Question: Extract lower limit of pvalue less than 2.2*10^-16
gravatar for titua19810
4.4 years ago by
titua198100 wrote:


I am using linear mixed model to get p values that are used to plot graph . The R package that I am using for the model is multcomp and the function that is used from multcomp package is glht. From the output I get some pvalues that are less than 2. 210^-16 but in the output it is shown as value of 0. I need to extract the exact p values for pvalues less than 2.210^-16 instead of 0's.

The following command are used to run the model and I extraced p values from the summary(model) in R

test <- xyz[1:115000,]
test$group <- rep(1:1000, each=115)

ngroup = max(test$group)

#code within tryCatch will proceed to next iteration even if error
pvals_drug_1 <- matrix(rep(-1,ngroup*6), nrow=ngroup, ncol=6)
pvals_time_1 <- matrix(rep(-1,ngroup*4), nrow=ngroup, ncol=4)
pvals_omni_interaction_1<- matrix(rep(-1,ngroup*1), nrow=ngroup, ncol=1)
for (i in 1:ngroup) {
    tryCatch ({
        m1 <- lmer(ratio ~ conc*week + (1|pool/cell), data=subset(test, test$group==i))
        s1 <- glht(m1, linfct=c("conc + 1*conc:week == 0",
                                              "conc + 2*conc:week == 0",
                                              "conc + 3*conc:week == 0",
                                              "conc + 4*conc:week == 0",
                                              "conc + 6*conc:week == 0",
                                              "conc + (3.2)*conc:week==0"))
    s2 <- glht(m1, linfct=c("week  == 0",
                                           "week + 8*conc:week == 0",
                                           "week + 25*conc:week == 0",
                                           "week + 75*conc:week == 0"))

   pvals_drug_1[i,] <- summary(s1)$test$pvalues[1:6]
   pvals_time_1[i,] <- summary(s2)$test$pvalues[1:4]
   pvals_omni_interaction_1[i,]<- summary(m1)[[10]][20]
  }, error=function(e){cat("Error on line ", i, ": ", conditionMessage(e), "\n")})

I would highly appreciate if anybody can guide me how to exact pvalues that are less than 2.2*10^-16 so that I can make a QTL plot with right hieght of the peaks of multiple hypothesis adjusted -logQ values against human genome.

R • 1.2k views
ADD COMMENTlink modified 4.4 years ago by Devon Ryan98k • written 4.4 years ago by titua198100
gravatar for Devon Ryan
4.4 years ago by
Devon Ryan98k
Freiburg, Germany
Devon Ryan98k wrote:

I'll just link this older cross-validated thread, which largely summarizes things.

ADD COMMENTlink written 4.4 years ago by Devon Ryan98k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1462 users visited in the last hour