TY - JOUR
T1 - Chromosomal Junction Detection from Whole-Genome Sequencing on Formalin-Fixed, Paraffin-Embedded Tumors
AU - Murphy, Stephen
AU - Smadbeck, James
AU - Eckloff, Bruce
AU - Lee, Yean
AU - Johnson, Sarah
AU - Karagouga, Giannoula
AU - Serla, Vishnu
AU - Sharma, Anurag
AU - Sikkink, Robert
AU - Voss, Jesse
AU - Harris, Faye
AU - Kline, Janet Schaefer
AU - Kosari, Farhad
AU - Feldman, Andrew
AU - Wieben, Eric
AU - Aubry, Marie Christine
AU - Kipp, Benjamin
AU - Jen, Jin
AU - Cheville, John
AU - Vasmatzis, George
N1 - Funding Information:
Supported by the Center for Individualized Medicine, the Biomarker Discovery Program and Genome Research Facility, and the Department of Laboratory Medicine and Pathology, Mayo Clinic .
Publisher Copyright:
© 2021 Association for Molecular Pathology and American Society for Investigative Pathology
PY - 2021/4
Y1 - 2021/4
N2 - DNA junctions (DNAJs) frequently impact clinically relevant genes in tumors and are important for diagnostic and therapeutic purposes. Although routinely screened through fluorescence in situ hybridization assays, such testing only allows the interrogation of single-gene regions or known fusion partners. Comprehensive assessment of DNAJs present across the entire genome can only be determined from whole-genome sequencing. Structural variance analysis from whole-genome paired-end sequencing data is, however, frequently restricted to copy number changes without DNAJ detection. Through optimized whole-genome sequencing and specialized bioinformatics algorithms, complete structural variance analysis is reported, including DNAJs, from formalin-fixed DNA. Selective library assembly from larger fragments (>500 bp) and economical sequencing depths (300 to 400 million reads) provide representative genomic coverage profiles and increased allelic coverage to levels compatible with DNAJ calling (40× to 60×). Although consistently fragmented, more recently formalin-fixed, specimens (<2 years’ storage) revealed consistent populations of larger DNA fragments. Optimized bioinformatics efficiently detected >90% of DNAJs in two prostate tumors (approximately 60% tumor) previously analyzed by mate-pair sequencing on fresh frozen tissue, with evidence of at least one spanning-read in 99% of DNAJs. Rigorous masking with data from unrelated formalin-fixed tissue progressively eliminated many false-positive DNAJs, without loss of true positives, resulting in low numbers of false-positive passing current filters. This methodology enables more comprehensive clinical genomics testing on formalin-fixed clinical specimens.
AB - DNA junctions (DNAJs) frequently impact clinically relevant genes in tumors and are important for diagnostic and therapeutic purposes. Although routinely screened through fluorescence in situ hybridization assays, such testing only allows the interrogation of single-gene regions or known fusion partners. Comprehensive assessment of DNAJs present across the entire genome can only be determined from whole-genome sequencing. Structural variance analysis from whole-genome paired-end sequencing data is, however, frequently restricted to copy number changes without DNAJ detection. Through optimized whole-genome sequencing and specialized bioinformatics algorithms, complete structural variance analysis is reported, including DNAJs, from formalin-fixed DNA. Selective library assembly from larger fragments (>500 bp) and economical sequencing depths (300 to 400 million reads) provide representative genomic coverage profiles and increased allelic coverage to levels compatible with DNAJ calling (40× to 60×). Although consistently fragmented, more recently formalin-fixed, specimens (<2 years’ storage) revealed consistent populations of larger DNA fragments. Optimized bioinformatics efficiently detected >90% of DNAJs in two prostate tumors (approximately 60% tumor) previously analyzed by mate-pair sequencing on fresh frozen tissue, with evidence of at least one spanning-read in 99% of DNAJs. Rigorous masking with data from unrelated formalin-fixed tissue progressively eliminated many false-positive DNAJs, without loss of true positives, resulting in low numbers of false-positive passing current filters. This methodology enables more comprehensive clinical genomics testing on formalin-fixed clinical specimens.
UR - http://www.scopus.com/inward/record.url?scp=85103045040&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85103045040&partnerID=8YFLogxK
U2 - 10.1016/j.jmoldx.2020.12.006
DO - 10.1016/j.jmoldx.2020.12.006
M3 - Article
C2 - 33387698
AN - SCOPUS:85103045040
SN - 1525-1578
VL - 23
SP - 375
EP - 388
JO - Journal of Molecular Diagnostics
JF - Journal of Molecular Diagnostics
IS - 4
ER -