genomic_wave.pl sub trainGCModel subsets lrr values differently to method described in paper

[joshuamschmidt]

I note that this conditional on line 177 of genomic_wave.pl :

if ($data[$i]->[3] > 15 and $data[$i]->[3] < 80 and $data[$i]->[1] > -1 and $data[$i]->[1] < 1)

shouldn't this be:

if ($data[$i]->[3] > 15 and $data[$i]->[3] < 80 and $data[$i]->[1] > -2 and $data[$i]->[1] < 1)

This would be in keeping with method as described in Diskin et. al.:

"To reduce the effect of markers within CNV regions on the regression coefficients, we restricted the analysis to markers with LRR between −2 and 1. "

and also line 186 of the same module:

if ($data[$i]->[3] > 15 and $data[$i]->[3] < 80 and $data[$i]->[1] > -2)

although it is unclear why this line drops the $data[$i]->[1] < 1

[joshuamschmidt]

Just wondering if there is potential for any clarification on this. Thank you.