[R] Compare correlation coefficients between samples
swaraj basu
projectbasu at gmail.com
Mon Feb 8 23:15:37 CET 2016
Dear All,
I have a dataframe of 1000 rows and 4 columns. Each row represents a pair
of vectors (in my case a pair of genes) while the columns represent the
following
estimate.A: spearman correlation coefficient of gene[i] and gene[j]
expression across 120 samples from cancer type A
prob.A: Probability value associated with the estimate.A as inferred from
cor.test
estimate.B: spearman correlation coefficient of gene[i] and gene[j]
expression across 48 samples from cancer type B
prob.B: Probability value associated with the estimate.A as inferred from
cor.test
To sum up the data.frame will look like
S<-data.frame(estimate.A=runif(1000,-1,1),prob.A=runif(1000,0,1),estimate.B=runif(1000,-1,1),prob.B=runif(1000,0,1))
I want to calculate the conditional probability for all the pair of genes
showing negative correlation in B (at p.value < 0.01) when they show a
strong negative correlation in A (again at p.value < 0.01). I was thinking
in the lines of using the "prob" package to estimate the conditional
probability but I am not able to figure out the correct way to do so in my
case. Thanks for any help in advance.
--
Swaraj Basu
Institute of Biomedicine
University of Gothenburg
[[alternative HTML version deleted]]
More information about the R-help
mailing list