a comprehensive target enrichment panel for fusion detection · for research use only. introduction...
TRANSCRIPT
www.nugen.comFor research use only.
I N T R O D U C T I O N
1NuGEN Technologies, 201 Industrial Road, San Carlos, CA 94070 USA. 2Avera Cancer Institute for Precision Oncology Genomics Laboratory, San Diego, CA, USA.
Ashesh A. Saraiya1, Brandon Young2, Tobias Meißner2, Brian L. Jones2, Stephanie C. Huelga1, Marie Edie1, and Doug A. Amorese1
A comprehensive target enrichment panel for fusion detection
®
C O N C L U S I O N S
VALIDATION OF WORKFLOW WITH CLINICAL FFPE SAMPLESFigure 5
Figure 5: Validation of the Ovation® Fusion Panel Target Enrichment System V2. A. Total DNA and RNA were isolated from clinical FFPE samples and were analyzed for genomic rearrangements and fusions using different methods. B. Representative results are shown. The whole transcriptomic analysis of sample CCD212 identified multiple fusions, including the GAB2-WNT11 fusion. Genomic rearrangements involving these genes were not identified using the commercial panel, potentially due to exclusion of these gene in the panel. The target enrichment method also detected the presence of the GAB2-WNT11 fusion in this sample, even with 50X less sequencing. Sample CCD201 showed no concordance between the genomic and transcriptomic analysis. The target enrichment method does not identify the fusion detected with the whole transcriptome analysis due to the absence of these genes in the target panel. Finally, sample CCD255 showed full concordance between the fusion identified during analysis of the whole transcriptome and genomic rearrangements. The expected fusion was also detected with the target enrichment method further validating this method as a alternative to whole transcriptome analysis for fusion detection.
A
・Integrated workflow with Ovation® Fusion Panel and NuFuseD was validated for use with FFPE sample clinical screening.
・Targeted sequencing provides similar results as whole transcriptome sequencing without the need for high sequencing depth.
・Panel targeting all exons from 502 genes known to be involved in gene fusion provides a comprehensive method for detection and discovery of known and novel gene fusions.
Figure 1
M E T H O D S
Figure 1: NuGEN’s Ovation® Fusion Panel Target Enrichment System workflow. A. This method consists of cDNA synthesis, adaptor ligation, probe hybridization and extension and library amplification. This protocol is completed over two half days. B. The full workflow includes library generation, sequencing and fusion analysis using NuGEN’s Ovation® Fusion detection software (NuFuseD).
cDNA synthesis
Adaptor ligation
Probe hybridization
Probe extension
Library amplification
©2017 NuGEN Technologies, Inc. All rights reserved. The Encore®, Ovation® and Applause® families of products and methods of their use are covered by several issued U.S. and International patents and pending applications (www.nugen.com). NuGEN, Ovation, SPIA, Ribo-SPIA, Applause, Encore, Prelude, Mondrian and Imagine More From Less are trademarks or registered trademarks of NuGEN Technologies, Inc. Other marks appearing in these materials are marks of their respective owners.
Sample Name
CCD212_fusion
CCD255_fusion
CCD201_fusion
SampleType
Breast
Breast
LymphNode
TargetedGenomic
Rearrangement Analysis
MIS18BP1−MIPOL1
FGFR3−TACC3 (F)
CDKN2A−CDKN2B−AS1 (T)
Ovation FusionPanel Target EnrichmentSystem V2
(~2 M reads)
Whole Transcriptome
Results (~100 M reads)
GAB2−WNT11AARB1−BCAS3PAK1−TENM4
KDM2A−MYO7APPFIA1−PRR11
GAB2−WNT11HIPK2−TRIM24
BRIP1 (R)TP53−TP53 (D)
NOTCH2−PHGDH (R)KDM5C−HUWE1 (R)WDR82−PBRM1
FGFR3−TACC3VCL−ZSWIM8GAN−PKD1L2BRAF−ENSA
ZFAND5−ALDH1A1MED16−GNB1
FRY−RXFP2PLB1−FOSL2
FAM63B−ADAM10UBE2G1−GGT6CCSER1−GRID2
GPR157−C1orf172PTPN13−AFF1
AKR1C1−AKT1E2EP300−EIF3B
CCNYL1−CPNE5
FGFR3−TACC3PTPN13−AFF1
MSI2−DCLRE1CGMDS−NUDT5FXBW7−ARFIP1DNAJC3−FOXP1
B
A Complete Solution for Fusion Discovery
Low inputs down to 10 ng of fresh frozen or FFPE total RNA
Simple, flexible sample prep
Total RNA
Ovation Fusion Panel Target Enrichment System
Sequencing
96 unique barcodes available for multiplexed sequencing
Ovation Fusion Detection Software Application
Simple click and go analysis available through BaseSpace or for download and local installation
ACTTCGGATGCTACACGATCCTGCCA
NuGEN Ovation Fusion Detection
Actionable Gene Fusion ResultsEvents prioritized by confidence P-values for follow-up
CCDC6 AGCGTGACCATCGAGGATCCAAAG RETCCDC6--RET p-value=1E-07
ACTTCGGATGCTACACGATCCTGCCA
A B
A C K N O W L E D G M E N T S
We’d like to thank Dr. Anand Patel for help with data analysis and Dr. Jennifer Finley for help with poster formatting and useful discussions.
Figure 3
C O M P L E T E , O P T I M I Z E D M E T H O D F O R F U S I O N D E T E C T I O N W I T H L E S S F A L S E P O S I T I V E S
Figure 3:. Simplified fusion detection. A. The NuGEN Ovation® Fusion Detection BaseSpace application provides a complete solution for fusion detection and performing all the necessary steps from demultiplexing to fusion detection. This pipeline is available as a local Linux installation. B. The app provides an analysis summary all samples in the run, with a graphical output of detected fusions, sequence, position, and a P-value. C. Higher level of false positive rate requiring more time consuming validation. Libraries were generated from 10 ng total RNA of an EML4-ALK fusion control (HorizonDX). The data was analyzed using both the NuGEN Ovation® Fusion Detection BaseSpace application and SOAPfuse. Both software applications identified the expected EML4-ALK fusion. SOAPfuse identified 30 additional fusions suggesting a high false positive rate which require extensive secondary analysis.
Ovation Fusion Detection App SOAPfuse
A
B
C
1
2
3
4
5
6
7
89
10
11
12
13
1415
1617
1819
2021
22
XY
ATP6AP1
RANBP17
KIAA1549
SLC29A1
NFATC2CAPNS1
CELSR3
ZNF703
RN
28S1
TRIP10TOP2B
STAT6
SPAG9
FUBP
3
DUS4L
SMTN
MAFK
ARN
T
MEN
1
GARS
SOX9
ERC1
PER1
PER2
JAG2
VIL2
ALK1
2
3
4
5
6
7
89
10
11
12
13
1415
1617
1819
2021
22
XY
SPAG9
BCL7A
MCM2
ALK
3 Detected Fusions High Rate of False Positives (30) Requiring
Extensive Validation
Demultiplexing Alignment and Fusion Detection
Summary report with p-value
Sample Demultiplexing
Probe Assignment
Targeted genomic rearrangement analysis using commercial panel
Whole transcriptome
analysis
Ovation Fusion Panel Target Enrichment System V2
Total
DNA Total RNA
Tota
l RN
A
Clinical FFPE sample less than 1 yr old
P R O B E D E S I G N F O R F U S I O N D E T E C T I O N
A
B
Figure 2: Schematic of probe design for fusion detection. A. Probes are designed to all exons in a targeted transcript. Multiple probes are designed to larger exons providing full coverage of the exon and the ability to detect potential fusions events that occur within exons as well as splice isoforms. B. Gene fusions can be detected by multiple probes. Multiple probes located adjacent to the fusion site (red arrows) or probes targeting near-by exons (green arrows) can provide evidence of a fusion event. Novel fusion events can be detected as long as one of the genes involved is targeted.
Figure 2
EXON 1 EXON 2 EXON 3
Gene Fusion (Gene A - Gene B)
Gene Fusion (Gene A - unknown gene)
EXON 1 EXON 2 EXON 5 EXON 6EXON 4
EXON 1 EXON 2 EXON 3EXON 2
Figure 4
DETECTION OF KNOWN AND NOVEL GENE FUSIONS
Figure 4: The Ovation® Fusion Panel Target Enrichment System enables targeted detection of known and novel fusion events. A. Using 100 ng of UHR total RNA, the known fusions in this sample (BCR-ABL and NUP214-XKR3) were easily detected as well as two novel fusions (FGF1R-WHSC1L1 and MLLT10-PICALM). The novel fusions were not detected using a untargeted RNA-Seq approach. B. Analysis of libraries constructed with 10 ng of HorizonDX CCDC6-RET fusion control RNA identified the expected fusion. A second high probability fusion EFNA5-IKBKB was also detected and validated with Sanger sequencing demonstrating the value of a comprehensive fusion panel for the detection of novel fusions.
0
50
100
150
200
250
300
3508.7 M Targeted Reads
125 M Untargeted Reads
1.6 M Targeted Reads
MLLT10-PICALMFGF1R-WHSC-1L1NUP-214-XKR3BCR-ABL
AGCAACCCCAGGCTGACCCACCCC
EFNA5 IKBKB
Junction File
(R1|R2)
Spanning File
(R1|R2)
S N P R L T H P In-Frame
CCDC6 RET
AGCGTGACCATCGAGGATCCAAAGS V T I E D P K In-Frame
Junction File Spanning File (R1|R2) (R1|R2)
Sample Name Sample Input Number of Reads Fusions Detected
CCDC6-RET control sample
(HorizonDX)10 ng 1913289
CCDC6−RETEFNA5−IKBKBBCL3−FAM84B
A
B
A G C A A C C C C A G G C T G A C C C A C C C C
60 70 80
Growing interest in cancer classification and progression has accelerated the rate of novel gene fusions discovery with increasing recognition of their roles as biomarkers. RNA-Seq is an attractive method for expressed fusion discovery and detection because of its ability to provide unbiased fusion sequencing information. The ability to detect low expressing fusion transcripts, however, require high sequencing depth and represents a significant financial barrier and identification of clinically relevant fusion sequences from a large data set can be a bioinformatics challenge.
To address these challenges we have tested the Ovation® Fusion Panel Target Enrichment System V2, a targeted RNA sequencing method using the Single Primer Enrichment Technology (SPET), with a number of control and clinical samples. Initial studies were performed using a comprehensive target enrichment panel targeting 502 genes with three samples from Horizon DX containing known fusions. Target enriched libraries were constructed with 10 ng and 100 ng inputs and the data was analyzed using the NuFuseD pipeline (available as a point and click BaseSpace application or downloadable Linux package) which has been optimized for fusion analysis from this data. Expected fusions were identified at both input levels, even when down sampled to 500K reads, with fewer fusion calls compared to other publicly available fusion detection software (Chimerascan and SOAPfuse), suggesting a lower false positive rate. NuFuseD fusion calls are provided with a P-value to help prioritize the identified fusions for subsequent validation. Additionally, NuFuseD detected novel fusions in the control samples demonstrating the advantage of a comprehensive panel compared to more restricted panels. We further validated the target panel using control RNA (UHR and Human Brain) and fresh or FFPE cell lines (NCI-H2228, HCC1937) to further demonstrate our ability to identify known fusions.
Finally, the system was evaluated at an external site using patient FFPE samples. These samples (N=8) were from a set of breast, liver and ovarian cancers, containing a unique fusion in 4 of the samples based on DNA based sequencing. Only 1 of the 4 expected fusions were identified using whole transcriptome data (100 million reads) while 3 of the 4 fusions were detected with this assay (10 million reads) demonstrating its ability to generate targeted RNA sequencing libraries with increased sensitivity of gene fusion detection and reduced sequencing costs compared to standard RNA-Seq methods. Integrated workflow with Ovation® Fusion Panel and NuFuseD was validated for use with FFPE sample clinical screening.