The I variable in the for loop changes

Question

[3095916029]

I set ‘n’ in advance, and then I do a for loop to extract the desired columns from a data box

n<-as.numeric(all_apa1$distal)

But, as we go through the loop, it's like I becomes 1,2,3

data1<-data.frame()
for(i in n){
data<-m6a_seq[which(as.numeric(m6a_seq$start) - i <200 & as.numeric(m6a_seq$start) - i >-200 ),]
data$position<-as.character(i)
data1<-rbind(data1, data)
}

So can someone please tell me how to deal with it?

Answer 1

[Michael]

I don't think you have given enough information for us to figure out if anything at all, not to mention what is wrong here. The only thing I can say is that R is rock-solid in guaranteeing that a variable will not change without an assignment to it. So, if i becomes 1, 2, 3, these were in the original vector n.

> for (i in c(10,1,42)) print(i)
[1] 10
[1] 1
[1] 42

The only reason you might get an unexpected result is, and that is very likely, if the original data was in fact a factor value:

n=as.numeric(as.factor(c(10,1,42)))
> n
[1] 2 1 3
> str(n)
 num [1:3] 2 1 3

That might seem unexpected, however, it is consistent and documented behavior.

A quick fix is to either:

n=as.numeric(as.character(as.factor(c(10,1,42))))
> n
[1] 10  1 42

Or better, define proper column classes while importing or turn off strings.as.factors.

Anyway, you seem to do some sort of interval overlap within a window of +/- 200bp around a genomic position. I highly recommend shifting your analysis to GenomicRanges to do this efficiently and correctly.

[3095916029]

Thank you for your reply. I hope to make a 200bp distribution density map of M6A above and below APA site. 'n' includes the location of APA site. M6a_seq is a Merip acquired bed file containing CHR,start,end, and value. I'm going to find every row in m6a_seq that is 200bp away from APA and do the following

[Michael]

Ok, please see my updated answer. I highly recommend doing this by the GenomicRanges package and its interval overlap methods if you want to perform the analysis in R. Also, when importing data, make sure numeric columns are imported as such and not converted to factors, which may happen sometimes.

[3095916029]

But, when i do this,

for ( i in n ) {print(i)}

I still get numeric in 'n', not others.

[Michael]

But that's what you wanted, isn't it?

[3095916029]

yes , i need it. but when i do the loop, it's still use 1, 2, 3, not these.

[Michael]

I don't see what could cause this because I have seen only fragments of your code. Anyway, simply scrap it and use GenomicRanges instead, then you need no loops.

[3095916029]

Hahaha, I think I need to try it. Thank you again, I am still a novice in bioinformatics, I still have a lot to learn, this is my first time to ask questions on this, thank you for giving me a good experience.

[Michael]

You are welcome, try stay on the well-trodden path (like using existing package), if you can, at least in the beginning. I think even seasoned programmers and bioinformaticians can and will make terrible mistakes when trying to code things from scratch. So, the best code is that that I didn't write at all.