| SNP.to.cn(FASeg) | R Documentation |
A function that converts plain text outputs of probe level data to SNP-level copy number signal
SNP.to.cn(filenames, ref=ref, probe.loc=probe.loc)
filenames |
A list of filenames to import. Each file should be plain text output from Gtype. |
ref |
Reference data frame. It should contain a column named "SNP", which are SNP names, and another column named "ALL_MEAN", which are the standardized mean signal strength from reference data. Please see "Hind.ref.100K" etc. |
probe.loc |
Probe location data frame. It should contain three columns. The first column is SNP names; the second is chromosome number; the third is physical position. Please see "Hind.loc.100K" etc. |
This function performs calculation based on a simplified version of Huang et.al. Human Genomics 1(4):287-299. COpy number = exp (0.659 + 0.939 * (scaled log-intensity of this SNP - mean scaled log-intensity of reference data)) Instead of first finding the genotype of the SNP and comparing with only those reference of the same genotype, here we don't distinguish genotype. The reason is we found very high correlation between the overall mean and the mean of each genotype.
A matrix is returned. The first column is chromosome number; the second is physical location; the third is the SNP-level copy number of the sample corresponding to the first file name; the fourth correspond to the second file name, and so on.
The Gtype output file should be one sample per file. The first row of the file is a single phrase "PIW". If extra tabs exist in the same row, the function will fail. In that case please hand-erase those tabs in a text editor.
Tianwei Yu (tyu8@sph.emory.edu)
Huang, J., Wei, W., Zhang, J., Liu, G., Bignell, G.R., Stratton, M.R., Futreal, P.A., Wooster, R., Jones, K.W., & Shapero, M.H. (2004) Whole genome DNA copy number changes identified by high density oligonucleotide arrays. Hum Genomics, 1, 287-99.
#Need to have plain text output file from Gtype to use this function.
#cn<-SNP.to.cn(c("226.txt","047.txt","067.txt"),ref,probe.loc)
#fitted<-faseg(cn,no.change=2,sig=1e-10)