CODEX

CODEX is a software tool used for analyzing genetic variations in humans. It uses high-throughput sequencing of DNA coding regions to identify copy number variations (CNV). However, detecting and characterizing CNV from exome sequencing can be difficult due to various biases and artifacts. To address this challenge, CODEX proposes a normalization and CNV calling procedure that employs a Poisson latent factor model to remove specific biases such as GC content, exon capture and amplification efficiency, and latent systemic artifacts. Additionally, it includes a recursive segmentation procedure that explicitly models count-based exome sequencing data. Compared to other methods, CODEX shows improved accuracy on several validation datasets and is able to effectively detect CNVs with complex structures in a neuroblastoma sample.