Load the ISLR package into R. Build an FDR (Procedure is called Benjamini Hochbe

Question

Load the ISLR package into R.
Build an FDR (Procedure is called Benjamini Hochberg in book) program in R where you input a vector of pvalues and Q (assume independence)
In the NCI60 data base, for each cancer type in lab
Calculate a T statistic for each gene.
Using 1 sided tests separately for increased activity.
Use FDR to identify genes with enhanced activity for the cancer types (tune FDR to get at most 50-100 interesting genes) but don't make q>0.2
Use FDR to identify genes with suppressed activity for the cancer types (tune FDR to get at most 50-100 interesting genes) but don't make q>0.2
For each direction try to use the results of your separate FDR tests to find genes that have the same effect in multiple (2 or more) cancer types

Write up the results with any appropriate graphs you can think of and hand in
FDR program
Sorting and identification procedure description

Explanation / Answer

To load the ISLR package, we first type the command

install.packages("ISLR") . Once this package is istalled, we then load the package with the follwing command

library(ISLR)

Since data is not given the general approach tosolving the FDR problem is as below:

If you have the dataset loaded in as a dataframe then to calculate the False Positve Rate we use this function FDR(pvals, qlevel ) where pvals is a vector of false positives and qlevel is the proportion of false positives desired. To create appropriate graphs use the ggplot2() package for generating advanced graphics.

Related Questions

Navigate

• Browse (All)
• Browse L
• Subjects

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.