How to get rid of "" object name and working in loop?

I do not know how to precisely define this problem; but I could not figure this out

    genx <- list(scaffold_1 = c("AAATTTTTATAT"),scaffold_2 = c("AAATTTTTATAT"),
    scaffold_3 = c("AAATTTTTATAT"),scaffold_4 = c("AAATTTTTATAT"),
    scaffold_5 = c("AAATTTTTATATA"),scaffold_6 = c("AAATTTTTATAT"),
scaffold_7 = c("AAATTTTTATAT"),scaffold_8 = c("AAATTTTTATATA"))


TATA = "TATA"   
      myobs <- paste("genx$scaffold_", 1:8, sep = "")

I want to apply the following function to every elments of the elements of myobs (are objects):

source("http://www.bioconductor.org/biocLite.R")

biocLite("Biostrings")
require((Biostrings)

countPattern (TATA, genx$scaffold_1, max.mismatch = 1)
[1] 3  

When I use the following:

countPattern (TATA, myobs[1], max.mismatch = 1)

Doesnot work as it is I believe interpreted as:

countPattern (TATA, "genx$scaffold_1", max.mismatch = 1)
 [1] 0 

Which is not same as the above one. How can get开发者_如何学编程 rid of "" and create a loop to perform this job, your suggestions are appreciated:

You can directly use the list and do an sapply which acts on every element of the list. Here is some sample code

sapply(genx, countPattern, pattern = TATA, max.mismatch = 1)

You can use the get function

a  <- 12
get("a") # returns 12

What about this:

sapply(genx, function (x) { countPattern(TATA, x, max.mismatch = 1) })

There is a problem with:

get("genx$scaffold_1")

because R thinks you are looking for an object with that full name, not the 'scaffold_1' component of 'genx'.

What should work is:

eval(parse(text="genx$scaffold_1"))

But see:

fortune(106)

More trouble spots along these lines can be found in 'The R Inferno': http://www.burns-stat.com/pages/Tutor/R_inferno.pdf

I would recommend that you define genx$scaffold as vector, and then you can easily use the apply function:

genx = data.frame(scaffold = c(1, 2, 3, 4)) # genx$scaffold is a vector

Now you can easily run the countPattern() function on every genx$scaffold's element

apply(as.matrix(genx$scaffold), 1, 
    function (x) { countPattern(TATA, x, max.mismatch = 1) })

You can easily extend the solution so that genx$scaffold can be also matrix etc.