Abstract
Through the identification of genomic runs of homozygosity (ROHs), homozygosity (autozygosity) mapping is a powerful approach to find autosomal recessive mutations, especially in consanguineous families. Many tools exists for
ROH discovery that rely on data from genotyping arrays or exome sequencing (ES). Since most of these tools had bad performances or were difficult to use, we have developed a new tool: AutoMap (Autozygosity Mapper). AutoMap
directly uses standard VCF files as primary source of data. A careful variant filtering step is implemented prior to the identification of ROHs by a sliding-window approach and subsequent filtering. The tool was first trained on ES data
from 26 samples from consanguineous families, and further validated in 26 additional samples. AutoMap displayed similar performance for both training and validation cohorts, with sensitivity of 90.5% and 92.4% respectively and showed the best sensitivity/specificity combination with respect to 8 existing tools. When applied to unpublished ES data, AutoMap allowed the discovery of ROHs containing homozygous non-coding variants that led to the identification of disease genes for five novel conditions including the Liberfarb syndrome. AutoMap is a reliable tool that can predict, in less than a minute, ROHs with high specificity and sensitivity by using a single VCF file. Our tool is available both via a web-based interface, for a quick analysis with default parameters, as well as a command-line package, allowing assessment of a large numbers of samples, the customization of many parameters and analysis of VCFs from whole-genome sequencing (WGS) experiments.
ROH discovery that rely on data from genotyping arrays or exome sequencing (ES). Since most of these tools had bad performances or were difficult to use, we have developed a new tool: AutoMap (Autozygosity Mapper). AutoMap
directly uses standard VCF files as primary source of data. A careful variant filtering step is implemented prior to the identification of ROHs by a sliding-window approach and subsequent filtering. The tool was first trained on ES data
from 26 samples from consanguineous families, and further validated in 26 additional samples. AutoMap displayed similar performance for both training and validation cohorts, with sensitivity of 90.5% and 92.4% respectively and showed the best sensitivity/specificity combination with respect to 8 existing tools. When applied to unpublished ES data, AutoMap allowed the discovery of ROHs containing homozygous non-coding variants that led to the identification of disease genes for five novel conditions including the Liberfarb syndrome. AutoMap is a reliable tool that can predict, in less than a minute, ROHs with high specificity and sensitivity by using a single VCF file. Our tool is available both via a web-based interface, for a quick analysis with default parameters, as well as a command-line package, allowing assessment of a large numbers of samples, the customization of many parameters and analysis of VCFs from whole-genome sequencing (WGS) experiments.
Original language | English |
---|---|
Pages (from-to) | 657-657 |
Number of pages | 1 |
Journal | European Journal of Human Genetics |
Volume | 28 |
Issue number | SUPPL 1 |
DOIs | |
Publication status | Published - Dec 2020 |