|
|
Segmentation
SENECA is a count-based method to assign copy number state. It compares coverage between tumor and normal samples. Specifically, it bins read coverage using nonoverlapping 1 kb windows to derive counts in tumor and normal samples, and it then takes the ratio of the 2 counts. Bins are skipped during segmentation when they overlap low alignability regions in more than 20% of their size.
Independently, SENECA calculates B allele ratios at dbSNP positions from a tumor BAM file, and it keeps only SNVs that are heterozygous in the corresponding normal sample. Segmentation is carried out independently for copy number and B allele ratios.