Hui Daniel, Fang Zhou, Lin Jerome, Duan Qing, Li Yun, Hu Ming, Chen Wei
Department of Computer Science, University of Pittsburgh, Pittsburgh, PA, 15213, USA.
Department of Biostatistics, University of Pittsburgh, Pittsburgh, PA, 15213, USA.
BMC Genet. 2017 Sep 6;18(1):83. doi: 10.1186/s12863-017-0546-y.
Inferring local ancestry in individuals of mixed ancestry has many applications, most notably in identifying disease-susceptible loci that vary among different ethnic groups. Many software packages are available for inferring local ancestry in admixed individuals. However, most of these existing software packages require specific formatted input files and generate output files in various types, yielding practical inconvenience.
We developed a tool set, Local Ancestry Inference Toolkit (LAIT), which can convert standardized files into software-specific input file formats as well as standardize and summarize inference results for four popular local ancestry inference software: HAPMIX, LAMP, LAMP-LD, and ELAI. We tested LAIT using both simulated and real data sets and demonstrated that LAIT provides convenience to run multiple local ancestry inference software. In addition, we evaluated the performance of local ancestry software among different supported software packages, mainly focusing on inference accuracy and computational resources used.
We provided a toolkit to facilitate the use of local ancestry inference software, especially for users with limited bioinformatics background.
推断混合血统个体的本地祖先有许多应用，最显著的是在识别不同种族群体中存在差异的疾病易感基因座方面。有许多软件包可用于推断混合个体的本地祖先。然而，这些现有的软件包大多需要特定格式的输入文件，并生成各种类型的输出文件，带来了实际的不便。
我们开发了一个工具集，即本地祖先推断工具包（LAIT），它可以将标准化文件转换为特定软件的输入文件格式，并对四种流行的本地祖先推断软件（HAPMIX、LAMP、LAMP-LD和ELAI）的推断结果进行标准化和汇总。我们使用模拟数据集和真实数据集对LAIT进行了测试，并证明LAIT为运行多个本地祖先推断软件提供了便利。此外，我们评估了不同支持软件包中本地祖先软件的性能，主要关注推断准确性和所使用的计算资源。
我们提供了一个工具包，以促进本地祖先推断软件的使用，特别是对于生物信息学背景有限的用户。
