Development and evaluation of ActSeq: A targeted next-generation sequencing panel for clinical oncology use

PLoS One. 2022 Apr 21;17(4):e0266914. doi: 10.1371/journal.pone.0266914. eCollection 2022.

Abstract

Purpose: The demand for high-throughput genetic profiling of somatic mutations in cancer tissues is growing. We sought to establish a targeted next generation sequencing (NGS) panel test for clinical oncology practice.

Methods: Customized probes were designed to capture exonic regions of 141 genes selected for the panel, which was aimed for the detection of clinically actionable genetic variations in cancer, including KRAS, NRAS, BRAF, ALK, ROS1, KIT and EGFR. The size of entire targeted regions is 0.8 Mb. Library preparation used NEBNext Ultra II FS kit coupled with target enrichment. Paired-end sequencing was run on Illumina NextSeq 500 at a read length of 150 nt. A bioinformatics workflow focusing on single nucleotide variant and short insertions and deletions (SNV/indel) discovery was established using open source, in-house and commercial software tools. Standard reference DNA samples were used in testing the sensitivity and precision and limit of detection in variant calling.

Results: The general performance of the panel was observed in pilot runs. Average total reads per sample ranged from 30 million to 48 million, 73% ~82% unique reads. All runs had more than 99% average mapping rate. Mean target coverage ranged from 727x to 879x. Depth of coverage at 50x or more reached 87% of targeted region and 60% of targeted region received 500x or more coverage depth. Using OncoSpan HD827 DNA, which bears 144 variants (SNV/indel) from 80 genes that are within the targeted region on the panel, our somatic variant calling pipeline reached 97% sensitivity and 100% precision respectively, with near 48 million reads. High concordance with orthogonal approaches in variant detection was further verified with 7 cancer cell lines and 45 clinical specimens.

Conclusion: We developed a NGS panel with a focus on clinically actionable gene mutations and validated the performance in library construction, sequencing and variant calling. High concordance with reference materials and orthogonal mutation detection was observed.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Computational Biology
  • High-Throughput Nucleotide Sequencing
  • Humans
  • Medical Oncology
  • Mutation
  • Neoplasms* / genetics
  • Protein-Tyrosine Kinases* / genetics
  • Proto-Oncogene Proteins / genetics

Substances

  • Proto-Oncogene Proteins
  • Protein-Tyrosine Kinases