longtests/testthat/test-check_impute_se_beta.R
In neurogenomics/MungeSumstats: Standardise summary statistics from GWAS
test_that("Check that imputation columns added correctly", {
## The following test uses more than 2GB of memory, which is more
## than what 32-bit Windows can handle:
is_32bit_windows <- .Platform$OS.type == "windows" #&&
#.Platform$r_arch == "i386"
if (!is_32bit_windows) {
#Our VCF file contains a BETA and SE column, let's impute these and
#compare the difference
# write the ALS GWAS, VCF file to a temp file for testing
vcf_path <- system.file("extdata", "ALSvcf.vcf",
package = "MungeSumstats")
sumstats_dt <- MungeSumstats:::read_vcf(path = vcf_path)
#add in known sample size of study
sumstats_dt[,N:=293723]
# write to temp dir
file <- tempfile()
data.table::fwrite(sumstats_dt, file)
#run control through
control <- MungeSumstats::format_sumstats(file,
ref_genome = "GRCh37",
on_ref_genome = FALSE,
strand_ambig_filter = FALSE,
bi_allelic_filter = FALSE,
allele_flip_check = FALSE,
imputation_ind = TRUE,
log_folder_ind = TRUE,
INFO_filter = 0,
dbSNP=144
)
control_dt <- data.table::fread(control$sumstats)
#rerun control through with ES not equal to beta
a<- data.table::fread(file)
a[,Beta:=ES]
control_es <- MungeSumstats::format_sumstats(a,
ref_genome = "GRCh37",
on_ref_genome = FALSE,
strand_ambig_filter = FALSE,
bi_allelic_filter = FALSE,
allele_flip_check = FALSE,
imputation_ind = TRUE,
log_folder_ind = TRUE,
INFO_filter = 0,
dbSNP=144,
es_is_beta = FALSE
)
control_dt_es <- data.table::fread(control_es$sumstats)
control_dt_es[,ES:=NULL]
testthat::expect_equal(all.equal(control_dt,control_dt_es,
ignore.col.order = TRUE),TRUE)
#now impute SE
sumstats_dt_org <- copy(sumstats_dt)
sumstats_dt[,SE:=NULL]
data.table::fwrite(sumstats_dt, file)
reformatted <- MungeSumstats::format_sumstats(file,
ref_genome = "GRCh37",
impute_se = TRUE,
on_ref_genome = FALSE,
strand_ambig_filter =
FALSE,
bi_allelic_filter = FALSE,
allele_flip_check = FALSE,
imputation_ind = TRUE,
log_folder_ind = TRUE,
INFO_filter = 0,
dbSNP=144
)
#first check it was imputed
imp_se_dt <- data.table::fread(reformatted$sumstats)
expect_equal("SE" %in% colnames(imp_se_dt), TRUE)
expect_equal("IMPUTATION_SE" %in% colnames(imp_se_dt), TRUE)
#check how close it is to actual
setkey(imp_se_dt,SNP)
setkey(control_dt,SNP)
control_dt[imp_se_dt,imp_se:=i.SE]
#check diff as percentage of value
control_dt[,diff_se:=abs((SE-imp_se)/SE)]
#expect a mean difference below 0.1%
testthat::expect_equal(mean(control_dt$diff_se)<0.001, TRUE)
#-----------
# Now let's impute beta
# use sumstats with Z and beta:
# Genome-wide association meta-analysis (N=269,867) identifies new
# genetic and functional links to intelligence.
# https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6411041/
# PMID: 29942086 PMCID: PMC6411041 DOI: 10.1038/s41588-018-0152-6
#manually writing 25 rows because of package size limits
sumstats_dt_org <-
data.table::data.table(SNP=c('rs12184267','rs12184277','rs12184279',
'rs116801199','rs12565286','rs2977670',
'rs28454925','rs116720794','rs4951859',
'rs146277091','rs3094315','rs3131972',
'rs3115860','rs2073813','rs12184312',
'rs12184325','rs3131969','rs3131968',
'rs12184313','rs3131967','rs3115859',
'rs10454459','rs3131966','rs3131965',
'rs3115858'),
CHR=c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1,1,1,1,1,1),
BP=c(715265,715367,717485,720381,721290,
723891,
726794,729632,729679,752478,752566,
752721,
753405,753541,754063,754105,754182,
754192,
754211,754334,754503,754629,754964,
755775,
755890),
A1=c('C','A','C','G','G','G','C','C','C',
'G','G',
'A','C','G','G','C','A','A','G','T',
'G','A','C','A','A'),
A2=c('T','G','A','T','C','C','G','T','G','A',
'A',
'G','A','A','T','T','G','G','A','C','A',
'G','T','G','T'),
FRQ=c(0.9591931,0.9589313,0.9594241,0.957838,
0.9576224,0.06312,0.9590545,0.9589005,
0.183523,0.9592701,0.180135,0.186495,
0.147228,0.848645,0.9590545,0.9590083,
0.154419,0.154389,0.9587773,0.153927,
0.185618,0.9585925,0.184909,0.177333,
0.147844),
Z=c(-0.916,-0.656,-1.05,-0.3,-0.566,0.253,
-0.539,
-0.27,0.208,-0.263,0.351,0.61,0.48,
-0.441,
-0.181,-0.217,0.344,0.434,-0.212,0.509,
0.823,-0.351,0.828,0.808,0.414),
BETA=c(-0.0068872979,-0.0049144905,
-0.0079116035,
-0.0022174032,-0.0041742154,
0.0015498434,
-0.0040387267,-0.0020185581,
0.0008058415,
-0.0019709245,0.0013480759,
0.0023117892,
0.0019985366,-0.0018164389,
-0.0013546785,
-0.0016223525,0.0013769675,
0.001760705,
-0.0015823145,0.0020819201,
0.003133556,
-0.0026146926,0.003159564,
0.0032010789,0.0017171769),
SE=c(0.007518884,0.007491601,0.00753486,
0.007391344,0.007374939,0.006125863,
0.007492999,0.007476141,0.003874238,
0.007494009,0.003840672,0.003789818,
0.004163618,0.004118909,0.007484411,
0.007476279,0.004002812,0.004056924,
0.007463748,0.004090216,0.00380748,
0.007449267,0.003815899,0.003961731,
0.00414777),
P=c(0.3598,0.5116,0.2938,0.7644,0.5711,
0.8006,
0.5896,0.7872,0.835,0.7924,0.7257,0.5421,
0.631,0.6592,0.8567,0.8281,0.7309,0.6646,
0.8322,0.6105,0.4103,0.7257,0.4074,0.419,
0.6787),
N=c(225955,226215,226224,226626,226528,
225312,226782,226989,222312,227870,
229517,229459,229723,229447,227303,
227552,238992,232696,227092,229485,
228163,227002,227830,218366,230684))
sumstats_dt <- copy(sumstats_dt_org)
sumstats_dt[,BETA:=NULL]
data.table::fwrite(sumstats_dt, file)
reformatted <- MungeSumstats::format_sumstats(file,
ref_genome = "GRCh37",
impute_beta = TRUE,
on_ref_genome = FALSE,
strand_ambig_filter =
FALSE,
bi_allelic_filter = FALSE,
allele_flip_check = FALSE,
imputation_ind = TRUE,
log_folder_ind = TRUE,
dbSNP=144
)
imp_beta_dt <- data.table::fread(reformatted$sumstats)
#first check it was imputed
testthat::expect_equal("BETA" %in% colnames(imp_beta_dt), TRUE)
#check imputation column created
testthat::expect_equal("IMPUTATION_BETA" %in% colnames(imp_beta_dt),
TRUE)
#check how close it is to actual
setkey(imp_beta_dt,SNP)
setkey(sumstats_dt_org,SNP)
sumstats_dt_org[imp_beta_dt,imp_beta:=i.BETA]
#check diff as percentage of value
sumstats_dt_org[,diff_beta:=abs((BETA-imp_beta)/BETA)]
#expect a mean difference below 0.1%
testthat::expect_equal(mean(sumstats_dt_org$diff_beta,na.rm = TRUE)<0.001,
TRUE)
} else {
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
}
})
neurogenomics/MungeSumstats documentation built on Aug. 10, 2024, 5:59 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
test_that("Check that imputation columns added correctly", {
## The following test uses more than 2GB of memory, which is more
## than what 32-bit Windows can handle:
is_32bit_windows <- .Platform$OS.type == "windows" #&&
#.Platform$r_arch == "i386"
if (!is_32bit_windows) {
#Our VCF file contains a BETA and SE column, let's impute these and
#compare the difference
# write the ALS GWAS, VCF file to a temp file for testing
vcf_path <- system.file("extdata", "ALSvcf.vcf",
package = "MungeSumstats")
sumstats_dt <- MungeSumstats:::read_vcf(path = vcf_path)
#add in known sample size of study
sumstats_dt[,N:=293723]
# write to temp dir
file <- tempfile()
data.table::fwrite(sumstats_dt, file)
#run control through
control <- MungeSumstats::format_sumstats(file,
ref_genome = "GRCh37",
on_ref_genome = FALSE,
strand_ambig_filter = FALSE,
bi_allelic_filter = FALSE,
allele_flip_check = FALSE,
imputation_ind = TRUE,
log_folder_ind = TRUE,
INFO_filter = 0,
dbSNP=144
)
control_dt <- data.table::fread(control$sumstats)
#rerun control through with ES not equal to beta
a<- data.table::fread(file)
a[,Beta:=ES]
control_es <- MungeSumstats::format_sumstats(a,
ref_genome = "GRCh37",
on_ref_genome = FALSE,
strand_ambig_filter = FALSE,
bi_allelic_filter = FALSE,
allele_flip_check = FALSE,
imputation_ind = TRUE,
log_folder_ind = TRUE,
INFO_filter = 0,
dbSNP=144,
es_is_beta = FALSE
)
control_dt_es <- data.table::fread(control_es$sumstats)
control_dt_es[,ES:=NULL]
testthat::expect_equal(all.equal(control_dt,control_dt_es,
ignore.col.order = TRUE),TRUE)
#now impute SE
sumstats_dt_org <- copy(sumstats_dt)
sumstats_dt[,SE:=NULL]
data.table::fwrite(sumstats_dt, file)
reformatted <- MungeSumstats::format_sumstats(file,
ref_genome = "GRCh37",
impute_se = TRUE,
on_ref_genome = FALSE,
strand_ambig_filter =
FALSE,
bi_allelic_filter = FALSE,
allele_flip_check = FALSE,
imputation_ind = TRUE,
log_folder_ind = TRUE,
INFO_filter = 0,
dbSNP=144
)
#first check it was imputed
imp_se_dt <- data.table::fread(reformatted$sumstats)
expect_equal("SE" %in% colnames(imp_se_dt), TRUE)
expect_equal("IMPUTATION_SE" %in% colnames(imp_se_dt), TRUE)
#check how close it is to actual
setkey(imp_se_dt,SNP)
setkey(control_dt,SNP)
control_dt[imp_se_dt,imp_se:=i.SE]
#check diff as percentage of value
control_dt[,diff_se:=abs((SE-imp_se)/SE)]
#expect a mean difference below 0.1%
testthat::expect_equal(mean(control_dt$diff_se)<0.001, TRUE)
#-----------
# Now let's impute beta
# use sumstats with Z and beta:
# Genome-wide association meta-analysis (N=269,867) identifies new
# genetic and functional links to intelligence.
# https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6411041/
# PMID: 29942086 PMCID: PMC6411041 DOI: 10.1038/s41588-018-0152-6
#manually writing 25 rows because of package size limits
sumstats_dt_org <-
data.table::data.table(SNP=c('rs12184267','rs12184277','rs12184279',
'rs116801199','rs12565286','rs2977670',
'rs28454925','rs116720794','rs4951859',
'rs146277091','rs3094315','rs3131972',
'rs3115860','rs2073813','rs12184312',
'rs12184325','rs3131969','rs3131968',
'rs12184313','rs3131967','rs3115859',
'rs10454459','rs3131966','rs3131965',
'rs3115858'),
CHR=c(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
1,1,1,1,1,1),
BP=c(715265,715367,717485,720381,721290,
723891,
726794,729632,729679,752478,752566,
752721,
753405,753541,754063,754105,754182,
754192,
754211,754334,754503,754629,754964,
755775,
755890),
A1=c('C','A','C','G','G','G','C','C','C',
'G','G',
'A','C','G','G','C','A','A','G','T',
'G','A','C','A','A'),
A2=c('T','G','A','T','C','C','G','T','G','A',
'A',
'G','A','A','T','T','G','G','A','C','A',
'G','T','G','T'),
FRQ=c(0.9591931,0.9589313,0.9594241,0.957838,
0.9576224,0.06312,0.9590545,0.9589005,
0.183523,0.9592701,0.180135,0.186495,
0.147228,0.848645,0.9590545,0.9590083,
0.154419,0.154389,0.9587773,0.153927,
0.185618,0.9585925,0.184909,0.177333,
0.147844),
Z=c(-0.916,-0.656,-1.05,-0.3,-0.566,0.253,
-0.539,
-0.27,0.208,-0.263,0.351,0.61,0.48,
-0.441,
-0.181,-0.217,0.344,0.434,-0.212,0.509,
0.823,-0.351,0.828,0.808,0.414),
BETA=c(-0.0068872979,-0.0049144905,
-0.0079116035,
-0.0022174032,-0.0041742154,
0.0015498434,
-0.0040387267,-0.0020185581,
0.0008058415,
-0.0019709245,0.0013480759,
0.0023117892,
0.0019985366,-0.0018164389,
-0.0013546785,
-0.0016223525,0.0013769675,
0.001760705,
-0.0015823145,0.0020819201,
0.003133556,
-0.0026146926,0.003159564,
0.0032010789,0.0017171769),
SE=c(0.007518884,0.007491601,0.00753486,
0.007391344,0.007374939,0.006125863,
0.007492999,0.007476141,0.003874238,
0.007494009,0.003840672,0.003789818,
0.004163618,0.004118909,0.007484411,
0.007476279,0.004002812,0.004056924,
0.007463748,0.004090216,0.00380748,
0.007449267,0.003815899,0.003961731,
0.00414777),
P=c(0.3598,0.5116,0.2938,0.7644,0.5711,
0.8006,
0.5896,0.7872,0.835,0.7924,0.7257,0.5421,
0.631,0.6592,0.8567,0.8281,0.7309,0.6646,
0.8322,0.6105,0.4103,0.7257,0.4074,0.419,
0.6787),
N=c(225955,226215,226224,226626,226528,
225312,226782,226989,222312,227870,
229517,229459,229723,229447,227303,
227552,238992,232696,227092,229485,
228163,227002,227830,218366,230684))
sumstats_dt <- copy(sumstats_dt_org)
sumstats_dt[,BETA:=NULL]
data.table::fwrite(sumstats_dt, file)
reformatted <- MungeSumstats::format_sumstats(file,
ref_genome = "GRCh37",
impute_beta = TRUE,
on_ref_genome = FALSE,
strand_ambig_filter =
FALSE,
bi_allelic_filter = FALSE,
allele_flip_check = FALSE,
imputation_ind = TRUE,
log_folder_ind = TRUE,
dbSNP=144
)
imp_beta_dt <- data.table::fread(reformatted$sumstats)
#first check it was imputed
testthat::expect_equal("BETA" %in% colnames(imp_beta_dt), TRUE)
#check imputation column created
testthat::expect_equal("IMPUTATION_BETA" %in% colnames(imp_beta_dt),
TRUE)
#check how close it is to actual
setkey(imp_beta_dt,SNP)
setkey(sumstats_dt_org,SNP)
sumstats_dt_org[imp_beta_dt,imp_beta:=i.BETA]
#check diff as percentage of value
sumstats_dt_org[,diff_beta:=abs((BETA-imp_beta)/BETA)]
#expect a mean difference below 0.1%
testthat::expect_equal(mean(sumstats_dt_org$diff_beta,na.rm = TRUE)<0.001,
TRUE)
} else {
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
expect_equal(is_32bit_windows, TRUE)
}
})
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.