Premium
Systematic comparative study of computational methods for HLA typing from next‐generation sequencing
Author(s) -
Yu Yuechun,
Wang Ke,
Fahira Aamir,
Yang Qiangzhen,
Sun Renliang,
Li Zhiqiang,
Wang Zhuo,
Shi Yongyong
Publication year - 2021
Publication title -
hla
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.347
H-Index - 99
eISSN - 2059-2310
pISSN - 2059-2302
DOI - 10.1111/tan.14244
Subject(s) - human leukocyte antigen , typing , dna sequencing , exome sequencing , computational biology , in silico , computer science , genotyping , biology , data mining , genetics , gene , antigen , genotype , mutation
The human leukocyte antigen (HLA) system plays an important role in hematopoietic stem cell transplantation (HSCT) and organ transplantations, immune disorders as well as oncological immunotherapy. However, HLA typing remains a challenging task due to the high level of polymorphism and homology among HLA genes. Based on the high‐throughput next‐generation sequencing data, new HLA typing algorithms and software tools were developed. But there is still a deficit of systematic comparative studies to assist in the selection of the optimal analytical approaches under different conditions. Here, we present a detailed comparison of eight software tools for HLA typing on different real datasets (whole‐genome sequencing, whole‐exome sequencing and transcriptomic sequencing data) and in‐silico samples with different sequencing lengths, depths, and error rates. We figure out the algorithms with the best efficiency in different scenarios, and demonstrate the effect of different raw reads on analytical performances. Our results provide a comprehensive picture of specifications and performances of the eight existing HLA genotyping algorithms, which could assist researchers in selecting the most appropriate tool for specific raw datasets.