STAARpipeline v0.9.7

.github/workflows/R-CMD-check.yaml

@@ -4,7 +4,7 @@ name: R-CMD-check
 
 jobs:
   R-CMD-check:
-    runs-on: macOS-latest
+    runs-on: macOS-13
     env:
       GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
     steps:

DESCRIPTION

@@ -2,7 +2,7 @@ Package: STAARpipeline
 Type: Package
 Title: STAARpipeline for Analyzing Whole-Genome/Whole-Exome Sequencing Data
 Version: 0.9.7
-Date: 2023-11-09
+Date: 2024-05-19
 Author: Xihao Li [aut, cre], Zilin Li [aut, cre], Sheila M. Gaynor [aut], Han Chen [aut]
 Maintainer: Xihao Li <[email protected]>, Zilin Li <[email protected]>
 Description: An R package for performing STAARpipeline in analyzing whole-genome/whole-exome sequencing data.

R/Individual_Analysis.R

@@ -56,17 +56,13 @@ Individual_Analysis <- function(chr,start_loc,end_loc,genofile,obj_nullmodel,mac
 	}
 
 	## residuals and cov
+	residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
 	if(SPA_p_filter)
 	{
 		### dense GRM
 		if(!obj_nullmodel$sparse_kins)
 		{
 			P <- obj_nullmodel$P
-			P_scalar <- sqrt(dim(P)[1])
-			P <- P*P_scalar
-
-			residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
-			residuals.phenotype <- residuals.phenotype*sqrt(P_scalar)
 		}
 
 		### sparse GRM
@@ -75,12 +71,7 @@ Individual_Analysis <- function(chr,start_loc,end_loc,genofile,obj_nullmodel,mac
 			Sigma_i <- obj_nullmodel$Sigma_i
 			Sigma_iX <- as.matrix(obj_nullmodel$Sigma_iX)
 			cov <- obj_nullmodel$cov
-
-			residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
 		}
-	}else
-	{
-		residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
 	}
 
 	## SPA
@@ -110,10 +101,6 @@ Individual_Analysis <- function(chr,start_loc,end_loc,genofile,obj_nullmodel,mac
 		if(!obj_nullmodel$sparse_kins)
 		{
 			P <- obj_nullmodel$P
-			P_scalar <- sqrt(dim(P)[1])
-			P <- P*P_scalar
-
-			residuals.phenotype <- residuals.phenotype*sqrt(P_scalar)
 		}
 
 		### sparse GRM

R/Individual_Analysis_cond.R

@@ -218,24 +218,7 @@ Individual_Analysis_cond <- function(chr,individual_results,genofile,obj_nullmod
 			if(!obj_nullmodel$sparse_kins)
 			{
 				P <- obj_nullmodel$P
-				P_scalar <- sqrt(dim(P)[1])
-				P <- P*P_scalar
 
-				residuals.phenotype <- obj_nullmodel$scaled.residuals
-				residuals.phenotype <- residuals.phenotype*sqrt(P_scalar)
-
-				if(method_cond == "optimal"){
-					residuals.phenotype.fit <- lm(residuals.phenotype~genotype_adj+obj_nullmodel$X-1)
-				}else{
-					residuals.phenotype.fit <- lm(residuals.phenotype~genotype_adj)
-				}
-
-				residuals.phenotype <- as.vector(residuals.phenotype.fit$residuals)
-				X_adj <- model.matrix(residuals.phenotype.fit)
-				if(n_pheno > 1)
-				{
-					X_adj <- as.matrix(Diagonal(n = n_pheno) %x% X_adj)
-				}
 				PX_adj <- P%*%X_adj
 				P_cond <- P - X_adj%*%solve(t(X_adj)%*%X_adj)%*%t(PX_adj) -
 							PX_adj%*%solve(t(X_adj)%*%X_adj)%*%t(X_adj) +
@@ -255,28 +238,24 @@ Individual_Analysis_cond <- function(chr,individual_results,genofile,obj_nullmod
 			}
 		}else
 		{
+			residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
 			### sparse GRM
 			if(obj_nullmodel$sparse_kins)
 			{
 				if(n_pheno == 1)
 				{
-					Score_test <- Individual_Score_Test(as.matrix(Geno[,k],ncol=1), Sigma_i, Sigma_iX, cov, obj_nullmodel$scaled.residuals)
+					Score_test <- Individual_Score_Test(as.matrix(Geno[,k],ncol=1), Sigma_i, Sigma_iX, cov, residuals.phenotype)
 				}
 				else
 				{
-					Score_test <- Individual_Score_Test_multi(as.matrix(Diagonal(n = n_pheno) %x% Geno[,k]), Sigma_i, Sigma_iX, cov, obj_nullmodel$scaled.residuals, n_pheno)
+					Score_test <- Individual_Score_Test_multi(as.matrix(Diagonal(n = n_pheno) %x% Geno[,k]), Sigma_i, Sigma_iX, cov, residuals.phenotype, n_pheno)
 				}
 				pvalue_cond_log10 <- c(pvalue_cond_log10,Score_test$pvalue_log/log(10))
 			}
 			### dense GRM
 			if(!obj_nullmodel$sparse_kins)
 			{
 				P <- obj_nullmodel$P
-				P_scalar <- sqrt(dim(P)[1])
-				P <- P*P_scalar
-
-				residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
-				residuals.phenotype <- residuals.phenotype*sqrt(P_scalar)
 
 				if(n_pheno == 1)
 				{

R/Individual_Analysis_cond_spa.R

@@ -3,9 +3,6 @@
 #' The \code{Individual_Analysis_cond_spa} function takes in chromosome, starting location, ending location,
 #' the object of opened annotated GDS file, and the object from fitting the null model to analyze the association between an
 #' imbalanced case-control phenotype and each individual variant in a genetic region by using score test.
-#' For multiple phenotype analysis (\code{obj_nullmodel$n.pheno > 1}),
-#' the results correspond to multi-trait score test p-values by leveraging
-#' the correlation structure between multiple phenotypes.
 #' @param chr chromosome.
 #' @param individual_results the data frame of (significant) individual variants for conditional analysis using score test.
 #' The first 4 columns should correspond to chromosome (CHR), position (POS), reference allele (REF), and alternative allele (ALT).
@@ -44,17 +41,13 @@ Individual_Analysis_cond_spa <- function(chr,individual_results,genofile,obj_nul
 	samplesize <- length(phenotype.id)
 
 	## residuals and cov
+	residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
 	if(SPA_p_filter)
 	{
 		### dense GRM
 		if(!obj_nullmodel$sparse_kins)
 		{
 			P <- obj_nullmodel$P
-			P_scalar <- sqrt(dim(P)[1])
-			P <- P*P_scalar
-
-			residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
-			residuals.phenotype <- residuals.phenotype*sqrt(P_scalar)
 		}
 
 		### sparse GRM
@@ -63,15 +56,9 @@ Individual_Analysis_cond_spa <- function(chr,individual_results,genofile,obj_nul
 			Sigma_i <- obj_nullmodel$Sigma_i
 			Sigma_iX <- as.matrix(obj_nullmodel$Sigma_iX)
 			cov <- obj_nullmodel$cov
-
-			residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
 		}
-	}else
-	{
-		residuals.phenotype <- as.vector(obj_nullmodel$scaled.residuals)
 	}
 
-
 	## SPA
 	muhat <- obj_nullmodel$fitted.values
 
@@ -92,7 +79,6 @@ Individual_Analysis_cond_spa <- function(chr,individual_results,genofile,obj_nul
 		XXWX_inv <- obj_nullmodel$XXWX_inv
 	}
 
-
 	## get SNV id
 	filter <- seqGetData(genofile, QC_label)
 	if(variant_type=="variant")

README.md

@@ -1,5 +1,4 @@
 [![R build status](https://github.com/xihaoli/STAARpipeline/workflows/R-CMD-check/badge.svg)](https://github.com/xihaoli/STAARpipeline/actions)
-[![Build Status](https://travis-ci.com/xihaoli/STAARpipeline.svg?branch=main)](https://app.travis-ci.com/github/xihaoli/STAARpipeline)
 [![Build status](https://ci.appveyor.com/api/projects/status/ltr225p13idh2934/branch/main?svg=true)](https://ci.appveyor.com/project/xihaoli/staarpipeline/branch/main)
 [![License: GPL v3](https://img.shields.io/badge/License-GPLv3-blue.svg)](https://www.gnu.org/licenses/gpl-3.0)
 
@@ -32,7 +31,7 @@ Please see the <a href="docs/STAARpipeline_manual.pdf">**STAARpipeline** user ma
 ## Data Availability
 The whole-genome functional annotation data assembled from a variety of sources and the precomputed annotation principal components are available at the [Functional Annotation of Variant - Online Resource (FAVOR)](https://favor.genohub.org) site and [FAVOR Essential Database](https://doi.org/10.7910/DVN/1VGTJI).
 ## Version
-The current version is 0.9.7 (February 16, 2024).
+The current version is 0.9.7 (May 19, 2024).
 ## Citation
 If you use **STAARpipeline** and **STAARpipelineSummary** for your work, please cite: