From 976953ab605ec38e1186b43654e3120b97835732 Mon Sep 17 00:00:00 2001
From: Vinicius Junqueira <junqueiravinicius@hotmail.com>
Date: Fri, 11 Jul 2025 14:18:20 -0300
Subject: [PATCH] adding new fQC function

---
 NAMESPACE                 |  1 +
 R/utils_fQC.R             | 36 ++++++++++++++++++++++++++++++++++++
 man/check.snp.hwe.chi2.Rd | 36 ++++++++++++++++++++++++++++++++++++
 3 files changed, 73 insertions(+)
 create mode 100644 man/check.snp.hwe.chi2.Rd

diff --git a/NAMESPACE b/NAMESPACE
index 5f52552..1418f51 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -13,6 +13,7 @@ export(check.sample.call.rate)
 export(check.sample.heterozygosity)
 export(check.snp.chromo)
 export(check.snp.hwe)
+export(check.snp.hwe.chi2)
 export(check.snp.maf)
 export(check.snp.mgf)
 export(check.snp.monomorf)
diff --git a/R/utils_fQC.R b/R/utils_fQC.R
index 225af25..19f2083 100644
--- a/R/utils_fQC.R
+++ b/R/utils_fQC.R
@@ -631,3 +631,39 @@ get.hwe.chi2 <- function(snp.summary) {
   pvalues <- pchisq(chi2stat, df = 1, lower.tail = FALSE)
   return(pvalues)
 }
+
+#' Check SNPs for Hardy-Weinberg equilibrium deviation using chi-square p-values
+#'
+#' This function identifies SNP markers whose Hardy-Weinberg equilibrium (HWE) chi-square p-values
+#' indicate significant deviation beyond a specified threshold. It uses the p-values computed by 
+#' \code{get.hwe.chi2} on the input summary data frame.
+#'
+#' @param snp.summary A data frame or matrix containing summary statistics for SNP markers. 
+#'        The row names should correspond to SNP identifiers. It must be compatible with 
+#'        the function \code{get.hwe.chi2}.
+#' @param max.dev A numeric value specifying the maximum acceptable p-value threshold.
+#'        SNPs with p-values below this threshold are considered as deviating from HWE.
+#'
+#' @return A character vector of SNP identifiers (rownames) that fail the HWE test (p-value < \code{max.dev}).
+#'         If no SNPs fail, an empty vector is returned.
+#'
+#' @details Any SNP with missing p-value (NA) is treated as not failing (returned as FALSE).
+#'
+#' @seealso \code{\link{get.hwe.chi2}}
+#'
+#' @examples
+#' # Example usage (assuming snp.summary is precomputed and get.hwe.chi2 is defined)
+#' # snps_failed <- check.snp.hwe.chi2(snp.summary, max.dev = 0.05)
+#'
+#' @export
+check.snp.hwe.chi2 <- function (snp.summary, max.dev) 
+{
+    pvalues <- get.hwe.chi2(snp.summary)
+    result <- pvalues < max.dev
+    result[is.na(result)] <- FALSE
+    snps <- NULL
+    if (sum(result) > 0) {
+        snps <- rownames(snp.summary[result, ])
+    }
+    return(snps)
+}
diff --git a/man/check.snp.hwe.chi2.Rd b/man/check.snp.hwe.chi2.Rd
new file mode 100644
index 0000000..53ac0c1
--- /dev/null
+++ b/man/check.snp.hwe.chi2.Rd
@@ -0,0 +1,36 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/utils_fQC.R
+\name{check.snp.hwe.chi2}
+\alias{check.snp.hwe.chi2}
+\title{Check SNPs for Hardy-Weinberg equilibrium deviation using chi-square p-values}
+\usage{
+check.snp.hwe.chi2(snp.summary, max.dev)
+}
+\arguments{
+\item{snp.summary}{A data frame or matrix containing summary statistics for SNP markers. 
+The row names should correspond to SNP identifiers. It must be compatible with 
+the function \code{get.hwe.chi2}.}
+
+\item{max.dev}{A numeric value specifying the maximum acceptable p-value threshold.
+SNPs with p-values below this threshold are considered as deviating from HWE.}
+}
+\value{
+A character vector of SNP identifiers (rownames) that fail the HWE test (p-value < \code{max.dev}).
+        If no SNPs fail, an empty vector is returned.
+}
+\description{
+This function identifies SNP markers whose Hardy-Weinberg equilibrium (HWE) chi-square p-values
+indicate significant deviation beyond a specified threshold. It uses the p-values computed by 
+\code{get.hwe.chi2} on the input summary data frame.
+}
+\details{
+Any SNP with missing p-value (NA) is treated as not failing (returned as FALSE).
+}
+\examples{
+# Example usage (assuming snp.summary is precomputed and get.hwe.chi2 is defined)
+# snps_failed <- check.snp.hwe.chi2(snp.summary, max.dev = 0.05)
+
+}
+\seealso{
+\code{\link{get.hwe.chi2}}
+}