When using scantwopermhk, the first phenotype column is always used regardless of user

Apologies for a less elegant example, but: <div class="snippet-clipboard-content n

scantwopermhk ignores pheno.col argument about qtl HOT 3 CLOSED

Wang-RJ commented on September 2, 2024

scantwopermhk ignores pheno.col argument

from qtl.

Comments (3)

kbroman commented on September 2, 2024

Do you have an example that demonstrates this? Because it seems to be using it appropriately.

Consider this example:

library(qtl)
data(hyper)
hyper <- calc.genoprob(hyper[1:2,])
hyper$pheno[,3] <- sqrt(hyper$pheno[,1])

set.seed(10347067)
out1 <- scantwopermhk(hyper, pheno.col=1, n.perm=3)
set.seed(10347067)
out3 <- scantwopermhk(hyper, pheno.col=3, n.perm=3)
set.seed(10347067)
out1b <- scantwopermhk(hyper, pheno.col=1, n.perm=3)

all(unlist(out1) == unlist(out1b))
all(unlist(out1) != unlist(out3))

from qtl.

Wang-RJ commented on September 2, 2024

Apologies for a less elegant example, but:

library(qtl)
load("goughF2_simple_v4.RData")

reduced_f2 <- calc.genoprob(f2, step = 1, error.prob = 0.0002, map.function = 'c-f')
reduced_f2 <- reduce2grid(reduced_f2)
reduced_f2 <- subset(reduced_f2, ind = reduced_f2$pheno$sex == "M")

nrow(reduced_f2$pheno)

out1 <- scantwopermhk(reduced_f2, pheno.col = 'wk5', assumeCondIndep = TRUE, n.perm = 1)
out2 <- scantwopermhk(reduced_f2, pheno.col = 2, assumeCondIndep = TRUE, n.perm = 1)

summary(out1);summary(out2)

wk1 (1 permutations)
full fv1 int add av1 one
5% 7.7 5.83 5.05 5.18 3.9 2.35
10% 7.7 5.83 5.05 5.18 3.9 2.35
wk1 (1 permutations)
full fv1 int add av1 one
5% 8.69 7.6 5.58 5.09 2.61 2.76
10% 8.69 7.6 5.58 5.09 2.61 2.76

from qtl.

kbroman commented on September 2, 2024

Ah; it's a problem in the pasting over of the appropriate column name. Got it fixed in the devel branch.

from qtl.

Recommend Projects