, B.S. (2002). Truncated Product Method for Combining p -Values. Genetic Epidemiology 22: 170–185.
To Specify a P -Value Adjustment Method:
*
For Additional Information
For more information about these methods, see the Methods of Smoothing p-Values in the SAS/Genetics User's Guide.