How to read Affymetrix CYCHP File

Question

[QVINTVS_FABIVS_MAXIMVS]

Howdy y'all,

I'm having a hard time intrepreting the cychp format.

Here's the affy documentation but it doesn't tell me what the symbols in the actual file mean. I have no idea http://media.affymetrix.com/support/developer/powertools/changelog/gcos-agcc/cychp.html

So has anyone worked with this .cychp file before? Is there an in house or 3rd party program that allows me to retrieve the data?

Tusen takk

[Istvan Albert]

this is not really a data exchange format, rather it is a special binary format, it feels unlikely that other people develop software for it.

Answer 1

[Christian]

You need to first convert the cychp file into a readable text format with Affymetrix Powertools and then you can parse it.

1. Convert to text

   apt-1.15.2-x86_64-intel-linux/bin/apt-chp-to-txt --chp-files samples.chpfile --out-dir .
   

   The file 'samples.chpfile' needs to be created first with a text editor and should contain the filenames of your .cychp files that you want to convert.

2. Parse text file to obtain log R ratios and allele peaks

   use warnings FATAL => qw( all );
   use strict;
   
   use Getopt::Long;
   
   my ($lrr_file, $baf_file, $sample_name);
   GetOptions
   (
    "lrr-file=s" => \$lrr_file,  
    "baf-file=s" => \$baf_file,  
    "sample-name=s" => \$sample_name,  
   );
   
   die "ERROR: --lrr-file not specified" if (!$lrr_file);
   die "ERROR: --baf-file not specified" if (!$baf_file);
   die "ERROR: --sample-name not specified" if (!$sample_name);
   
   open(LRR, ">$lrr_file") or die "ERROR: Could not write to file $lrr_file\n";
   open(BAF, ">$baf_file") or die "ERROR: Could not write to file $baf_file\n";
   
   print LRR "\tchrs\tpos\tLRR_$sample_name\n";
   print BAF "\tchrs\tpos\tBAF_$sample_name\n";
   
   while(<>)
   {
    last if (/^ProbeSetName\tChromosome\tPosition\tLog2Ratio\tWeightedLog2Ratio\tSmoothSignal/);
   }
   
   while(<>)
   {
    last if  (/^#/);
    my ($ProbeSetName, $Chromosome, $Position, $Log2Ratio, $tWeightedLog2Ratio, $tSmoothSignal) = split(/\t/);
    print LRR "$ProbeSetName\t$Chromosome\t$Position\t$Log2Ratio\n";
   }
   
   while(<>)
   {
    last if (/^ProbeSetName\tChromosome\tPosition\tAllelePeaks0\tAllelePeaks1/);
   }
   
   while(<>)
   {
    last if (/^#/);
    my ($ProbeSetName, $Chromosome, $Position, $AllelePeaks0, $AllelePeaks1) = split(/\t/);
    my $baf = ($AllelePeaks0 - 250)/(750-250);
    $baf = 0 if ($baf < 0);
    $baf = 1 if ($baf > 1);
    print BAF "$ProbeSetName\t$Chromosome\t$Position\t$baf\n";
   }
   
   while(<>) {}; # read pipe to the end to avoid SIGPIPE error status 141
   
   close(LRR);
   close(BAF);
   

[hongen.hugo.xu]

Dear Christian,

would you please explain the logic of the following line?

my $baf = ($AllelePeaks0 - 250)/(750-250);

thanks

[Christian]

This is just an ad-hoc heuristic I came up with to transform allele peaks into BAF values with range between 0 and 1.