Converting SNP datasets from A/B to ACGT format

Question

[User000]

Hello,

I have two sets of SNP data that were aligned, one in AGCT format and one in A/B format. For A/B format I know the alleles. However, there is difference in allele designation in two formats. Something like this:

marker1  chr  pos  alleles_set1  snp1_set1  snp2_set1  snp3_set1  alleles_set2  snp1_set2  snp2_set2  snp3_set2
m1       1    0    A/G           G          A          A          A/G           A          A          B
m2       1    0    A/G           A          A          G          T/C           A          B          B
m3       1    0    G/C           G          G          C          C/G           A          A          B

I need to produce a hapmap file to make association analyses.

So, my questions are: 1) How to change set2 from A/B format to AGCT? 2) When there is difference in alleles like in marker2, how to treat this data? Especially C/G and G/C?

Answer 1

[User000]

OK, I solved the problem.

Apparently if the possible SNP variation is A, then it is assigned as Allele A, while C and G as Allele B. Similar to rules of complementarity, if the possible SNP variation is T, then it is assigned as Allele A, while C and G as Allele B. So in m1 A is allele A, and G is allele B. In m2 since this variations are complementary we can change T/C to A/G follow the rules above and merge sets. In m3 we need to know the TOP/BOT information of the allelic variations. The algorithm used in this cased if called "sequence walking".

[dorcus-kholofelo.malomane]

Hi, how did you solve the problem.

I have run a file on plink in AB format and now trying to do imputation on beagle 4. but beagle4 requires that i have ACGT coding instead of B. I would like to convert the AB to ACGT. your help will be highly appreciated

[User000]

Hi, do you have an information about the alleles? if yes, you can apply the rules above but you will also need to know the TOP/BOT information of the allelic variations. I had to request an additional file from the company. Hope I was helpful. Here is a link you might find helpful as well: http://www.illumina.com/documents/products/technotes/technote_topbot.pdf

[dorcus-kholofelo.malomane]

Hi sorry for late reply, i ended up using the old beagle version which accept the AB format. thanks for the file, I already had it

[SheelS]

Hi Malomane, probably a question not related to the topic, but do you know for Beagle, if a parent has many offspring, how could we order / arrange the genotype input file for unphased trio data? Is it Male1, Female1, Off1, Male1, Female1, Off2 or simply as Male1, Female1, Off1, Off2, Off3..... ?? Thanks.

Answer 2

[dr.akilalshawi]

Hi every one

I have ped file with alleles coding as A B ,so I want to recode this ped file it as AGCT alleles . I used plink but not work with me to recode it to AGCT alleles. Could you please help here?

Thanks

Akil

[User000]

Hi, see the solution I posted above.