Calculate LogR ratio and BAF from X and Y intensities

Question

[dp0b]

Hi,

Does any one have a recommendation of a method or a program to convert X and Y intensities from a genotyping final report text file to their Log R ratio and B allele frequency estimates.

The LRR and BAF were not provided in the final report and I dont have access to Illumina Beadstudio. I was hoping to complete cnv analysis if possible

Any help appreciated.

Final report file format

[Header]
GSGT Version    1.9.4
Processing Date 6/19/2015 11:59 AM
Content     IDBv2_15051423_C.bpm
Num SNPs    16223
Total SNPs  18004
Num Samples 41
Total Samples   576
[Data]
SNP Name    Sample ID   Allele1 - Forward   Allele2 - Forward   Allele1 - AB    Allele2 - AB    Allele1 - Top   Allele2 - Top   GC Score    X   Y
ARS-BFGL-BAC-10972  1234    G   C   A   B   C   G   0.8651  0.726   0.905   
ARS-BFGL-BAC-10975  1235 A  A   A   A   A   A   0.9303  0.983   0.060   
ARS-BFGL-BAC-11025  1236 T  G   A   B   A   C   0.9092  0.434   0.684

[ivivek_ngs]

Show us some lines from your report. Usuallu beadstudio can perform CNV and LOH and give reports with LRR and BAF values with plots else show us what kind of information you posses and what you intend to perform and describe with a bit more detail your question then people can be more helpful.

[dp0b]

Apologies I have attached the final report format. I was hoping to complete CNV analysis.

[taghizadehkobra]

hi dp0b! i am recently work on snp like your study..i have some question of you.could you please give me a way for connection with you..?i need help you .thank you very much

Answer 1

[Irsan]

BAF = Y / (X + Y)
LRR = log2( (X + Y)sampleOfInterest / (X+Y)baselineSample)

But you might want to do some normalization first, possbily with beadstudio or with tQn normalization. See here http://bmcbioinformatics.biomedcentral.com/articles/10.1186/1471-2105-9-409

[dp0b]

Thank you very much for your help. Is there a recommendation for choosing the baseline sample?

[Irsan]

A sample for which you now the ploidy = 2. So any non-cancer sample would be fine. I think most of the time in a SNP-array run such a sample is included, for example a male and female reference DNA sample. It depends all on your research context but most of the time, best scenario is when you have a patient / family member matched control, because this way you can distinghuish between somatic / de novo / germ line copy number variants