Illumina Trusight Myeloid Panel validation A R FHAN R A FIQ G E NETIC T E CHNOLOGIST MEDICAL G E NETICS, CARDIFF
To Cover Background to the project Choice of panel Validation process Genes on panel, Protocol LOD analysis Analysis and Reporting Future impact
Background Currently, the laboratory is using several separate assays to detect a limited number of mutations in genes involved in the development of leukaemia. We are currently unable to test for all clinically relevant mutations using our current testing methodology due to financial and staffing constraints. The challenge the laboratory faces is to develop a single, cost effective test to detect gene mutations in myeloid leukaemia. The implementation of next generation sequencing would enable us to expand the number of genes we test. This will allow the detection of all the clinically relevant mutations in a single test in a cost efficient manner.
Gene Panel Specification Illumina 54 genes & 568 amplicons on panel Compatible with MiniSeq, MiSeq, NextSeq 500 DNA (not FFPE compatible) Read length 2 x 250 bp. ~48hrs sequencing Thermo Fisher 69 genes in panel. Targeted NGS and whole genome copy number change. DNA based. Ion torrent.
Validation 50,000 ETTF Wales (Efficiency through Technology) Association of Clinical Genomic Science Guidelines 60 variants (=mutations) Known abnormal samples JAK2, CALR, NPM1 Leeds laboratory Horizon controls Limit of detection Sensitivity Specificity Failure rate Clinical Utility Technical workflow Bioinformatic processing Interpretation of results Report format
Diseases AML High risk MDS MPN PRV s likely to require JAK2 only, ET & MF likely to require full panel. CLL TP53 and NOTCH1
Genes included Fusion Driver Expression Other Hotspot genes genes Full genes gene JAK3 ABL1 IDH2 SETBP1 MLL ASXL1 SMC1A KDM6A BRAF JAK2 SF3B1 PDGFRA BCOR ATRX CBL KIT SRSF2 CEBPA CUX1 CSF3R KRAS U2AF1 ETV6/TEL BCORL1 DNMT3A MPL WT1 EZH2 FBXW7 FLT3 MYD88 CALR IKZF1 GATA1 GATA2 NPM1 NOTCH1 PHF6 GNAS HRAS NRAS RUNX1 CBLB IDH1 PTPN11 STAG2 CBLC CDKN2A PTEN RAD21 SMC3 TET2 TP53 ZRSR2
Which genes/exons to analyse Asked UHW Haematology consultants which genes they would like on the panel. Online database search - Mycancergenome - COSMIC - Journals Decided to do whole gene sequencing on some genes i.e. TP53 exon 2-11 Hotspot i.e. NOTCH1, JAK2 V617F Whole exon CALR exon 9 Used Alamut to get genomic coordinates and decide which ref seq to use for Bioinformatics pipeline. 2 confirmed 25536776 using 25536858 DNMT3A Mutalyzer exon 2 2 25523004 25523117 DNMT3A exon 3 2 25505304 25505585 DNMT3A exon 4 2 25498366 25498417 DNMT3A exon 5 2 25497804 25497961 DNMT3A exon 6 2 25475057 25475071 DNMT3A exon 6A 2 25472522 25472598 DNMT3A exon 6B Exon 23 DNMT3A NM_022552.4 NM_022552.4:c.*3 Exon 23 DNMT3A NM_022552.4 NM_022552.4:c.*2 Exon 23 DNMT3A NM_022552.4 NM_022552.4:c.*1 Exon 23 DNMT3A NM_022552.4 NM_022552.4:c.2739 Exon 23 DNMT3A NM_022552.4 NM_022552.4:c.2738 Exon 23 DNMT3A NM_022552.4 NM_022552.4:c.2737
Myeloid NGS Service Targeted panel Illumina TruSight Myeloid Sequencing Panel Sequenced on the Illumina miseq ASXL1 (exon 12), BCOR (exons 3-16), CALR (exon 9), CBL (exons 8 and 9), CEBPA (exon 1), DNMT3A (exons 2 to 23), ETV6 (exons 1 to 8), EZH2 (exons 2 to 20), FLT3 (exons 13, 14, 15 and 20), GATA2 (exons 2-6), IDH1 (codon 132 within exon 4), IDH2 (codon 172 within exon 4), JAK2 (exon 12 and codon 617 within exon 14), KIT (exon 17), KRAS (exons 2 and 3), MPL (exon 10), NOTCH1 (exon 34), NPM1 (exon 11), NRAS (exons 12 and 13), PDGFRA (exons 12, 14 and 18), RUNX1 (exons 2 to 10), SETB1 (exon 4), SF3B1 (exons 13, 14, 15 and 16), SRSF2 (exon 1), TET2 (exons 3 to 11), TP53 (exons 2 to 11), U2AF1 (exons 2 and 6)
Protocol D AY 1 Sample dilution to 5 ng/ul Hybridise oligo pool Remove unbound oligos Extend and ligate bound oligos D AY 2 Clean up library with AMPure Normalize library (beads based) Pool libraries Loading on miseq (loading volume 6-13 µl)
Limit of Detection (LOD) LOD analysis carried out on Horizon controls Take random 10% of reads Analysis to see how many reads have the variant Gene Variant Expected Allelic Frequency Reported AF Reported Depth Reported LOD Qual Subsample Depth BRAF V600E 8.00% 7.40% 10283 100 1% 92-112 BRAF V600K 4.00% 3.41% 10283 100 10% 980-1057 EGFR G719S 16.70% NaN NaN NaN NaN NaN EGFR T790M 4.20% NaN NaN NaN NaN NaN FLT3 ΔI836 5.00% 4.15% 10082 100 5% 463-526 IDH1 R132C 5.00% 5.75% 5498 100 5% 234-296 JAK2 V617F 5.00% 11.34% 2699 100 5% 119-151 KRAS G12A 5.00% 3.88% 9576 100 5% 446-500 KRAS G12R 5.00% 3.77% 9599 100 5% 446-514 KRAS G13D 25.00% 24.77% 9592 100 1% 85-105 MEK1 P124L 5.00% NaN NaN NaN NaN NaN NOTC L1600 4.80% H1 P 6.36% 3394 100 5% 150-191 NRAS Q61K 5.00% 4.31% 1416 100 25% 334-379 PIK3C A H1047 R 30.00% NaN NaN NaN NaN NaN
Analysis Created an Excel macro which consisted of various tabs that would breakdown the data into a manageable size. Tabs were: Patient Demographics Depth Compare depth of patient vs NTC and depth > 500 for chosen location NTC Variant Raw VCF Sorted VCF Filter out calls less than 5% Variant calls Categorise variants as Genuine, Artifact or Known poly s. Number of reads Quality score of that base Position in the genome Some regions are at high risk of sequencing errors/artefacts Mutation & SNP s Screen Gaps Identify which gene/ exon the gap relates to. No gap filling. Report Summary of all above.
IGV Trusight Cancer Myeloid
Variant Interpretation Two stages Decide if variant is genuine If genuine decide if actionable and therefore reportable ACGS germline guidelines 2017 No national guidelines for somatic variant interpretation National working group AMP guidelines Local guidelines Consistent with other labs and working group Ensures consistency Haematological and solid tumour
Reporting Genes and regions of genes analysed will be stated Actionable variants only Variants will be fully interpreted in the context of the clinical information received May require more clinical information Some regions may not have sufficient sequencing coverage Vary by sample Stated in Basis of test Possible to repeat on fresh sample if crucial
Limitations NGS not suitable for monitoring disease Can t identify copy number changes, gene fusions or large deletions and duplications eg FLT3 Sensitivity of 5% Not always 100% coverage Interpretation can be difficult
Future Genomics England Test Directory Uplift in funding across laboratory NovoSeq Expand myeloid panel - > haematological panel Some to this panel eg CLL TP53 Tender for new panel Run weekly Increase throughput Reduce reporting time Aim 10 working day Gene Fusions Panel for FFPE samples Lymphoma Shorter DNA fragments Electronic reporting