Method to get all possible combinations of genotypes for a group of SNPs
1
0
Entering edit mode
3.2 years ago
Volka ▴ 160

Hi all,

I have with me now a group of about 20 SNPs that I would like to acquire all possible genotype combinations for. As an example, let's start off with three SNPs and their alleles.

SNP      A1         A2
SNP1      A          T
SNP2      C          G
SNP3      T          A


I want to start off by generating a list of all possible genotype permutations/combinations of these three SNPs, for example:

SNP1 SNP2 SNP3
AA   CC   TT
AA   CC   TA
AA   CC   AA
AA   CG   TT
AA   CG   TA
AA   CG   AA
AA   GG   TT
AA   GG   TA
AA   GG   AA
...


And so on, for what I expect to be 3^3 = 27 possible combinations.

From here, I hope to scale this up to my full group of ~20 SNPs. What is a good way of doing this, in Python or even in R?

combination genotype permutation python R • 1.1k views
2
Entering edit mode
3.2 years ago

Here is a R-based solution using expand.grid():

> SNP1
[1] "A" "T"
> SNP2
[1] "C" "G"
> SNP3
[1] "T" "A"
> expand.grid(SNP1_alleleA = SNP1, SNP1_alleleB = SNP1, SNP2_alleleA = SNP2, SNP2_alleleB = SNP2,
SNP3_alleleA = SNP3, SNP3_alleleB = SNP3)
SNP1_alleleA SNP1_alleleB SNP2_alleleA SNP2_alleleB SNP3_alleleA SNP3_alleleB
1             A            A            C            C            T            T
2             T            A            C            C            T            T
3             A            T            C            C            T            T
4             T            T            C            C            T            T
...
61            A            A            G            G            A            A
62            T            A            G            G            A            A
63            A            T            G            G            A            A
64            T            T            G            G            A            A


SNP1, SNP2 and SNP3 are character vectors of the possible polymorphisms for that particular SNP.

If you need to scale this up, it might be handy to use something along the lines of

> SNPs
[1] "SNP1" "SNP2" "SNP3"
> eval(parse(text = paste0("expand.grid(",
paste0(rep(SNPs, each = 2), c("_alleleA", "_alleleB"),
" = ", rep(SNPs, each = 2), collapse = ", "),
")")))


which gives you the exact same output, and where you can vary the amount of SNPs you include in your analysis by use of the SNPs variable. FYI, eval(parse(text = SOME_CHARACTER_STRING)) parses and evaluates the expression denoted by SOME_CHARACTER_STRING.

0
Entering edit mode

I have marked this as accepted answer, it looks like OP is already using this script. @Volka please attend to previous posts, upvote, and accept the answer or leave a comment. Opening a chain of new posts for the same problem is discouraged.