Renault Victor, Tost Jörg, Pichon Fabien, Wang-Renault Shu-Fang, Letouzé Eric, Imbeaud Sandrine, Zucman-Rossi Jessica, Deleuze Jean-François, How-Kit Alexandre
Laboratory for Bioinformatics, Fondation Jean Dausset-CEPH, Paris, France.
Laboratory of Excellence GenMed, Paris, France.
PLoS One. 2017 Dec 19;12(12):e0189334. doi: 10.1371/journal.pone.0189334. eCollection 2017.
Copy number variations (CNV) include net gains or losses of part or whole chromosomal regions. They differ from copy neutral loss of heterozygosity (cn-LOH) events which do not induce any net change in the copy number and are often associated with uniparental disomy. These phenomena have long been reported to be associated with diseases and particularly in cancer. Losses/gains of genomic regions are often correlated with lower/higher gene expression. On the other hand, loss of heterozygosity (LOH) and cn-LOH are common events in cancer and may be associated with the loss of a functional tumor suppressor gene. Therefore, identifying recurrent CNV and cn-LOH events can be important as they may highlight common biological components and give insights into the development or mechanisms of a disease. However, no currently available tools allow a comprehensive whole-genome visualization of recurrent CNVs and cn-LOH in groups of samples providing absolute quantification of the aberrations leading to the loss of potentially important information.
To overcome these limitations, we developed aCNViewer (Absolute CNV Viewer), a visualization tool for absolute CNVs and cn-LOH across a group of samples. aCNViewer proposes three graphical representations: dendrograms, bi-dimensional heatmaps showing chromosomal regions sharing similar abnormality patterns, and quantitative stacked histograms facilitating the identification of recurrent absolute CNVs and cn-LOH. We illustrated aCNViewer using publically available hepatocellular carcinomas (HCCs) Affymetrix SNP Array data (Fig 1A). Regions 1q and 8q present a similar percentage of total gains but significantly different copy number gain categories (p-value of 0.0103 with a Fisher exact test), validated by another cohort of HCCs (p-value of 5.6e-7) (Fig 2B).
aCNViewer is implemented in python and R and is available with a GNU GPLv3 license on GitHub https://github.com/FJD-CEPH/aCNViewer and Docker https://hub.docker.com/r/fjdceph/acnviewer/.
拷贝数变异(CNV)包括部分或整个染色体区域的净增加或减少。它们不同于杂合性拷贝中性缺失(cn-LOH)事件,后者不会引起拷贝数的任何净变化,且通常与单亲二体性相关。长期以来,这些现象一直被报道与疾病相关,尤其是癌症。基因组区域的缺失/增加通常与较低/较高的基因表达相关。另一方面,杂合性缺失(LOH)和cn-LOH是癌症中的常见事件,可能与功能性肿瘤抑制基因的缺失有关。因此,识别复发性CNV和cn-LOH事件可能很重要,因为它们可能突出常见的生物学成分,并深入了解疾病的发展或机制。然而,目前没有可用的工具能够对样本组中的复发性CNV和cn-LOH进行全面的全基因组可视化,从而无法对导致潜在重要信息丢失的畸变进行绝对定量。
为克服这些限制,我们开发了aCNViewer(绝对CNV查看器),这是一种用于在一组样本中对绝对CNV和cn-LOH进行可视化的工具。aCNViewer提供了三种图形表示:树状图、显示具有相似异常模式的染色体区域的二维热图,以及便于识别复发性绝对CNV和cn-LOH的定量堆叠直方图。我们使用公开可用的肝细胞癌(HCC)Affymetrix SNP阵列数据对aCNViewer进行了说明(图1A)。1q和8q区域的总增加百分比相似,但拷贝数增加类别显著不同(Fisher精确检验的p值为0.0103),另一组HCC验证了这一点(p值为5.6e-7)(图2B)。
aCNViewer用Python和R实现,可在GitHub(https://github.com/FJD-CEPH/aCNViewer)和Docker(https://hub.docker.com/r/fjdceph/acnviewer/)上以GNU GPLv3许可获得。