4. Page 4 of 51
Abstract
The Promega® PowerPlex® Y23 System is the most recent Y-STR system developed by
Promega® Corporation (Madison, WI) to replace the original PowerPlex® Y system. The
PowerPlex® Y23 amplification kit contains 11 more loci than the PowerPlex® Y system and
includes two rapidly mutating loci which allows for potentially greater discrimination between
paternally related males (1). PowerPlex® Y23 system can be used for both casework and direct
amplification applications and with a shortened amplification time provides a more efficient
analysis process. The Y23 kit demonstrates a large degree of sensitivity, even in the presence of
excessive female DNA (1).
An internal validation was performed on the PowerPlex® Y23 PCR Amplification kit in
accordance with the Scientific Working Group for DNA Analysis Methods (SWGDAM) validation
guidelines, and the FBI Quality Assurance Standards for Forensic DNA Testing Laboratories
(September 2011 revision) (2) for the North Carolina State Crime Laboratory Forensic Biology
section. Automated DNA extractions were performed throughout the project using the
Qiagen® EZ1 Advanced® Robot. DNA quantification was performed on an Applied Biosystems®
(AB; Foster City, California)7500 real-time PCR instrument using the AB Quantifiler® Duo kit and
AB Human Identification (HID) Real-Time PCR Analysis software version 1.1 for data analysis.
PCR amplification was performed on the Applied Biosystems GeneAmp 9700 thermal cycler
following the Promega PowerPlex® Y23 Technical Manual Protocols at 30 cycles (1). Capillary
electrophoresis was performed on the Applied Biosystems® 3500xL Genetic Analyzer using data
collection software version 2.0, and where all data was analyzed using Applied Biosystems
GeneMapper® ID-X v 1.4 (3).
5. Page 5 of 51
Internal validation studies included the following: precision, sensitivity, concordance,
reproducibility, contamination, mixtures (to include male/male and male/female scenarios),
stochastic evaluation of the DYS385 locus, minimum threshold assessment, and non-
probative/mock sample studies. Sensitivity results demonstrated that the Y23 system could
consistently generate full profiles at concentrations of 0.03125ng, and full male profiles were
also observed in several samples at concentrations as low as 0.0156ng. Male/female mixture
study results indicated that full male profiles could be consistently obtained at ratios as
extreme as 1:16,000, illustrating the specificity of the Y23 for male DNA amplification.
Additionally, a study was performed to explore the viability of Y23 PCR product over a period of
several weeks. These studies became the basis of the efficient and reliable operating
procedures for the PowerPlex® Y23 amplification kit for the North Carolina State Crime
Laboratory Forensic Biology section
1. Introduction
Y-STR systems can be an effective tool in distinguishing between males of different
paternal lineage. Generating male profiles can be useful in identifying missing persons and
human remains, distinguishing male contributors in complex autosomal DNA mixtures, and
potentially excluding male contributors in samples containing minor male components (4).
Y-STRs are implemented in the Combined DNA Index System (CODIS) in conjunction with
traditional STRs to provide more information on missing person cases and unidentified remains
at the National DNA Index System (NDIS) level (5). Y-STRs are also beneficial for sexual assault
6. Page 6 of 51
evidence where the female contributor overwhelms the male contributor, or when differential
extractions cannot effectively separate the male and female contributors (6).
Internal validation studies included the following: precision, sensitivity, concordance,
reproducibility, contamination, mixtures (to include male/male and male/female scenarios),
stochastic evaluation of the DYS385 locus, minimum threshold assessment, and non-
probative/mock sample studies.
The PowerPlex® Y23 System is a 23-loci multiplex that uses a five-dye chemistry
to allow for the amplification of loci including DYS576, DYS389 I, DYS448, DYS389 II, DYS19,
DYS391, DYS481, DYS549, DYS533, DYS438, DYS437, DYS570, DYS635, DYS390, DYS439, DYS392,
DYS643, DYS393, DYS458, DYS385(a/b), DYS456, and YGATAH4 (7). As compared to PowerPlex
Y with 12 loci identified, the increase in Promega PowerPlex® Y23 loci allows for greater
discrimination among males with no paternal relation while the two rapidly mutating loci,
DYS570 and DYS576, allow for possible discrimination among paternally related males.
Paternally related males can be used to identify unidentified remains, missing persons, and/or if
the actual suspect standard is unavailable, but one can be obtained from a brother, father, or
other paternal relative (7).
All extraction processes were completed using stain extraction buffer (SEB), 10 µL of
Proteinase K, and 1 µL of carrier RNA in 2mL tube with lyse and spin baskets. A portion of each
sample was incubated for at least one hour to overnight in a thermomixer set to 56°C.
The samples were then spun down in a microcentrifuge for 5 minutes and subsequently
transferred to a Qiagen® EZ1 Advanced XL robot for the extraction. All known reference
7. Page 7 of 51
samples were eluted in 100 µL of TE while sperm fractions of differential extractions were
eluted in 40 µL of TE (8).
All quantitation steps were performed using the Applied Biosystems® Quantifiler® Duo
kit and the Applied Biosystems® Real-Time PCR 7500 instrument. The quantitation plate was
prepared using the Qiagen® (Qiagen, Inc., Valencia, CA) QIAgility® Robot and in addition to the
standard curve prepared by the robot, a duplicate standard curve was added manually
following the Qiagen® QIAgility® set up.
The amplification steps were performed on an Applied Biosystems® GeneAmp® 9700
thermal cycler with the target amplification volume of 25 μL. The protocols listed in the
Promega® PowerPlex® Y23 System Technical Manual were utilized to establish thermal cycling
parameters (Figure 1). For the amplification plate set up, several different methods were
utilized based on the specific study that was being performed (1).
Figure 1. Thermal Cycling Parameters Set for the Validation Studies
8. Page 8 of 51
2. Methods
2.1 Precision Studies
The purpose of a precision study is to ensure that the amplification system can
accurately distinguish between allele calls within one base pair. PowerPlex® Y23 Allelic ladder
was loaded into wells A01 through H03 for a total of 24 wells on a single 96-well plate. Each
well contained 10 µL of Hi-Di Formamide, 1.0 µL of CC5 Internal Lane Standard (ILS), and 1.0 µL
of Y23 Ladder described in the PowerPlex® Y23 System Technical Manual (1). The plate was
injected consecutively four times on the AB 3500xL Genetic Analyzer for a total of 96 individual
samples, while the CC5 ILS 500 was used to size each allele.. The standard 24 second injection
protocol was utilized for this study as well as throughout the entire validation. The size of each
allele was analyzed per capillary, per injection, and by both capillary and injection, for three
separate analyses. Microsoft® Excel® was utilized to generate statistics for all allele sizes which
included the minimum, maximum, average, and standard deviations of each allele for all three
analyses.
2.2 Sensitivity Studies
The purpose of a sensitivity study is to demonstrate how much target DNA is required to
obtain full male profiles, and show how sensitive the amplification system is in regards to small
amounts of input sample DNA. Three unrelated male samples were chosen from the set of
known standards and 2-fold serial dilutions were prepared with TE buffer starting at 0.5 ng
down to 0.0156 ng for each of the three male samples. These dilutions were quantified on the
7500 instrument and the resulting data was used to more accurately reflect target input
9. Page 9 of 51
amounts of DNA for each dilution. For the sensitivity studies, each dilution was amplified in
triplicate using the GeneAmp® 9700 thermal cycler set to the parameters listed in the
PowerPlex® Y23 Technical Manual (1). All of the samples and controls were amplified with 5 µL
of 5X Master Mix and 2.5 µL of 10X Primer Pair Mix. The dilutions were normalized with the
Amplification Grade Water to combine for a total of 17.5 µL to equal the desired total DNA
input.
2.3 Minimum and Analytical Threshold Analysis
The purpose of the minimum and analytical threshold analysis is to determine at what
RFU value a called allele can be stated as a true allele or artifact and not background noise from
the amplification kit or instrument. The minimum threshold analysis was performed in order to
set limits for assessing the peaks in the electropherogram for the baseline noise of the
instrument. Instrument noise, true DNA fragments and a variety of artifacts result in the peaks
as seen in the electropherogram. The analytical threshold is defined by the 2010 SWGDAM
Interpretation Guidelines for Autosomal STR Typing as “the minimum height requirement at
and above which detected peaks can be reliably distinguished from background noise; peaks
above this threshold are generally not considered noise and are either artifacts or true alleles”
(9). The analytical threshold can be calculated using the various data collected throughout the
validation, including minimum threshold, and it can be generated using the following:
𝐴𝑛𝑎𝑙𝑦𝑡𝑖𝑐𝑎𝑙 𝑇ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑 = 2(𝑀𝑎𝑥𝑖𝑚𝑢𝑚 𝑝𝑒𝑎𝑘 ℎ𝑒𝑖𝑔ℎ𝑡 − 𝑀𝑖𝑛𝑖𝑚𝑢𝑚 𝑝𝑒𝑎𝑘 ℎ𝑒𝑖𝑔ℎ𝑡)
𝐿𝑖𝑚𝑖𝑡 𝑜𝑓 𝐷𝑒𝑡𝑒𝑐𝑡𝑖𝑜𝑛 = 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑝𝑒𝑎𝑘 ℎ𝑒𝑖𝑔ℎ𝑡 + (3 × 𝑆𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝐷𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 𝑝𝑒𝑎𝑘 ℎ𝑒𝑖𝑔ℎ𝑡)
10. Page 10 of 51
𝐿𝑖𝑚𝑖𝑡 𝑜𝑓 𝑄𝑢𝑎𝑛𝑡𝑖𝑡𝑎𝑡𝑖𝑜𝑛 = 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑝𝑒𝑎𝑘 ℎ𝑒𝑖𝑔ℎ𝑡 + (10 × 𝑆𝑡𝑑. 𝐷𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛 𝑝𝑒𝑎𝑘 ℎ𝑒𝑖𝑔ℎ𝑡)
Various negative controls, positive controls, positive samples and known standards were
selected from projects created throughout the validation for a total of 24 samples.
GeneMapper® ID-X software was used to analyze the 24 various samples. The analysis
threshold was set to 1 Relative Fluorescent Unit (RFU) and all known artifacts (pull up, stutter, -
A, etc) and alleles were removed for data analysis. The remaining calls were considered the
background noise from the instrument or amplification kit itself, and the calls were separated
by the dye color in order to calculate the minimum RFU, maximum RFU, average RFU, and
standard deviation for each dye color using Microsoft Excel.
2.4 Stochastic Threshold
The purpose of establishing the stochastic threshold is determine at what RFU value, a
called allele can be stated as a true homozygous allele and not heterozygous with dropout of a
sister allele. A stochastic threshold is the RFU level at a locus where a homozygous peak called
above that threshold can be considered a true homozygous peak and not a possible
heterozygous locus with dropout of the sister allele. For the PowerPlex® Y23 Kit, there is only
one locus, DYS385, which can have two peaks as it is a multi-copy number marker (DYS385a and
DYS385b). The stochastic threshold does not apply to the remaining loci in this kit due to the
single peaks expected at those loci. Seven various known male samples were selected based on
being the most genetically distinct from each other across all loci specifically at the DYS385
locus (Table 2.4.1). Of the samples used for the stochastic study, no two samples had the same
11. Page 11 of 51
two alleles at the DYS385 locus, but there were common alleles present in several of them.
Dilutions of 0.03125ng and 0.0156ng were prepared from each of the seven samples. All
dilutions were quantified to verify accuracy of each dilution’s quantity. The dilutions were
normalized using the Amplification Grade Water provided with the PowerPlex® Y23 kit and each
was amplified in triplicate.
Table 2.4.1. Allele Calls of Samples at DYS385 Locus
2.5 NIST/ Contamination
The purpose of the NIST/Contamination study is to demonstrate that the amplification
system is in compliance with the FBI Quality Assurance standards. Per the FBI Quality
Assurance Standard 9.5.5, an appropriate NIST standard reference material must be used in
order to check the laboratory’s procedures (10). NIST SRM 2391-c Components B, C, and D
were used to ensure the expected full male profiles were obtained and to demonstrate
compliance with QAS Standard 9.5.5. Component A, which was female, was also used to verify
that the kit fails to amplify DNA from female contributors. Components E and F were not
required to be tested because Component E was a female stain on 903 paper while Component
F was a male stain on FTA paper, which are both designed more for direct amplification
Sample DYS385
K01 16,19
K03 11,12
K05 15,16
K08 13,14
K10 13,18
K14RE 10,14
K23 11,14
12. Page 12 of 51
applications (11). After quantitation of the NIST components, the appropriate dilutions were
made targeting 0.25ng/ µL of DNA. Then each standard was amplified with 5 µL of 5X Master
Mix and 2.5 µL of 10X Primer Pair Mix, while 2 µL of each component for a target input of 0.5 ng
of DNA was added with 15.5µL of the Amplification Grade Water for a total 25 µL per well.
Contamination was monitored throughout the validation through evaluation of all of the
controls from each process. The extraction and amplification negative controls were observed
to see if any contamination occurred in what should be a blank sample. Known single source
samples and positive amplification controls were observed to see if any other source could
contaminate the samples.
2.6 Non –Probative/ Mock Study
The purpose of the non-probative/mock study is to demonstrate the ability of the kit to
obtain expected results from mock casework samples. A variety of 26 samples were chosen or
prepared in order to demonstrate kit performance with non-probative and mock forensic
samples. An adjudicated case was chosen and items from the case included the victim
standard, vaginal swab, oral swab, and a cutting from panties. A differential extraction was
performed on all items but the victim standard, which underwent extraction protocols for
known samples. Three aspermic post-coital samples at three different time intervals (24, 48,
and 72 hours) were also tested. These samples did not need to have a differential extraction
performed. Three different male mixture samples were chosen as well and a differential
extraction was performed on each. Three previous proficiency tests were selected and samples
from each were tested. One test included a sperm sample from a bed sheet, a sample from the
13. Page 13 of 51
suspect’s shirt (no sperm identified), and standards from the suspect and victim. The remaining
proficiency tests both included differential stain samples, suspect standards, and victim
standards. Differential extractions were performed on all samples previously identified as
having sperm present. Additionally, three semen dilutions were prepared at 1:200, 1:150, and
1:200; and three male saliva dilutions were prepared at 1:50, 1:75, and 1:100; and all were
extracted as aspermic samples. The lab’s protocols for differential, unknown, and known
extractions were followed for the appropriate mock and non-probative samples. An extraction
control was used for the sperm, nonsperm, aspermic, and standard samples. All of the known
and standard samples for the study were extracted on the EZ1 together and separate from the
sperm and non-sperm sample fractions. All of the samples for the study were quantified,
amplified, and run as described previously.
2.7 Mixture Studies
The overall purpose of the mixture studies was to illustrate the ability of the PowerPlex®
Y23 kit to produce Y-profiles when mixed with highly concentrated female DNA samples, as well
as other male DNA samples. The ability of the kit to distinguish between male contributors in
mixed samples was determined through various ratios of male to male mixtures as well as male
to female mixtures. In order to determine the sensitivity of the kit for mixtures with high
concentration of female DNA, a constant, large volume of female DNA was mixed with varying
low concentrations of male DNA (Study A). Large, varying amounts of female DNA was mixed
with a constant amount of male DNA (Study B). Varying ratios of male:male mixtures were
used to determine the ability of the kit to produce profiles representative of those ratios while
14. Page 14 of 51
holding the total input amount of DNA constant (Study C). Stochastic levels of the male DNA
mixtures from study C were also tested to determine how well profiles at low input levels are
representative of the male to male ratios (Study D).
Mixture Study A:
The purpose of the mixture study A is to demonstrate that the amplification system
correctly obtains the male profile without inhibition from the constant overwhelming amount
of female DNA. A known female sample, K12, was extracted and quantified at a concentration
of 31.4395 ng/µL (stock solution). The known male sample was extracted and quantified, then
a 4-fold serial dilution was prepared at 0.5ng, 0.25ng, 0.125ng, 0.0625ng, 0.03125ng, and
0.0156ng concentrations (see Table 2.7.1a). Those dilutions were then re-quantified and
normalized along with the female stock solution. Table 2.7.1b shows the preparation of each
sample into one master mix tube.
Table 2.7.1a. Mixture Study A: High constant female contribution; variable male
Target Concentration (in ng)
Ratio Combo # [Female] [Male] Ratio (:1M)
1 250 0.5 500
2 250 0.25 1000
3 250 0.125 2000
4 250 0.0625 4000
5 250 0.03125 8000
6 250 0.015625 16000
15. Page 15 of 51
Table 2.7.1b. Normalization of mixtures to equal total input volume of desired female target
concentration and male target concentration for each reaction.
Sample ID
K12
(volume
)
K05
(volume
)
Water
(volume
)
5X
Master
Mix
(volume)
10X
Primer
(volume
)
Total
Volume
MixA_250F:0.5M 7.95 1.47 8.08 5 2.5 25.00
MixA_250F:0.25M 7.95 0.87 8.68 5 2.5 25.00
MixA_250F:0.125M 7.95 0.93 8.62 5 2.5 25.00
MixA_250F:0.0625M 7.95 1.15 8.40 5 2.5 25.00
MixA_250F:0.03125M 7.95 1.17 8.38 5 2.5 25.00
MixA_250F:0.015625
M 7.95 3.40 6.15 5 2.5 25.00
Mixture Study B:
The purpose of the mixture study B is to demonstrate the kit obtains the full male
profile without inhibition from the varying amounts of overwhelming female DNA. A stock of a
0.5 ng/µL male DNA sample and a stock of 0.5 ng/µL female DNA sample were prepared and
quantified in order to calculate the input amount of each sample to create the most accurate
target ratios. The female to male ratios were determined and prepared keeping the amount of
male DNA constant while increasing the amount of female DNA (Table 2.7.2). The mixtures
were quantified in triplicate and then normalized for amplification with a total input at 0.5ng of
DNA.
Table 2.7.2. Mixture Study B: Male:Female Mixtures
Female
(K28_0.25ng stock)
Male
(K10_0.25ng stock)
Starting [ ]
Total
Target
Vol to
add Starting [ ]
Total
Target
Vol to
add
F:M
Ratio
0.2834 24.5 86.5 0.2065 0.5 2.4 49:1
16. Page 16 of 51
0.2834 23.75 83.8 0.2065 1.25 6.1 19:1
0.2834 22.5 79.4 0.2065 2.5 12.1 9:1
0.2834 18.75 66.2 0.2065 6.25 30.3 3:1
0.2834 12.5 44.1 0.2065 12.5 60.5 1:1
Mixture Study C:
The purpose of the mixture study C is to determine whether or not two male
contributors can be distinguished at varying ratio amounts in a male:male mixture profile.
Stock solutions of 500µL of 0.25ng/µL for two separate and distinguishable known male
samples were prepared and subsequently quantified in order to calculate the input amount of
each sample. This ensured the most accurate target ratios for the amplification and mixture
analysis. Table 2.7.3 describes the preparation of the stock solutions for each sample and
mixture. The mixture C samples were prepared for a total target DNA input of 0.5 ng.
Table 2.7.3. Mixture Study C: Male:Male Mixture Ratios Preparation
M1 (K07_0.25ng stock) M2 (K10_0.25ng stock) M1:M2
RatioStarting [ ] Total
Target
Vol to
add
Starting [ ] Total
Target
Vol to add
0.3393 23.75 70.0 0.3702 1.25 3.4 19:1
0.3393 22.5 66.3 0.3702 2.5 6.8 9:1
0.3393 18.75 55.3 0.3702 6.25 16.9 3:1
0.3393 12.5 36.8 0.3702 12.5 33.8 1:1
0.3393 6.25 18.4 0.3702 18.75 50.6 1:3
0.3393 2.5 7.4 0.3702 22.5 60.8 1:9
0.3393 1.25 3.7 0.3702 23.75 64.2 1:19
Mixture Study D:
The purpose of a mixture study D is to look at stochastic levels of the male DNA
mixtures from study C to determine how well profiles at low input levels are representative of
the male to male ratios. The mixture ratios that were prepared in the mixture C studies were
17. Page 17 of 51
used for the mixture D studies with targeted input values of 31pg and 15pg. This study
examined the resulting peak height ratios for the same mixtures as used in Study C, but at
lower total template amounts. The purpose of the study was to determine whether or not
major and minor contributors could be distinguished reliably in mixture samples at very low
DNA input amounts. Internal validations from other laboratories including the Wisconsin State
Crime Laboratory Bureau have documented the phenomena referred to as “flip-flopping”
where the expected higher input allele actually has the lower RFU value and the lower expected
allele has the higher RFU value (12). In other words, the predominant allele observed at low
input levels may not be attributable to the major contributor. This was analyzed through visual
inspection of the electropherograms and comparing the observed results to expected
contributor profiles.
Table 2.7.4. Mixture D Preparation at Stochastic Levels
Sample
Name
Quant Duo
Concentration
Final DNA
Input
Quantity
Dilution
Per Rxn
Water Per
Rxn
Amount of
Dilution
(4 Rxns)
Water
(4 Rxns)
MixD1_19:1 0.2776 0.03125 0.11 17.39 0.3 52.2
MixD1_19:1 0.2776 0.01563 0.06 17.44 0.2 52.3
MixD1_9:1 0.2898 0.03125 0.11 17.39 0.3 52.2
MixD1_9:1 0.2898 0.01563 0.05 17.45 0.2 52.3
MixD1_3:1 0.2875 0.03125 0.11 17.39 0.3 52.2
MixD1_3:1 0.2875 0.01563 0.05 17.45 0.2 52.3
MixD1_1:1 0.2995 0.03125 0.10 17.40 0.3 52.2
MixD1_1:1 0.2995 0.01563 0.05 17.45 0.2 52.3
2.8 Robustness of Amplification Product
The strength of the PCR product was expressed through the robustness of the PCR
products over time. The study of the amplification product stability over time can also look at
18. Page 18 of 51
reproducibility of the samples, but it demonstrates the actual reproducibility of 3500xL Genetic
Analyzer, rather than the amplification system. A full amplification plate was initially prepared,
amplified, and injected on the genetic analyzer in order to determine the known reference
samples for comparison. This plate was stored at -4°C and every seven days a new plate was
set up for a run on the genetic analyzer from the stored amplification plate. The results for
alleles, ILS, and RFU values were analyzed in Excel® for comparison to the initial data collected.
The 2800M stability was determined in order to see the period of time in which the dilutions of
the 2800M can still give acceptable results under the stored conditions at -4°C. For each
amplification prepared during the validation, a new 2800M dilution was prepared at the
recommended 20:1 ratio of amplification grade water to DNA. At the conclusion of the
validation, all of the 2800M dilutions prepared were amplified in triplicate and injected on the
3500xL Genetic Analyzer. The average RFU values for the alleles at each locus were compared
sample to sample and analyzed in Excel®.
2.9 Reproducibility
The purpose of reproducibility studies is to demonstrate the ability of the amplification
system to produce the same, expected results when the analysis process is repeated.
Reproducibility was assessed throughout the validation from the single source standards and
positive controls (2800M) that were amplified at the optimal target DNA input of 0.5 ng. The
ability of the PowerPlex® Y23 system to produce the same profiles was determined from
looking at the repeated amplifications and profiles of the known single source standards as well
as the 2800M throughout the validation studies.
19. Page 19 of 51
3. Results
3.1 Precision Results
The PowerPlex® Y23 precision studies included calculations for minimum and maximum
base pair size, average base pair size, and the standard deviation of the base pair size all the
alleles over every loci. These statistics were generated by injection, by capillary, and all
injections and capillaries combined. Microsoft® Excel® was utilized to generate statistics for
each allele’s base pair size in the allelic ladders for all of the instrument conditions over all
injections and capillaries. The lowest standard deviation and highest standard deviation for all
of the alleles by each capillary was determined and shown in Table 3.1.1. The lowest and
highest standard deviation was also determined by each injection and shown in Table 3.1.2.
Figure 2 demonstrates the standard deviations at each loci with the average base pair size of
the alleles. The larger alleles and loci seemed to have the larger standard deviations however
there is only a slight pattern to demonstrate that trend. Table 3.1.3 shows average standard
deviations at each locus from largest to smallest with the loci greater than 300 base pairs is size
highlighted. This shows that some larger loci had a greater standard deviation, but there is no
distinct trend. Figure 2 demonstrates that the larger allele calls resulted in higher standard
deviations. DYS19 was the only larger locus that did not fall in the higher standard deviation
range. The highest standard deviation was at the DYS385 locus which could be a result of the
DYS385 locus multi-copy number characteristic, which was not taken into account for during
the calculations of the standard deviations. This locus is not one of the larger loci but has a
range of 220 – 320 base pairs is size.
20. Page 20 of 51
Table 3.1.1. Standard deviation range of all alleles (per capillary)
Capillary
Lowest Standard
Deviation
Highest
Standard
Deviation
1 0.0050 0.1153
2 0.0000 0.1040
3 0.0000 0.1063
4 0.0000 0.1207
5 0.0000 0.1127
6 0.0050 0.1273
7 0.0000 0.1130
8 0.0000 0.1008
9 0.0000 0.1215
10 0.0000 0.1215
11 0.0000 0.1066
12 0.0000 0.1215
13 0.0000 0.1124
14 0.0000 0.0929
15 0.0050 0.1226
16 0.0000 0.1253
17 0.0000 0.1201
18 0.0000 0.1103
19 0.0000 0.0980
20 0.0050 0.1182
21 0.0000 0.0927
22 0.0000 0.1008
23 0.0000 0.1049
24 0.0000 0.0918
Across all
capillaries
0.0091 0.0276
Table 3.1.2. Range of allelic ladder standard deviations by injection (across all capillaries)
Injection
Lowest
Standard
Deviation
Highest
Standard
Deviation
1 0.0000 0.0573
2 0.0163 0.0580
3 0.0051 0.0549
4 0.0167 0.0542
21. Page 21 of 51
Figure 2. Allele Size Standard Deviations versus Average Allele Size across All Injections
Table 3.1.3. Standard Deviations at Each Loci by Dye Channel
Locus Average Std Dev
DYS385 25.695
DYS448 19.036
DYS570 18.263
DYS643 17.405
DYS458 16.620
DYS635 16.098
DYS438 16.032
DYS456 15.161
DYS390 15.157
DYS576 14.987
DYS392 14.724
DYS391 14.239
DYS393 14.085
DYS389 II 13.915
DYS439 13.833
DYS481 13.616
DYS19 12.876
DYS533 12.830
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0 50 100 150 200 250 300 350 400 450
StandardDeviation(bp)
Average Allele Size (bp)
Allele Size Standard Deviations Across All Plate
Injections
DYS576
DYS389 I
DYS448
DYS389 II
DYS19
DYS391
DYS481
DYS549
DYS533
DYS438
DYS437
DYS570
DYS635
DYS390
DYS439
DYS392
DYS643
DYS393
DYS458
DYS385
DYS456
YGATAH4
22. Page 22 of 51
YGATAH4 12.769
DYS549 12.735
DYS389 I 10.437
DYS437 9.147
3.2 Sensitivity Studies
The average peak heights for all alleles at each input target shown in Table 3.2.1 reveal
the positive correlation between the peak heights and the input target. Similarly, as input
target decreases so does the corresponding peak height. The peak height ratios for the DYS385
locus are listed in Table 3.2.2. The average peak height ratios are above 50% for all of the input
target amounts. However, there were several minimum peak height ratios that were below
50% for the 31 pg and 15 pg input amounts. The minimum peak height ratio for the 62 pg
amount was at 47% which is just below 50% making that input amount questionable. The total
allele calls for each input target per sample is indicated in Table 3.2.3, which also shows the
majority of the profiles were full profiles. However, there were some incomplete profiles at the
31 pg and 15 pg input amounts. Figure 3 shows the trend that the peak height decreases as the
target input decreases along with the standard deviation.
Table 3.2.1. Total Average Peak Heights per Input Target Amount
Input
Target
Average of
Height (RFU)
0.5 ng 13655.57
0.25ng 6022.22
0.125 ng 3816.91
0.0625 ng 1497.96
0.03125
ng
1562.32
0.0156 ng 669.25
23. Page 23 of 51
Table 3.2.2. Peak Height Ratio for DYS385 at All Input Targets
Input Target
Min of %PHR
(DYS385)
Average of %PHR
(DYS385)
0.5 ng 86% 89%
0.25 ng 63% 84%
0.125 ng 64% 81%
0.0625 ng 47% 77%
0.03125 ng 34% 73%
0.0156 ng 26% 69%
Table 3.2.3. Total Alleles Called for All Replicates at Each Input Target
Input Target Sample Name
Total Alleles
(all replicates)
Expected Total
(all replicates)
%
Complete
Profiles
0.5 ng K05 69 69 100%
K07 69 69 100%
K08 69 69 100%
0.25ng K05 69 69 100%
K07 69 69 100%
K08 69 69 100%
0.125 ng K05 69 69 100%
K07 69 69 100%
K08 69 69 100%
0.0625 ng K05 69 69 100%
K07 69 69 100%
K08 69 69 100%
0.03125 ng K05 64 69 93%
K07 69 69 100%
K08 69 69 100%
0.0156 ng K05 25 69 36%
K07 63 69 91%
K08 69 69 100%
24. Page 24 of 51
Figure 3. Average Peak Height across All Loci versus the DNA Input Target with the Standard
Deviations
3.3 Minimum and Analytical Threshold
Calculations for minimum threshold were assessed for each dye using the limit of
detection and limit of quantitation. The analytical threshold was calculated for each dye using
the minimum and maximum RFU values (Table 3.3.1). Once the thresholds for each dye were
determined, a final overall analytical threshold was determined to be 150 RFUs.
Table 3.3.1. Analytical Threshold Calculations for Each Dye Channel.
Dye
Minimum
RFU
Maximum
RFU
Average
RFU
Standard
Deviation
(SD)
Average
RFU + 3
SD (LOD)
Average
RFU +
10 SD
(LOQ)
MTS
(1)
MTS
(2)
Analytical
Threshold
B 1 65 5.59 3.56 16.26 41.14 45 128 130
G 1 72 9.86 4.75 24.12 57.39 60 142 145
Y 2 74 11.16 4.91 25.89 60.26 65 144 145
R 2 72 12.95 5.59 29.73 68.87 70 140 140
MTS (1): minimum threshold, LOQ rounded up to nearest five
MTS (2): analytical threshold, 2*(max RFU - min RFU)
Analytical Threshold: MTS (2) rounded to nearest five
0.00
5000.00
10000.00
15000.00
20000.00
25000.00
0.5 ng 0.25ng 0.125 ng 0.0625 ng 0.03125 ng 0.0156 ng
AveragePeakHeight(RFU)
Input Target (ng)
Average Peak Height (RFU)
K05
K07
K08
25. Page 25 of 51
3.4 Stochastic
Full DNA profiles were obtained at input levels of 31pg and 15pg as indicated in the
sensitivity study. The majority of the samples were analyzed at a target input of 31pg resulted
in full profiles. For the PowerPlex® Y23 system, the majority of the loci are expected to have
only one allele call per locus. The DYS385 is the only locus expected to result in the presence of
two alleles, which can manifest as a heterozygotic pair of alleles. All of the samples used in this
study were heterozygous at the DYS385 locus and no dropout was observed at the 31pg DNA
input target. There were only two observed dropout occurrences at the DNA input of 15pg
(Table 3.4.2). There was dropout of allele 14 for one of the K08_0.0156ng samples; however
there is a peak present at the 14 but it is below the analytical threshold (Figure 4). The second
dropout occurrence did not show any sign of the missing allele, which was allele 18 in the
profile of sample K10_0.0156ng (Figure 5). Additionally, there were several profiles with
severe peak imbalances present at the DYS385 locus (Figure 6).
Table 3.4.1. Average PHR and the Minimum PHR for DYS385 locus for All Stochastic Samples
Input Target % Average of
PHR
% Min of
PHR
0.0156 ng 63% 37%
0.03125 ng 61% 18%
Table 3.4.2. Two Instances of Drop Out (purple) for 15pg and Alleles Called at 31pg
DYS385
Allele
Peak Height
0.03125 ng (RFU)
Peak Height
0.0156 ng (RFU)
K08 13 1050 606
14 586
13 1440 582
14 1861 723
13 1191 656
26. Page 26 of 51
14 1193 242
K10 13 602 356
18 1056
13 946 384
18 811 527
13 1903 302
18 872 309
Figure 4. Dropout of the 14 Allele at DYS385 Locus for Sample K08 at 15pg
27. Page 27 of 51
Figure 5. Dropout of Allele 18 at DYS385 Locus of Sample at 15pg
Figure 6. a) One example of peak imbalance at the DYS385 locus. b) Second example of a peak
imbalance at the DYS385 locus
a). b).
28. Page 28 of 51
3.5 NIST
The NIST standards were extracted and quantified according to laboratory protocol, and
amplified with the PowerPlex® Y23 kit. Only the NIST samples A-D were utilized and necessary
for the validation studies (11). SRM Component A was a female sample and therefore no
profile was obtained, as expected. SRM Components B, C and D all contained male DNA and
the expected profiles were obtained. Component D was a mixture of a female (SRM A) and a
male (SRM C) and only the male profile was obtained which correlates with the profile obtained
from SRM C (Table 3.5.1).
Table 3.5.1. Allele Calls for NIST Standards Containing Male DNA
Item
Number SRM_B SRM_C SRM_D
Markers Allele 1 Allele 2 Allele 1 Allele 2 Allele 1 Allele 2
DYS576 17 - 16 - 16 -
DYS389 I 13 - 12 - 12 -
DYS448 20 - 19 - 19 -
DYS389 II 31 - 27 - 27 -
DYS19 14 - 15 - 15 -
DYS391 10 - 11 - 11 -
DYS481 25 - 26 - 26 -
DYS549 12 - 13 - 13 -
DYS533 11 - 10 - 10 -
DYS438 10 - 11 - 11 -
DYS437 14 - 16 - 16 -
DYS570 18 - 20 - 20 -
DYS635 20 - 21 - 21 -
DYS390 23 - 24 - 24 -
DYS439 11 - 12 - 12 -
DYS392 11 - 13 - 13 -
DYS643 9 - 12 - 12 -
DYS393 12 - 13 - 13 -
DYS458 17.2 - 17 - 17 -
DYS385 13 17 13 15 13 15
DYS456 15 - 15 - 15 -
YGATAH4 11 - 11 - 11 -
29. Page 29 of 51
3.6 Non-probative/ Mock
Multiple non-probative and mock samples yielded no male profiles. Many additional
samples were incomplete with very few full male profiles obtained. The average number of
alleles called per sample was calculated to be approximately 20 alleles called per sample (Table
3.6.1). This did include the samples that contained known mixture samples and possible
mixture samples. Of three semen dilutions tested, only one yielded a partial DNA profile, while
the other two yielded 1 or 2 loci with low RFU values. The three saliva dilutions yielded the
expected profiles with alleles at strong RFU values. From the adjudicated case, only a partial
profile was obtained from the panties cutting, and no other results were obtained. One of the
differential mixture samples did reveal a full male profile, but also indicated possible
degradation. (Figure 7). The three proficiency tests resulted in full profiles with reportable
results. Table 3.6.1 shows the alleles called per sample and the total alleles called for all of the
non-probative samples including the mixture and differential samples.
Table 3.6.1. Average Allele Calls Per Sample and Total Average Allele Calls
Row Labels
Count of
Allele 1
Count of
Allele 2
NP01-AS 10 NONE
NP02-AS 4 NONE
NP03-AS 3 NONE
NP05-NS 6 NONE
NP07-NS 20 3
NP07-SP 6 NONE
NP08-NS 21 3
NP08-SP 22 1
NP09-NS 22 1
NP09-SP 11 NONE
NP10-NS 22 5
NP10-SP 15 6
NP11-NS 22 4
NP11-SP 22 5
30. Page 30 of 51
NP13 22 6
NP14-AS 22 3
NP16 22 4
NP17-NS 22 5
NP17-SP 22 4
NP19 22 6
NP20-NS 22 3
NP20-SP 22 4
NP21-AS 4 NONE
NP22-AS 1 NONE
NP23-AS 2 NONE
NP24-AS 22 1
NP25-AS 21 1
NP26-AS 22 1
Grand Total 454 66 Average Per Sample
Average Per Sample 16.21 3.47 19.69
Figure 7. NP08 Sperm Fraction Sample in the Yellow Dye Channel
3.7 Mixture Studies
Mixture A
The extremely high constant concentration of female did not interfere with the
amplification and subsequent interpretation of the male component for any of the varying
31. Page 31 of 51
input male DNA template amounts, even when the male contributor was as low as 15pg (Figure
8). Full male profiles were obtained for all mixture samples.
Figure 8. Female to Male Mixture Ratio 250ng Female to 15pg Male in the Blue Dye Channel
Mixture B
The increasing female to constant male mixture ratios also showed no interference of
the female DNA when attempting to obtain male DNA profiles. Full male DNA profiles were
obtained from each mixture tested (Figure 9). The mixture B set was also amplified using the
Identifiler® Plus Amplification kit for autosomal DNA analysis in order to compare the results
strictly for the male profiles for each system. Figure 10 shows the Identifiler® Plus results(top),
which includes the female autosomal DNA with the small amount of male autosomal DNA,
versus the Powerplex® Y23 results(bottom) with only the Y-STR profile of the male. The
Identifiler® Plus results showed the major contributor as female and dropout of the male at
most loci.
32. Page 32 of 51
Figure 9. Female to Male Mixture Ratio 49F:1M in the Blue Dye Channel
Figure 10. Mixture B Ratio 49F:1M Results of the Green Dye Channel for Identifiler® Plus (top)
and Powerplex® Y23 (bottom)
33. Page 33 of 51
Mixture C
Male to Male mixtures with ratios of 1:1 revealed that the difference in peak heights
between the contributors corresponded well with the expected contribution of each person.
(Figure 11). The data showed that the peak heights were representative of major and minor
contributors for all ratios. However, the peak height ratios were not completely representative
of the input ratios as the ratio difference increased (Figure 12 and Figure 13). This trend
occurred across all loci and Table 3.7.1 demonstrates the ratio differences at three of the loci
that were included in these calculations.
Figure 11. Male to Male Mixture Ratio 1:1 in Blue Dye Channel
Figure 12. Male to Male Mixture Ratio 1:9 in Blue Dye Channel
34. Page 34 of 51
Figure 13. Male to Male Mixture Ratio 1:19 in Blue Dye Channel
Table 3.7.1. Actual Peak Height Ratios versus the Expected PHR of Mixture C per Allele (at three
of the loci)
Contributor
(Average RFUs)
Marker Ratio (A:B) Allele A B Actual Ratio Expected Ratio
DYS389 I
1:19
12 4783.333
0.050 0.050
13 241.000
1:9
12 4159.667
0.103 0.110
13 428.333
1:3
12 4202.333
0.371 0.330
13 1557.000
1:1
12 3214.667
1.203 1.000
13 3866.000
3:1
12 1485.333
2.487 3.000
13 3693.667
9:1
12 535.333
6.769 9.000
13 3623.667
19:1
12 545.333
8.482 19.000
13 4625.333
DYS448
1:19
20 581.333
0.084 0.050
21 6921.333
1:9
20 934.000
0.155 0.110
21 6015.667
1:3
20 2566.000
0.477 0.330
21 5382.667
1:1
20 6399.667
1.242 1.000
21 5152.000
35. Page 35 of 51
3:1
20 5732.667
3.373 3.000
21 1699.333
9:1
20 6548.000
12.402 9.000
21 528.000
19:1
20 6610.333
26.266 19.000
21 251.667
DYS389 II
1:19
28 6422.333
0.046 0.050
30 294.000
1:9
28 4999.000
0.088 0.110
30 440.667
1:3
28 4452.333
0.358 0.330
30 1594.333
1:1
28 3274.667
1.077 1.000
30 3527.667
3:1
28 1495.333
2.525 3.000
30 3776.333
9:1
28 688.000
5.955 9.000
30 4097.333
19:1
28 491.333
9.921 19.000
30 4874.667
Mixture D
In the mixture D set, there was drop out present in several profiles at both 15pg and 31pg
input amounts across all ratios. Figure 14 shows that at low template amount, the peak heights
are too similar to be able to distinguish between major and minor contributors and that the
ratios were not preserved at these amounts. The mixture ratios at low template amount of
input DNA from the Mixture C studies exhibited the phenomena referred to as “flip flopping”;
which refers to instances where the expected major contributor’s alleles is actually smaller than
the allele of the expected minor contributor(12). Figure 15 exhibits the “flip flopping” of the
major contributor at the DNA input amount of 31pg versus the target input amount from
Mixture C set. The male to male ratio is 19:1 however, the 16 allele from the expected minor
contributor has a greater peak height than the major contributor which is the 17 allele.
36. Page 36 of 51
Figure 14. Mixture D Ratio 3:1 at 15pg input amount
Figure 15. Occurrence of “Flip Flopping” from Optimal DNA Input Amount (left) and Stochastic
Input Amount of 31pg (right).
3.8 Robustness of Amplification Product
In both the amplification product and 2800M stability studies, all DNA profiles obtained
contained the expected results. In the majority of the samples, week 8 showed the highest
RFU values when compared to previous weeks (Figure 16 and 17). The 2800M stability study
showed that there was some degradation of the first 2800M dilution prepared. The RFU values
37. Page 37 of 51
were much lower in the oldest dilution, but the RFU values were still largely above the
stochastic levels (Table 3.8.1).
Figure 16. Sample K01 Average RFU Values at Each Loci for Weeks 1 through 4
Figure 17. Sample K01 Average RFU Values at Each Loci for Weeks 5 through 8
0
5000
10000
15000
20000
25000
DYS576
DYS389I
DYS448
DYS389II
DYS19
DYS391
DYS481
DYS549
DYS533
DYS438
DYS437
DYS570
DYS635
DYS390
DYS439
DYS392
DYS643
DYS393
DYS458
DYS385
DYS456
YGATAH4
AveragePeakHeight(RFU)
Locus
K01 Reproducibility Summary
Week 1
Week 2
Week 3
Week 4
0
5000
10000
15000
20000
25000
DYS576
DYS389I
DYS448
DYS389II
DYS19
DYS391
DYS481
DYS549
DYS533
DYS438
DYS437
DYS570
DYS635
DYS390
DYS439
DYS392
DYS643
DYS393
DYS458
DYS385
DYS456
YGATAH4
AveragePeakHeight(RFU)
Locus
K01 Reproducibility Summary
Week 5
Week 6
Week 7
Week 8
38. Page 38 of 51
Table 3.8.1. Average Peak Heights across all Loci for each 2800M Dilution
Date Prepared
Average Peak Height
(RFU)
5.26.2015 3519.565
6.10.2015 7068.826
6.18.2015 7963.333
7.1.2015 7957.58
7.8.2015 9670.29
7.14.2015 7397.319
7.22.2015 10046.04
4. Discussion
4.1 Precision Studies
The alleles for the allelic ladder must fall within +/- 0.5 base pair (bp) window set by
GeneMapper® ID-X Software in order to satisfy the acceptable degree of precision. A standard
deviation of 0.15 bp or less must be met for precision studies, to allow for error where the
sample allele would acceptably fall outside the base pair window. The Promega® Internal
Validation Guide of Y-STR Systems in Forensic Laboratories recommends that three times the
standard deviation of each allele be less than 0.5 base pairs (13). The calculated standard
deviations for all of the 96 allelic ladders for the precision study were all below the 0.15 bp
mark. The highest standard deviation that was observed over the 24 capillaries as well as all 4
injections was 0.1273 bp which is less than the maximum 0.15 bp recommended for data
analysis. The precision study demonstrated the ability of the PowerPlex® Y23 system to
produce accurate and reliable results.
39. Page 39 of 51
4.2 Sensitivity Studies
The sensitivity results demonstrated the increased sensitivity of the PowerPlex® Y23
system which shows that even as low as 15pg input DNA a full profile can be obtained. The
average peak height ratio (PHR) for all of the samples over all of the input levels was over 50%
where the lowest average PHR for an individual sample was 64% at 31 pg and the lowest
average PHR at the 15 pg input level was 67% (Table 3.2.1). The lowest PHR of the DYS385
locus obtained for the optimal DNA target input amount of 0.5 ng was 86% and for the 0.25 ng
and 0.125 ng input amounts the lowest expected PHR is approximately 63%. The imbalance in
the peak heights at the DYS385 locus could become problematic when distinguishing between
male contributors in mixture samples. The peak height ratios only apply to the DYS385 locus,
which is the only locus with a possibility of 2 alleles. In order to determine the range of DNA
input amounts that would obtain reliable results, the percentage of complete profiles obtained
at each input amount was observed (Table 3.2.3). All input amounts from 0.5 ng to 62.5 pg
obtained 100% complete profiles. Only the 31 pg and 15 pg input amounts resulted in less than
100% complete profiles. Therefore, based on the data obtained, the PowerPlex® Y23 system
can reliably produce full single source profiles down to 62 pg input amounts of DNA. The only
drop out that occurred at any loci was at the 31 pg and 15 pg amounts. There was no dropout
of the sample K08 alleles and very few dropout of the sample K07 alleles. The most drop out
occurred for the sample K05, and respectively at the loci, DYS391, DYS438, DYS392, and
DYS643. However, the only corresponding dropout for the K07 sample occurred at the DYS643,
which still had minimum drop out. There is no excessive drop out at any specific loci, but there
was also minimum drop out throughout the entire study. It is recommended that more studies
40. Page 40 of 51
should be performed in order to determine if there is a trend present for drop out at specific
loci.
4.3 Minimum and Analytical Threshold
The minimum and analytical thresholds are calculated to determine at what RFU value a
peak can be considered a true allele or artifact and not just background noise. No one channel
contained more background noise than the others and the background noise was below 80
RFUs for each dye. The red, yellow, and green dye channel did have slightly higher average
heights (RFU) than the blue dye channel, which had the lowest average (Table 3.1.1). The
minimum threshold was calculated using the limit of quantitation (LOQ) then the analytical
threshold was calculated from the LOQ values for each dye except the Orange dye channel
(CC5). The maximum calculated analytical threshold was 145 RFU in the green and yellow dye
channels; this value was rounded up to 150 RFU, for simplicity, for an overall analytical
threshold across all dye channels.
There were a few reoccurring artifacts present throughout the study. One artifact was
not mentioned in any literature or previous studies and was recorded in the green dye channel
at around 70 base pairs (Figure 19). No artifacts acknowledged by Promega® for the
PowerPlex® Y23 system were observed in the studies during the internal validation process.
There were some artifacts listed in the internal validation study performed by The Wisconsin
Crime Laboratory Bureau (12). However, none of these artifacts were observed during the
validation process. Throughout the internal validation process, there was a large amount of
stutter observed and the Promega® guidelines for the stutter percentages were followed (1).
41. Page 41 of 51
Further studies should be performed in order to develop internal lab protocols specific to
stutter.
Figure 19. Reoccurring Artifact in the Green Dye Channel of MTS Study
4.4 Stochastic
Due to the consistency and sensitivity of the PowerPlex® Y23 system, minimal data was
obtained which could be utilized to assess a stochastic threshold for the Y23 amplification kit.
The two instances of dropout were both at the input amount of 15pg. Because there was a lack
of dropout at the DYS385 locus for input target amounts of 31 pg and 15 pg, more studies need
to be performed at lower target amounts to come to a more accurate and conclusive set
stochastic threshold. Of the two drop out occurrences, one for sample K08 was not a true drop
out, due to presence of the dropped out allele that was not called because it was lower than
42. Page 42 of 51
the analytical threshold. The second occurrence of drop out was for sample K10 and was
considered a true drop out and the sister allele that remained present had a peak height of 356
RFU. The sister allele remaining for the K08 dropout occurrence had a peak height of 606 RFU.
The sensitivity of the PowerPlex® Y23 system indicates that male input DNA quantities can
obtain full profiles lower that of autosomal STR systems. From the sensitivity studies, it was
determined that full male profiles may be obtained from input amounts as low as 15 pg.
Therefore, the lack of dropout at the DYS385 locus is a result of the ability of the kit to detect
low input template amounts of DNA. It is recommended that for a stochastic threshold to be
determined, more studies be performed at the low input DNA template amounts of 15 pg and
at even lower amounts. However, based on the validation data obtained a stochastic threshold
would be within the range of 150 RFU to 625 RFU.
4.5 Contamination
Throughout the entire validation, contamination was not observed in any of the positive
or negative controls used throughout the validation. There was a possible contamination of
sample NP07-NS sample, where there was an extra allele, 22, present at the DYS570 locus.
There was an allele 22 called for the NP07-SP sample at the DYS570 locus, which could have
contaminated the NP sample during preparation. However, the SP sample was not located near
the NP sample in the 96 well plate. Therefore, the presence of the extra allele could be due to
a mixture present in the original sample and that the differential extraction did not effectively
separate the two contributors. These results demonstrate the sensitivity of the system without
contamination issues. It should be noted that the preparation of all of the samples was either
43. Page 43 of 51
automated or performed by a female. Due to the noted high level of kit sensitivity, it is
recommended that more samples be prepared and analyzed by a male analyst in order to test
the limitations of kit sensitivity.
4.6 Non-Probative/ Mock
The results obtained from the non-probative studies were not as complete as expected due
to lack of profiles obtained. Even the partial profiles that were obtained contained very few
called alleles. Many of the samples that yielded partial profiles had a zero quantitation value in
the Quantifiler® Duo results, which demonstrates that the quantitation results does not reflect
the sensitivity of the amplification kit and its ability to produce at some results. The NCSCL
protocols require analysis to continue even if a quantitation results in zero DNA, for autosomal
STR analysis which would also be implemented for YSTR testing. These samples included the
aspermic postcoital samples which observed 4-9 alleles for each sample, and a spermic sample
of a two person mixture which observed a partial profile of 11 alleles called. There were also
very low quantitation result samples (below 15pg) that resulted in a full DNA profile. A two
person male to male mixture sperm fraction from sample NP08 with a quantitation value of
0.0113 ng and another two person male to male mixture nonsperm fraction from sample NP09
with a quantitation value of 0.0051 ng, both resulted in full single source profiles. The observed
full profiles could be a combination of the two contributors, but the quantitation values are
extremely low and raise a concern with the correlation between the Quantifiler® Duo results
and the profiles obtained from the Powerplex® Y23 system. Because the Y23 system is
44. Page 44 of 51
extremely sensitive, there could be several low male DNA amounts that would have a zero
quantitation value but yield at least a partial profile.
The adjudicated case samples also resulted in partial and very few allele calls, but the alleles
that were called were in concordance with the YSTR results previously obtained with Applied
Biosystems® YFiler® kit in 2008. The panties stain resulted in an almost full profile with only
two loci failing to amplify. The original results collected from the adjudicated case yielded full
results, therefore the lack of profiles could be a result of DNA degradation or that the majority
of the sample was consumed for the first analysis with autosomal STRs followed by the previous
YSTR analysis in 2008. The adjudicated case samples had already been partially used for a
previous validation and there was limited sample available for the current validation. Further
studies using samples representative of true casework should be performed due to the limited
results collected.
The saliva dilution profiles resulted in the expected full profiles; however the semen
dilutions had no observed profiles. The semen dilutions were 1:100, 1:150, and 1:200 while the
saliva dilutions were 1:50, 1:75, and 1:100 and were both measured on sheer volume and not
measured in quantity or concentration. The dilutions for both semen and saliva were tested
with Rapid Stain Identification Series (RSID™) prior to YSTR analysis. The only alleles observed
in the semen dilution samples were present in 1:200 dilution, which raises some questions as to
why the lower volume of DNA resulted in more allele calls than the higher volume of DNA. This
could be a result of human error in the preparation of the sample or the preparation
throughout the analysis process. More semen as well as saliva dilution samples should be
studied in order to answer those questions and establish a more definitive conclusion. Further
45. Page 45 of 51
studies should also be performed in conjunction with the Forensic Biology/ DNA section body
fluid presumptive tests in order to help construct a workflow and cutoff point for YSTR analysis
for the implementation in casework. It is also recommended that more post-coital samples be
utilized for representation of mock evidence samples. Other possible representative evidence
samples could include saliva and semen mixture samples and could benefit the Non-probative
and mock studies.
4.7 Mixture Studies
Full male profiles were obtained for all mixture samples in mixture study A and B. This
demonstrates that large amounts of female DNA as high as 1:16,000 ratio will not inhibit or
interfere with the PowerPlex® Y23 amplification system. In sexual assault kits or other cases
where female victim DNA is likely to be the major contributor in the sample, a male profile
could still be obtained at lower input amounts when using the Y23 amplification kit. Mixture B
samples were also amplified with the autosomal amplification system Applied Biosystems®
Identifiler® Plus where the quantities of female DNA overwhelmed the minor contributing male
DNA and drop out of the male contributor was observed. However, with the PowerPlex® Y23
amplification, full profiles were obtained for minor male contributors at all mixture ratios. The
49:1, female to male mixture resulted in zero alleles called from the Identifiler® Plus results
however, a full male profile was obtained from the PowerPlex® Y23 results. The PowerPlex®
Y23 system could also be beneficial for cases where a consensual partner was involved beyond
the traditional 72 hour collection window for rape kits.
46. Page 46 of 51
Mixture C results revealed the extreme sensitivity of the Y23 system but also resulted in
unexpected contributor ratios per locus (RFU Values). The RFU values in the mixture ratio
samples 9:1 and 19:1 consistently deviated from the expected ratio by over 50%. This was
observed throughout all of the samples per loci. For example, an expected ratio of 19:1 would
result in a calculated ratio of 8.5:1 (Table 3.7.1). The inverse ratios of 1:19 and 1:9 did not show
this much variation from the expected results. This may be a result of human error in
preparation of the mixtures but more samples should be prepared and studied before any
conclusions should be made. This possible error in preparation would also affect the results
from the mixture D studies.
Mixture set D resulted in profiles with unexpected intensities of the alleles that were not
always representative of the major and minor contributor ratios. In the low DNA input
amounts of 15 pg, there were some instances of “flip flopping” of the alleles (12). This could be
difficult for analysts being unable to confidently distinguish between the contributors within a
profile. In conclusion, the validation data did not support the assessment of unambiguous
major and minor haplotypes in this amplification range. Therefore, it is recommended that
more studies be performed in assessing male to male mixture at low DNA input amounts.
4.8 Robustness of Amplification Product
The signal intensity of each sample remained between 7500 and 11000 RFUs for the
overall average peak height (all alleles at all loci) for the entire study. The data showed no
obvious pattern throughout the study. The average peak heights across all samples and loci
were calculated which included 24 samples in the 96-well plate for each week. The average
47. Page 47 of 51
peak height ratio from week 1 to week 8 runs was 85% and from week 5 to week 8 the average
peak height ratio was 87%. This ratio is still greatly above the acceptable peak height ratio,
which suggests that amplification product stored at the recommended conditions of 4°C can
still result in acceptable results with proper running, preparation conditions, and barring visible
observation of evaporation of the amplification product in the plate. The observed signal
intensity throughout the study was expected to decrease overtime, however the actual results
showed that the intensity fluctuated slightly but maintained at a constant level. Week 8
showed about a 25% increase in the signal intensity. This might have been a result of
evaporation of the amplification product causing the fluorescence to become more
concentrated, which resulted in higher signal. Amplification products should be studied over a
longer period of time to come to a more definitive conclusion on the robustness of
amplification product. It would also be recommended to assess the stability of the
amplification product for low copy number samples.
The 2800M dilutions can result in accurate and complete profiles after being stored for
about 8-9 weeks. The average peak height difference between the first dilution and the last
dilution was 6527 RFUs, resulting in a peak height decrease of approximately 65%. The data
indicates that there is a steady decrease of the peak heights over time. Proper interpretation of
the positive control 2800M is still possible despite the decrease in the RFU values after the total
9 week period between the least recent and most recent dilution. However, it is recommended
that more studies should be performed testing the 2800M dilutions for longer periods of time
to see how long a 2800M dilution may be stored and used to for casework purposes. The
48. Page 48 of 51
2800M stability demonstrated that a dilution may be prepared, stored and used for nine weeks
and still produce strong, full profiles.
4.9 Reproducibility
The reproducibility of the PowerPlex® Y23 Amplification kit was demonstrated through
the repeated amplification of the known standards as well as the 2800M positive amplification
control. All of the known standards resulted in all of the same expected profiles for each
repeated analysis. The same results occurred for all of the amplification positive controls that
were analyzed from the various studies throughout the validation process.
5. Conclusions
A thorough review of the internal validation data revealed a set analytical threshold at
150 RFU values and demonstrated the high sensitivity of the kit. A stochastic threshold could
not be confidently identified from the validation study due to more studies needed to be
performed at lower DNA input amounts. Based upon the validation data obtained, it may fall
within a range of 150 RFU to 625 RFU. More data should be collected at the 31 pg and 15 pg
input amounts of DNA and possibly run samples at even lower input amounts to determine the
stochastic threshold of the Y23 System. The mixture studies revealed a strong confidence in
detecting small amounts of male DNA when amplified in the presence of large amounts of
female DNA without any interference or inhibition.
This internal validation demonstrates the potential benefit of implementing the
PowerPlex® Y23 kit in other forensic casework laboratories and will assist the North Carolina
State Crime Laboratory’s Forensic Biology section in evaluating the addition of Y-STR analysis in
49. Page 49 of 51
the processing of sexual assault evidence. The NCSCL Biology Section will perform future
studies regarding stutter values, half-reactions, and other mixture studies that reflect actual
casework samples. Other studies will be designed to establish the correlation between high
mutation rate loci and related male samples. The implementation of the PowerPlex® Y23
system will expand the testing capabilities of the Forensic Biology section.
50. Page 50 of 51
References
1. Promega® PowerPlex® Y23 System Technical Manual, March 2015
2. Scientific Working Group of DNA Analysis Methods (SWGDAM). Quality Assurance
Standards for Forensic DNA Testing Laboratories. September, 2011. Available:
http://www.swgdam.org/FBI%20Director%20Forensic%20Standards%20%20Revisions%
20APPROVED%20and%20Final%20effective%209-01-2011.pdf.
3. Applied Biosystems. Applied Biosystems 3500/3500xL Genetic Analyzer User Guide. Rev
06/2010.
4. “SWGDAM Interpretation Guideline for Y-Chromosome STR Typing by Forensic DNA
Laboratories.” Scientific Working Group on DNA Analysis Methods. 9 January 2014.
5. “Frequently Asked Questions (FAQs) on the CODIS Program and the National DNA Index
System,” Federal Bureau of Investigation: Laboratory Services. Available:
https://www.fbi.gov/about-us/lab/biometric-analysis/codis/codis-and-ndis-fact-sheet.
6. Vermeulen M, Wollstein A, van der Gaag K et al. (2009) Improving global and regional
resolution of male lineage differentiation by simple single-copy Y-chromosomal short
tandem repeat polymorphisms. Forensic Science International: Genetics, 3, 205–213.
7. “Promega® PowerPlex® Y23 System” Available:
http://www.promega.com/products/genetic-identity/str-amplification/5-dye-y-str-
analysis/
8. “ISO Procedures.” North Carolina State Crime Laboratory. Available:
http://www.ncdoj.gov/About-DOJ/Crime-Lab/ISO-Procedures.aspx.
51. Page 51 of 51
9. “SWGDAM Validation Guidelines for DNA Analysis Methods.” Scientific Working Group
on DNA Analysis Methods. December 2012.
10. Federal Bureau of Investigation. “Quality Assurance Standards for Forensic DNA Testing
Laboratories.” July, 2009. Available: http://www.fbi.gov/about-us/lab/biometric-
analysis/codis/qas_testlabs
11. Kline, Margaret C., Butts, Erica L.R., Hill, Carolyn R., Coble, Michael D., Duewer, David L.,
and Butler, John M. “The New Standard Reference Material® 2391c: PCR-based DNA
Profiling Standard”. U.S. National Institute of Standards and Technology.
12. Buscher, A., Zastrow-Arkens, S., Kauraka, D., Hinton, N., Culhane, S., and Degroot, G.
“Internal Validation and Implementation of the PowerPlex® Y23 System. Wisconsin State
Crime Laboratory Bureau.
13. Promega® Internal Validation Guide of Y-STR Systems for Forensic Laboratories.
November 2012.
14. Maryland State Police, Validation Summary: Y-STR DNA Profiling Using PPY23. January
2014.