A Guide to NGS Applications Using cfDNA SamplesOverview__Liquid+Biopsy... · Introducing the Swift...
Transcript of A Guide to NGS Applications Using cfDNA SamplesOverview__Liquid+Biopsy... · Introducing the Swift...
A Guide to NGS Applications Using
cfDNA Samples
Swift Snapshot
• Swift Biosciences
– Develops and commercializes NGS
library kits
– Compatible with all NGS platforms
• Illumina®
• Ion Torrent™
• Pacific Biosciences
– Developed on broad IP portfolio
• Leader in molecular biology
– Fundamentally different from all commercially available kits
– Enables better performance across an expanding set of applications
• Whole genome sequencing and hybridization enrichment
• Bisulfite sequencing (WGBS)
• Amplicon sequencing for screening applications
Downtown Ann Arbor, MI
Introducing the Swift Portfolio for cfDNA
Product Application Technical Advantage Benefits
Accel-NGS® 2S Liquid biopsy for hyb
capture or WGS
• 90% library conversion rate for
cfDNA
PCR-free input down
to 10 ng
Accel-
Amplicon™
Liquid biopsy for somatic
mutations
• Single tube assay (< 2 hrs.)
• > 100s of primer pairs in one
tube
10 ng input for 1-5%
mutation detection
Accel-NGS
Methyl-Seq
Methylation sequencing
using bisulfite-converted
DNA
• Generates a library post-
bisulfite treatment
• 10-100X higher recovery
Sensitive detection
of hypomethylation
in cancer
Accel-NGS 1S
Plus
Compatible with single-
stranded DNA
• Allows the detection of
damaged or nicked cfDNA
Conversion of
ssDNA and dsDNA
Outline
• Introduction to cfDNA
– Application of liquid biopsy
– cfDNA extraction from plasma
– Low DNA input requirements
– Concentration and integrity of cfDNA
• Accel-NGS® 2S DNA Library Kits
– 90% library conversion for cfDNA samples
– PCR-free library prep from 10 ng of cfDNA
– Compatibility with multiple hybridization capture technologies
• Accel-Amplicon™ Panels
– Performance with Horizon’s Quantitative Multiplex Reference Standard
– Liquid biopsy samples experimental design
• Accel-NGS Methyl-Seq DNA Library Kits
– Out performs other kits
– Detecting genome-wide hypomethylation
Introduction to cfDNA
Bettegowda, C., et al Sci. Trans. Med., (2014) vol6
Limitations for Using cfDNA to Detect Cancer
Burden
• Over 75% of samples had
detectable cancer mutations from
17 different cancer types
• Lower levels of detection correlated
with tumor types that had some type
of physical barrier (blood brain)
• Frequency of detection of cancer
burden within cfDNA increased with
disease severity
Stage I II III IV
Bettegowda, C., et al Sci. Trans. Med., (2014) vol6
cfDNA Extraction from Plasma
• Validated blood collection into Streck Cell-Free DNA BCT® that preserves
cfDNA and prevents lymphocyte lysis
• Validated manual extraction using Qiagen QIAamp® circulating nucleic acid kit
and automated extraction using Perkin Elmer chemagicTM system
Sample mL of Plasma ng/mL of Plasma
1 2.5 6.3
2 5.0 4.3
3 3.8 4.4
4 4.0 10.5
5 3.0 6.7
6 4.5 7.1
7 4.5 2.6
8 4.5 2.9
Normal (5) Avg. 3.5 Avg. 3.5
Average results obtained from manual extraction
Low DNA Input Requires a High Conversion
Rate
Conversion Rates
Library Molecules
The number of genome copies available for detection by sequencing is impacted
by low input DNA quantities and the conversion rate of DNA fragments
constructed into library molecules.
cfDNA inputs range from 1-30 ng input
1 ng = 330 genome copies
30 ng = 9,900 genome copies
Accel-NGS® 2S Kits can achieve 90% library conversion for
cfDNA samples.
Achieving a 1% Limit of Detection from Low
DNA Input
cfDNA inputs range from 1-30 ng extracted from 10 mL blood draw
1 ng = 330 genome copies
1% allele frequency = 3 mutant copies
10 ng = 3,300 genome copies
1% allele frequency = 33 mutant copies
30 ng = 9,900 genome copies
1% allele frequency = 99 mutant copies
Swift’s Accel-NGS® 2S Kits and Accel-Amplicon™ Panels can
achieve 1% mutant allele detection from 10 ng cfDNA input.
Concentration and Integrity of cfDNA
• cfDNA is ~ 165 bp in size
• High molecular weight cellular DNA can contaminate cfDNA samples
• Swift library kits include primers for a qPCR assay designed for Alu
repeat sequences from human gDNA at two different sizes
For more information, see our Instructional Manual: “Quantification and
Quality Assessment of Human DNA Samples”
qPCR quantification
(Alu Amplicons)
115 bp
247 bp
Total DNA Concentration
HMW Cellular DNA Concentration
(cfDNA + HMW Cellular DNA Contamination)
DNA Integrity Index = Alu247/Alu115
T.B. Hao in the British Journal of Cancer (2014) 111, 1482-1489
Accel-NGS® 2S DNA
Library Kits
Technology for dsDNA
Accel-NGS® 2S Plus Library Kit:How Improved Performance Is Achieved
Dephosphorylation of both the 3’ and 5’ termini
• Reduces chimera formation
• Improves ligation
Repair both the 3’ and 5’ bases
• Eliminating any damaged bases
Sequential ligation of P7 and P5 adapters
prevents adapter dimer formation
• Eliminates need for adapter titration
• P7 adapter is attached to the repaired 3’ termini
• P5 adapter ligation to the repaired 5’ termini
Optional PCR
• PCR-free: inputs down to 10 ng
• With PCR: inputs down to 10 pg
Equivalent adapter ligation efficiency at all input levels
PCR-free Library Prep from 10 ng of cfDNA
• Up to 90% library conversion:
• 10 ng input (91 fmol) = PCR-free library yield of 4.3 nM (82 fmol)
• Sample integrity: qPCR Alu247/Alu115 ratio = 0.4
• PCR-free sequencing eliminates read duplicates and provides more uniform
coverage of AT-/GC-rich sequences
# PE Reads % Mapped to Human
Genome
Estimated Library Size Chimera Dimers
69,853,401 99.5% 1.2 X 109 2.6% 0.02%
Size Distribution
Median
insert size:
169 bp
0.010
0.100
Nu
mb
er
of
rea
ds
pe
r b
as
e a
t e
ac
h c
hro
mo
so
me
no
rma
lize
d
to m
ap
ab
ilit
y
chromosomes
Mapability and read length, Li W. Freudenberg, J.; Front Genet 2014 Nov
Superior Coverage Uniformity from cfDNA
WGS
Two cfDNA samples were prepared PCR-free with the Accel-NGS® 2S PCR-free DNA Library Kit
and sequenced on a HiSeq® 2500 Rapid Run with v3 chemistry. Analysis was performed using a
custom pipeline using BWA as the aligner and enrichment metrics were collected using Picard tools.
Sequencing Metrics Sample A Sample B
Yield 1.8 nM 3.8 nM
Total Reads 238,230,712 254,136,768
% Aligned 99.4% 99.4%
Average Genome Coverage 14.6X 15.6X
% Genome Missing 1.9% 2.0%
% Genome Covered ≥ 5X 99.6% 99.7%
% Genome Covered ≥ 10X 94.9% 95.8%
% Genome Covered ≥ 15X 92.6% 93.2%
% Duplication 0.04% 0.08%
Median Insert Size 172 bp 168 bp
Accel-NGS® 2S Hyb Library Kit:Compatibility with Multiple Hyb Capture Technologies
Compatible with cfDNA samples
Simple with bead protocol
Broad input range 10 pg-1 µg
Repairs both 5’ and 3’ termini to enhance ligation
efficiency
Increased library complexity
Balanced coverage of AT-/GC-rich regions.
A variety of indexing kits allow for compatibility with
multiple hybridization capture technologies: Agilent
SureSelectXT and SureSelectXT2, NimbleGen™
SeqCap™ EZ, and IDT xGen® Lockdown® Probe
Limit of Detection for Accel-NGS® 2S Hyb Kit
• To assess the limit of detection of Accel-NGS 2S Hyb, DNA samples from two
individuals with different ethnic backgrounds were used to prepare libraries.
• The DNA from Sample A was pre-screened for point mutations using the Accel-
Amplicon™ 56G Oncology Panel.
• Sample A was spiked into Samples B and C at 1% final concentration.
• Once libraries were prepared, they were hybridized to IDT xGen® Lockdown®
Pan-Cancer probes and SNPs were detected within this panel.
cfDNA Spike-in to Determine Limit of Detection
for the Accel-NGS® 2S Hyb Kit
• To determine if SNPs present at 1% allele frequency could be detected,1% of
cfDNA sample (A) with a unique ethnic background was spiked into two 10 ng
cfDNA samples (B and C) of different ethnic backgrounds.
• SNPs at 100% frequency unique to sample A could be detected near the
expected 1% allele frequency in the spiked-in samples. Libraries were
sequenced to an average coverage of 8000X.
cfDNA input with xGen® Pan-Cancer Panel
CHR: POS%A
Background
%B
Background
% C
Background
1% A into
10 ng B
1% A into
10 ng C
2: 212244718 100 0 0 0.6 1.0
12: 25361074 100 0 0 1.6 1.9
12: 25361142 100 0 0 1.1 0.9
12: 25361646 100 0 0 1.9 1.6
12: 40688695 100 0 0 0.5 1.1
12:115108136 100 0 0 0.7 2.0
Accel-NGS 2S® Hyb Kit:Low Duplication Rates from Low Input cfDNA
• cfDNA extracted using the PerkinElmer chemagicTM 360
• Libraries generated from as little as 1 ng and 10 ng of cfDNA
• IDT xGen® Lockdown® Pan-Cancer Panel
Hybridization Enrichment of 1 ng cfDNA
Swift Libraries
• Pooled normal human plasma DNA (Innovative Research Inc.)
• 1 ng input, 10 PCR cycles, and library yield ~1 µg
HGSC Standard Library
• NA12878, 500 ng input
Developed by Donna Muzny,
Harsha Doddapaneni, Qingchang
Meng, and Richard Gibbs
Coverage
Library Input Total
MB
Dup.
Rate
Avg.
Coverage
Reads on Hit
Target/Buffer
Targets
Hit
Buffer
Aligned
Reads
Target
Aligned
Reads
> 1X > 20X > 40X
Swift
(Avg. 4
Samples)
1 ng 10,325 20% 93X 55.94% 99.75% 9.79% 46.15% 99.63% 96.53% 89.15%
HGSC
Standard
Library
500 ng 12,925 3% 152X 76.44% 99.65% 14.84% 61.60% 99.35% 97.61% 95.44%
Accel-Amplicon™ Panels
Targeted NGS Panels
Accel-Amplicon™ Technology
Simple, Single-Tube Workflow
10 ng of input from genomic DNA and cfDNA was used to generate libraries with the
Accel-Amplicon 56G Oncology Panel. The coverage uniformity, as the percentage of the
bases covered at least 20%, 30%, 40%, or 50% of the average depth, was determined
across two sample types. The percentage of reads on target was > 95% for all sample
types.
56G Coverage Uniformity Across
Sample Types
A complete solution offering primer pairs, indexing, and
sequencing adapters all in a single kit.
75
80
85
90
95
100
gDNA cfDNA
Co
vera
ge u
nif
orm
ity
0.2x mean 0.3x mean 0.4x mean 0.5x mean
Genomic DNA
Multiplex PCR
90 minutes
Indexing Step
20 minutes
+
Dual-Indexed Amplicon Libraries
Accel-Amplicon™: Available Panels
Targeted sequencing to screen for either germline or somatic variants
Amplicon Panel Description Single Tube
Comprehensive TP53 Comprehensive coverage of all exons ✔
56G Oncology Panel Hotspot coverage of 56 cancer-related genes ✔
EGFR Pathway Panel Hotspot coverage of 4 EGFR pathway genes ✔
Sample_ID Panel 95 high MAF SNPs to generate a unique genetic
fingerprint for every sample
✔
Custom Panels Customer-driven content ✔
Coming Soon:
BRCA1 & BRCA2 Comprehensive coverage of all exons ✔
CFTR Complete exonic coverage of coding regions ✔
Pharmacogenomics 170 genes involved in drug metabolism ✔
Accel-Amplicon™ 56G Oncology Panel v2
• Single-tube oncology panel with 2-hour workflow
• 263 amplicons sized 92-184 bp (avg. 138 bp) for compatibility with cfDNA and FFPE
• 104 Sample_ID amplicons spiked in at a low percentage (4%) for genetic
fingerprinting
• Coverage of hotspots and contiguous regions of 56 clinically-relevant, oncology-
related genes
• Total target size of 23.7 kbTable of Genes and Number of Amplicons Per Gene
ABL1 5 CSF1R 2 FBXW7 6 GNAS 2 KIT 14 NPM1 1 STK11 5
AKT1 2 CTNNB1 1 FGFR1 2 HNF1A 4 KRAS 3 NRAS 3 SMAD4 10
ALK 2 DDR2 1 FGFR2 4 HRAS 2 MAP2K1 5 PDGFRA 4 SMARCB1 4
APC 9 DNMT3A 1 FGFR3 6 IDH1 1 MET 6 PIK3CA 11 SMO 5
ATM 19 EGFR 9 FLT3 4 IDH2 2 MLH1 1 PTEN 14 SRC 1
BRAF 2 ERBB2 4 FOXL2 1 JAK2 2 MPL 1 PTPN11 2 TP53 21
CDH1 3 ERBB4 8 GNA11 2 JAK3 3 MSH6 4 RB1 12 TSC1 1
CDKN2A 2 EZH2 1 GNAQ 2 KDR 9 NOTCH1 3 RET 6 VHL 3
█ Contiguous, overlapping coverage is included for APC, ATM, EGFR, FBXW7, FGFR3,
HNF1A, KIT, MSH6, PIK3CA, PTEN, SMAD4, and TP53.
█ Comprehensive coding exon coverage is included for TP53.
56G Performance with Horizon’s Quantitative
Multiplex Reference Standard
The Accel-Amplicon™ 56G Oncology Panel consistently detected validated variants at the expected frequency in replicates by five different users
from 10 ng of the Horizon Diagnostics Quantitative Multiplex DNA Reference Standards HD701. The variants were called by LoFreq 2.1.1
(Genome Institute of Singapore) and GATK HaplotypeCaller (Broad Institute). When examining sporadic variants among the 10 replicates, the
majority of background variants were present at less than 0.6%. No sporadic variants greater than 0.6% were detected.
Gene AA Change CHR POS REF ALT Expected Allele
Frequency
Detected Allele
Frequency (N=10)
Standard
Deviation
EGFR G719S 7 55241707 G A 24.5 23.8 1.5
PIK3CA H1047R 3 178952085 A G 17.5 17.5 1.3
KRAS G13D 12 25398281 C T 15.0 15.0 1.8
NRAS Q61K 1 115256530 G T 12.5 13.4 1.2
BRAF V600E 7 140453136 A T 10.5 9.9 0.3
KIT D816V 4 55599321 A T 10.0 10.3 1.1
PIK3CA E545K 3 178936091 G A 9.0 8.5 1.1
KRAS G12D 12 25398284 C T 6.0 6.6 1.2
EGFR L858R 7 55259515 T G 3.0 2.7 0.5
EGFR ΔE746-A750 7 55242465-
55242479
Del15bp 2.0 1.4 0.5
EGFR T790M 7 55249071 C T 1.0 1.0 0.3
Accel-Amplicon™ Sample_ID Panel
• SNP Specific Sample_ID Panel
– Generate a genetic fingerprint to track samples within and between studies
– Minimum 1 in 90k discrimination between samples when using a subset of 24
of the 104 amplicons
– Based upon the Genome Medicine publication: Pengelly, RJ et al (2013)
• http://www.genomemedicine.com/content/5/9/89
• 104 Total Amplicons
– 95 for exonic SNPs (all 95 SNPs are covered by most exome panels)
– 9 specific sex ID amplicons
• Can be run as stand-alone panel or combined with
Accel-Amplicon gene panels
– Verification that multiple samples are derived from the same individuals
– When analyzing distinct samples from the same individual (i.e., plasma,
metastatic, normal, tumor compared to normal adjacent, etc.)
Sample_ID Panel as a Spike-In for 56G
cluster
5000x
200x
Sequencing Depth
per Amplicon
Flow CellPCR Reaction
Somatic
Mutation Detection
Germline
Variant Detection
56G
Sample_ID
Experimental Design
Isolate cfDNA with
QIAamp® Circulating
Nucleic Acid kit
5 ng 10 ng
Healthy control blood, n = 5
(Streck Cell-Free DNA BCT)
Tumor bearing blood, n = 8
(Streck Cell-Free DNA BCT®)
Accel-Amplicon™ 56G Oncology
Panel library construction
Accel-NGS® Methyl-Seq
Liquid Biopsy Samples:Tumor Compared to Normal cfDNA Samples
SAMPLE PATHOLOGY
1 Fallopian tube high-grade papillary serous carcinoma pT3c N1 with 2 nodes involved by
micrometasasis
2 5 cm ovarian ‘borderline’ serous content (cancer-like)
3 Recurrent pT2, pN0 mammary carcinoma, 2.15 cm
4 pT1/pN1 pancreatic adenocarcinoma with neoadjuvant therapy
5 Metastatic colon cancer to the liver (previously treated)
6 14 cm ovarian ‘borderline’ serous content (cancer-like)
7 Colon-cancer, non-resectable Adenocarcinoma T4a by imaging
8 Metastatic colorectal adenocarcinoma with liver metastasis, 2 cm primary
Sequencing cfDNA and Matching Tumor
(FFPE) Samples
Sample Cancer Type Gene Hg19
Coordinate
Amino Acid
Change
% Mutant in
FFPE Normal
Adjacent
% Mutant in
FFPE Tumor
% Mutant in
cfDNA
1 Fallopian Tube
Adenocarcinoma
TP53 chr17:757708
5
E285K 0 48 0
TP53 chr17:757848
8
D148H 0 0 5
2 Ovarian
Cystadenofibroma
BRAF chr7:1404531
36
V600E 0 23 1
3 Mammary
Carcinoma
PIK3CA chr3:1789520
85
H1047R 0 17 0
TP53 chr17:757848
8
D148H 0 0 9
8 Metastic
Colorectal
Adenocarcinoma
PIK3CA chr3:1789360
91
E545K 0 23 11
APC chr5:1121755
76
Q1429* 0 20 5
TP53 chr17:757957
5
Q38* or
intron
0 21 14
KRAS chr12:253982
81
G13D 0 22 5
Accel-NGS® Methyl-Seq
DNA Library Kit
For Bisulfite-Converted DNA
Workflow Comparison for Bisulfite Converted
Samples
Accel-NGS® Methyl-Seq Outperforms Other Kits
Accel-NGS
Methyl-Seq
Traditional
Workflow
Random Primer
Approach
Input (ng) 1 1000 50
# PCR Cycles 10 10 10
Yield (nM) 7 15 20
nM Output Per ng Input 7 0.015 0.4
Bias Low Low High
cfDNA Sequencing to Detect Cancer Burden:A Global Biomarker for Cancer
PNAS, vol 110, no 47 (2013) pp18761-18768
Experimental Design
Isolate cfDNA with
QIAamp® Circulating
Nucleic Acid kit
5 ng 10 ng
Healthy control blood, n = 5
(Streck Cell-Free DNA BCT)
Tumor bearing blood, n = 8
(Streck Cell-Free DNA BCT®)
Accel-Amplicon™ 56G Oncology
Panel library construction Accel-NGS® Methyl-Seq
Detecting Genome-Wide Hypomethylation from
10 Million Reads with Accel NGS® Methyl-Seq
• 5 ng of cfDNA from 5 healthy controls and
sample 8 (Metastatic colorectal
adenocarcinoma with liver metastasis.
• Circos plot represents hypomethylation
status on chromosomes 1-22 between.
• Percent hypomethylation calculated by
comparing the Methylation Density in 1
Mb genome bins to healthy controls.
• Red is hypomethylated (> 3 SD lower
than normal mean MD) and green is
comparable to normal.
• Bins assigned hypomethylated if the MD
was > 3 SD lower than the healthy MD
average.
Liquid Biopsy Samples:Tumor Compared to Normal cfDNA Samples
SAMPLE PATHOLOGY
1 Fallopian tube high-grade papillary serous carcinoma pT3c N1 with 2 nodes involved by
micrometasasis
2 5 cm ovarian ‘borderline’ serous content (cancer-like)
3 Recurrent pT2, pN0 mammary carcinoma, 2.15 cm
4 pT1/pN1 pancreatic adenocarcinoma with neoadjuvant therapy
5 Metastatic colon cancer to the liver (previously treated)
6 14 cm ovarian ‘borderline’ serous content (cancer-like)
7 Colon-cancer, non-resectable Adenocarcinoma T4a by imaging
8 Metastatic colorectal adenocarcinoma with liver metastasis, 2 cm primary
Sequencing cfDNA and Matching Tumor
(FFPE) Samples
Sample Cancer Type Gene Hg19
Coordinate
Amino Acid
Change
% Mutant in
FFPE Normal
Adjacent
% Mutant in
FFPE Tumor
% Mutant in
cfDNA
1
Fallopian Tube
Adenocarcinoma
TP53 chr17:757708
5
E285K 0 48 0
TP53 chr17:757848
8
D148H 0 0 5
2Ovarian
Cystadenofibroma
BRAF chr7:1404531
36
V600E 0 23 1
3
Mammary
Carcinoma
PIK3CA chr3:1789520
85
H1047R 0 17 0
TP53 chr17:757848
8
D148H 0 0 9
8
Metastic
Colorectal
Adenocarcinoma
PIK3CA chr3:1789360
91
E545K 0 23 11
APC chr5:1121755
76
Q1429* 0 20 5
TP53 chr17:757957
5
Q38* or
intron
0 21 14
KRAS chr12:253982
81
G13D 0 22 5
Summary
• Accel-NGS® 2S DNA Library Kits– 90% library conversion
– Superior coverage uniformity with cfDNA WGS
– Compatibility with multiple hybridization capture technologies
• Accel-Amplicon™ Panels– Performance with Horizon’s Quantitative Multiplex Reference Standard
– Liquid biopsy samples experimental design
• Accel-NGS Methyl-Seq DNA Library Kits– Accel-NGS Methyl-Seq outperforms other kits
– Detecting genome-wide hypomethylation
THANK YOUwww.swiftbiosci.com
© 2016, Swift Biosciences, Inc. The Swift logo and Accel-Amplicon are trademarks and Accel-NGS is a registered
trademark of Swift Biosciences. Illumina and HiSeq are registered trademarks of Illumina, Inc. Ion Torrent is a trademark
of Thermo Fisher Scientific Inc. Cell-Free DNA BCT is a registered trademark of Streck, Inc. QIAamp is a registered
trademark of QIAGEN. Chemagic is a trademark of PerkinElmer Inc. SureSelectXT and SureSelect XT2 are products of
Agilent Technologies. NimbleGen and SeqCap are trademarks and xGen and Lockdown are registered trademarks of
Roche NimbleGen Inc. 16-1041 08/16