Skip to main content
. 2015 Jan 23;43(6):e39. doi: 10.1093/nar/gku1363

Figure 2.

Figure 2.

A flowchart outlining the procedures of CODEX in normalizing WES read depth and calling CNV. The first step is computing GC content, mappability and depth of coverage using Rsamtools with QC measures. The multi-sample normalization model by CODEX is then applied to remove biases and artifacts introduced by GC content, exon targeting and amplification efficiency and latent systemic artifacts. The Poisson likelihood-based segmentation algorithm gives final CNV calls with copy number estimates.