Entering edit mode

2.8 years ago

always_learning
★
1.1k

Hi,

I have Allele Frequenct frequency data for Population A and I wanted to compare this with Population B. Below is the script that I wrote for Chi Square test between AF for these two population.

```
library("dplyr")
input = fread("GENE_PANEL_MUTATION.tsv", sep = "\t") inputTtest = filter(input, !complete.cases(gnomadWGS_AF, new_AF))
inputTtest = input %>% select(gnomadWGS_AF,new_AF) %>%
filter(complete.cases("NA")) %>%
na.omit() %>%
filter(!is.na(as.numeric(gnomadWGS_AF)))
inputTtest$gnomadWGS_AF=as.numeric(inputTtest$gnomadWGS_AF) inputTtest$new_AF=as.numeric(inputTtest$new_AF)
chisq <- function(a,b){ data <- matrix(c(a,b),ncol=2) mode(data) = "numeric" print (data) c(p = chisq.test(data)$p.value,
OR = chisq.test(data)$estimate) }
inputTtest %>% rowwise() %>% mutate(p=chisq(gnomadWGS_AF,new_AF)[[1]])
```

I am not getting any significant P-value while I can see that I have SNP's where frequencies are significantly different between both populations. Can someone help me with this? Which test I should apply?

Please provide example data. You might be inspired by How to ask good questions in technical and scientifc forums.

What is not clear in my Question?

Please find dropbox link for example file. https://www.dropbox.com/s/1gjnoh1ft6k0aat/example.txt?dl=0

Please find dropbox link for example data.

https://www.dropbox.com/s/1gjnoh1ft6k0aat/example.txt?dl=0

Do the Chi Square test. While you're at it, calculate the OR, confidence intervals, and Z-score, too: A: SNP dataset and Z Score

Thanks !! But I have just frequencies !! So I don't think that I can do similar to that.