Combine p-values Using the Sum of Logs Method — metap_sumlog

This function combines multiple p-values using Fisher's method (sum of logarithms of p-values), which is often used in meta-analysis. It computes a chi-squared statistic and the corresponding p-value, based on the assumption that the input p-values are independent. This function is adapted from the sumlog function in the metap package.

Usage

metap_sumlog_pvals(p, log.p = FALSE)

Arguments

p: A numeric vector of p-values. Each p-value must lie in the range (0, 1]. Invalid values (e.g., <= 0 or > 1) are excluded.
log.p: Logical. If TRUE, the returned p-value is given as log(p).

Value

A list with the following components:

chisq: The chi-squared statistic computed as -2 * sum(log(p)).
df: The degrees of freedom, equal to 2 * length(validp).
p: The combined p-value corresponding to the chi-squared statistic.
validp: A vector of valid p-values after filtering out invalid ones.

Details

This function applies Fisher's method for combining p-values. It first filters out invalid p-values (values <= 0 or > 1). If fewer than two valid p-values remain, a warning is issued, and NA values are returned. If some p-values are filtered out, the function issues a warning indicating that some studies were omitted.

Fisher's method is based on the chi-squared distribution, with degrees of freedom proportional to the number of valid p-values (2 * number of valid p-values).

For further details on the original implementation, please refer to the sumlog function in the metap package: https://cran.r-project.org/web/packages/metap/metap.pdf

References

sumlog function from the metap package: https://cran.r-project.org/web/packages/metap/metap.pdf

Examples

# Example with valid p-values
pvals <- c(0.01, 0.03, 0.05, 0.2)
result <- metap_sumlog_pvals(pvals)
print(result)
#> $chisq
#> [1] 25.4338
#> 
#> $df
#> [1] 8
#> 
#> $p
#> [1] 0.001312015
#> 
#> $validp
#> [1] 0.01 0.03 0.05 0.20
#> 

# Example with some invalid p-values
pvals <- c(0.01, 0, 1.2, 0.05)
result <- metap_sumlog_pvals(pvals)
#> Warning: Some studies omitted
print(result)
#> $chisq
#> [1] 15.2018
#> 
#> $df
#> [1] 4
#> 
#> $p
#> [1] 0.004300451
#> 
#> $validp
#> [1] 0.01 0.05
#>