A comparison of algorithms for identifying copy number variants in family-based whole-exome sequencing data and its implications in inheritance pattern analysis

作者全名："Ye, Bo; Tang, Xia; Liao, Shixiu; Ding, Keyue"

作者地址："[Ye, Bo] Chongqing Med Univ, Sch Basic Med, Dept Bioinformat, Chongqing 400016, Peoples R China; [Tang, Xia] Fudan Univ, Ctr Genet & Dev, Sch Life Sci, State Key Lab Genet Engn & Collaborat Innovat, Shanghai 200438, Peoples R China; [Liao, Shixiu; Ding, Keyue] Henan Univ, Peoples Hosp Zhengzhou Univ, Henan Prov Peoples Hosp, Med Genet Inst Henan Prov,Henan Key Lab Genet Dis, Zhengzhou 450003, Henan, Peoples R China; [Ding, Keyue] Mayo Clin, Dept Cardiovasc Med, Rochester, MN 55905 USA"

通信作者："Liao, SX; Ding, KY (通讯作者)，Henan Univ, Peoples Hosp Zhengzhou Univ, Henan Prov Peoples Hosp, Med Genet Inst Henan Prov,Henan Key Lab Genet Dis, Zhengzhou 450003, Henan, Peoples R China.; Ding, KY (通讯作者)，Mayo Clin, Dept Cardiovasc Med, Rochester, MN 55905 USA."

来源：GENE

ESI学科分类：MOLECULAR BIOLOGY & GENETICS

WOS号：WOS:000950519400001

JCR分区：Q2

影响因子：2.6

年份：2023

卷号：861

期号：　

开始页：　

结束页：　

文献类型：Article

关键词：Germline copy number variants (gCNVs); Whole-exome sequencing; Algorithms; Performance

摘要："There remain challenges in accurately identifying constitutional or germline copy number variants (gCNVs) based on whole-exome sequencing data that have implications for genetic diagnosis for 'rare undiagnosed dis-ease' in the clinical setting. Although multiple algorithms have been proposed, a systematic comparison of these algorithms for calling gCNVs and analyzing inherited pattern have yet to be fully conducted. Therefore, we empirically compared seven exome-based algorithms, including XHMM, CLAMMS, CODEX2, ExomeDepth, DECoN, CN.MOPS, and GATK gCNV, for calling gCNVs in 151 individuals from 44 pedigrees, together with the gold standard of genotyping-derived gCNVs in the same cohort for the performance assessment. These algorithms demonstrated varied powers in identifying gCNVs, although the distribution of gCNVs size was similar. The number of shared gCNVs across these algorithms was limited (e.g., only four gCNVs shared among seven algo-rithms); however, several algorithms showed varying degrees of consistency (e.g., 1,843 gCNVs shared between DECoN and ExomeDepth). CLAMMS and CODEX2 outperformed the remaining algorithms according to a rela-tively higher F-score (i.e., 0.145 and 0.152, respectively). In addition, these algorithms exhibited different Mendelian inconsistencies of gCNVs and significant challenges remained in inheritance pattern analysis. In conclusion, selecting good algorithms may have important implications in gCNVs-based inheritance pattern analysis for family-based studies."

基金机构：National Natural Science Foundation of China [81650010]; Science and Technology Coop-eration Project of Henan Province [182106000058]

基金资助正文："Funding sources This work was supported by the National Natural Science Foundation of China [No. 81650010, S.L.] , and the Science and Technology Coop-eration Project of Henan Province [No.182106000058, S.L.] ."