Transcript Table 1

GVRG ABRF 2011 Research Study
Evaluation of DNA Whole Genome Amplification (WGA) Technologies for Genotyping
Amy Hutchinson1, Casey Dagnall1, Charles Nicolet2, Helaman Escobar3, Sean Blake4, Brian Sanderson5, Bruce Kingham6, Karen Jonscher7
1Core
Genotyping Facility, SAIC-Frederick, Inc., NCI-Frederick, Frederick, MD; 2Data Production Facility, USC Epigenome Center, Los Angeles, CA; 3Eurofins MWG|Operon, Huntsville, AL;
4DNA Core Facility, University of Mis souri, Columbia, MO; 5Life Technologies, Austin, TX; 6DNA Sequencing & Genotyping Center, University of Delaware, Newark, DE; 7SBCF and TATC Proteomics, UC Denver, Denver CO
Introduction
Amplified Sample QC
The evolution of genomic technologies is occurring rapidly and often requires large amounts
of source DNA. There is also an expanded desire to analyze smaller numbers of cells for
higher resolution studies as well as to take advantage of large numbers of archived samples
(eg. FFPE, serum, etc.). To provide enough material for the newest genomic technologies,
whole genome amplification (WGA) has reemerged as an important and necessary
technique.
With new WGA products on the market, the GVRG has completed a benchmarking study
evaluation of 6 commercially available WGA kits using several QC and genotyping assays.
Utilizing 6 samples, the WGA kits were tested following the manufacturer’s protocols.
Quality metrics provided measurements on yield, fragment size, and concentration of
material derived from each kit. Several widely adopted genotyping methods were then used
to evaluate the amplification products, including Illumina Human Omni1 Quad Beadchip,
TaqMan copy number and SNP genotyping assays, and STR genotyping.
All amplified and genomic control samples (66 wgaDNA + 6 gDNA) were subjected to several
quality control (QC) measures:
 Quantification via NanoDrop (OD) and PicoGreen (Invitrogen Quant-iT™ dsDNA Reagent)
 Agilent BioAnalyzer analysis
 STR Fingerprinting via Applied BioSystem’s AmpFℓSTR® Identifiler® assay
Figure 1 – Fold increase of material yield by
sample and kit
Illumina Infinium Genotyping
Figure 7 – Call Rates on Infinium
platform of all samples run
across all SNPs by kit
The samples used for the experiment include Coriell DNA from a trio of CEPH individuals, 2
FFPE DNAs, both from the same individual (1 newly extracted and the other extracted 2 years
ago), and 1 sample from the Coriell trio that was fragmented prior to amplification for a total
of 6 samples tested.
•
FFPE Derived Sample
• CEPH/Utah Pedigree 1463 Trio from HapMap Project
 All processing following the manufacturer’s provided protocol
 Following amplification, all samples were sent to a single laboratory for QC and
genotyping.
Table 1 – WGA Kit List and Comparison
NuGen Ovation
Sigma Complete
(WGA2)
Qiagen REPLI-g
Figure 9 – 9a) Concordance rates of all
samples on 18 duplicated SNPs by kit
and 9b) Discordance type by kit
For concordance checking, eighteen (18) TaqMan assays were selected and run on all WGA and
genomic samples for SNPs also present on the Infinium Omni1 Quad BeadChip.
Figures 3 & 4 – Average
fragment length data by
sample and kit (left) and
BioAnalyzer traces from
amplified sample across
kits (right)
The WGA kits included 6 options from Sigma, NuGen, Qiagen and GE Life Sciences. Samples
were amplified as blind duplicates (n = 12) at several GVRG member lab sites using 5 of the
kits for a total of 60 products. Additionally, the Ovation FFPE kit provided reagents for the 6
samples only (no duplicates) adding another 6 WGA products for a grand total of 66 samples.
10 ng
Figure 8 – Call rates on Infinium platform
of all samples by 18 duplicated SNPs by kit
18 SNP Platform Comparison – TaqMan and Infinium
Amplification Kits
GE Life Sciences
GenomiPhi
Fig. 9b
Newly Extracted
Extracted 2 years ago
• Fragmented Sample
NA12891 (average fragment length: 477 bp)
NA12891 (Father)
NA12892 (Mother)
NA12878 (Daughter)
Kit
Fig. 9a
Figure 2 – Summary of allelic distribution by sample
and kit from the Identifiler assay.
Samples
Required
Input
 Infinium Human Omni1Quad Chips
• Concordance with genomic greater than 99.8% using 90% call rate cut-off
• Genotypes 100% consistent with Mendelian inheritance among trio
Input Types
Expected
Yield
Expected
Product Length
(in Kb)
Average: 10
Range: 2-100
2 hours
Hands-On
Processing
Time
20 min
5 hours
2 hours
3.5 hours
1 hour
17 hours
40 min
Process
Time
4 - 7 ug
10 ng
Purified gDNA, whole
blood, mouth wash,
buccal, dried blood cards
Purified gDNA
10 ng
Purified gDNA
10 ug
10 ng
10 ug
Millionfold
Average: 0.4
Range: 0.1-1
3.5 hours
1 hour
3.5 - 5 ug
Average: 0.2
Range: 0.05-1.5
8.5 hours
4 hours
Sigma Single Cell
(WGA4)
1 cell or
< 100 pg*
Purified gDNA, whole
blood, mouth wash,
buccal, dried blood cards
Single cell or
Purified gDNA
NuGen Ovation
FFPE
100 ng
Purified gDNA
3 - 6 ug
Average: 0.2
Range: 0.05-1.5
Average: 0.4
Range: 0.1-1
Average: 10
Range: 2-100
* 64pg was used as the input amount for the Sigma Single Cell kit (WGA4)
TaqMan & Copy Number Genotyping
TaqMan® SNP Genotyping Assays (18 SNPs chosen from SNPs on Infinium chips) and TaqMan®
Gene Copy Number Assays (2 genes)
Figure 5 – 5a)Concordance Rate by Kit and 5b)
Discordances by Sample and Type for TaqMan SNP
Assays
Figure 6 – 6a) Call Rates by Gene and Kit, 6b) Concordance Rate
By Kit, and 6c) Types of Discordances by Gene and Kit for
TaqMan Copy Number Assays
Fig. 6a
Figure 10 – Call variance between TaqMan
and Infinium across duplicated assays by kit
Figure 11 – Breakdown of call variance between TaqMan and Infinium
by kit and variance type
Discussion
Whole genome amplification is a useful tool, and possibly a necessary step, for amplifying unique
and rare samples in order to have enough input DNA material for use with current technology.
This study has evaluated the practical use of several commercially available WGA kits in relation to
ease of use, quality of amplification and maintenance of genetic integrity across various
genotyping platforms in relation to regular, degraded, and FFPE DNA sources. There is excellent
performance across the kits tested though there are several things to take into account when
selecting a kit for use:
 Take into account the downstream application or platform (required input, fragment sizes, etc.)
 Consider QC measure to ensure the amplification products meet needs
• Identifiler assay is representative of LOH and no amplification in samples tested here
 FFPE samples can be accurately genotyped using a WGA sample
 Ease of kit automation if high-throughput is desired
Acknowledgements
Fig. 5a
Fig. 5b
Fig. 6b
Fig. 6c
The ABRF Genomic Variation Research Group (GVRG) is very thankful for the support of its
corporate sponsors whose generous help enabled the GVRG 2011 study. Specifically, the GVRG
would like to thank GE Healthcare Life Sciences, NuGen Technologies, Qiagen and Sigma-Aldrich,
who generously donated the whole genome amplification kits for this study. The GVRG also
thanks Bruce Kingham for supplying the FFPE samples used in this study.