Breakthroughs in Forensic DNA Analysis Using Next-Generation Sequencing
The growing use of NGS in forensic genetic applications has resulted in the analysis of alternative markers that can complement CE-STR results for challenging samples- bones, mixtures, or degraded samples- to generate investigative leads. Marta Diepenbroek will discuss her collaborative work with colleagues from Poland and Austria, which concerns the first genetic and phylogeographic evidence for Jewish Holocaust victims found at the former Nazi death camp in Sobibór. Diepenbroek will discuss show how forensic analysis of haploid markers can shed light on historical events. Jesse Cox will describe the underpinnings of an approach to deconvolute mixtures of two individuals to provide ancestry and phenotypic information via NGS for both a major and minor contributor. In this webinar you’ll hear about: NGS analysis of the mitochondrial genome can be combined with CE-based Y STR markers to enable lineage determination of remains SNP analysis by NGS is used to deconvolute mixtures for investigative lead generation
Hello, and welcome to Breakthroughs and Forensic DNA Analysis using Next
Generation Sequencing,
brought to you by Forensic and sponsored by Thermo Fisher Scientific.
This is the fifth and final webinar in the 2021 Future Trends and Forensic DNA
Technology
Series.
My name is Michelle Taylor, Editor-in-Chief of Forensic and I will be your
moderator throughout.
For today's webinar, you can earn one hour of continuing education credit.
Following the conclusion of the webinar, you will receive an email with
information on
how to obtain CE Credit Documentation.
We have a great lineup scheduled to present to you today, but before we begin,
I'd like
to take just a moment to cover a few logistics.
At the end of the presentation, we will hold a question and answer session.
To ask a question, click on the "Ask a Question" tab in the upper right corner
of
your screen.
Please also take note that the right side of the screen features an overview of
today's
webinar as well as more information about our speakers.
If you have a technical question during today's event, click on the "Test your
Connection"
button on the bottom of your screen.
From there, you can access additional webinar support.
We also invite you to use a social media widget beneath the webinar to share
with your friends
and colleagues.
Today, you will hear from Marta Deepenbroke, a post-doctoral scholar at the
Ludwig Maximilian
University in Munich.
Marta will discuss her collaborative work with colleagues from Poland and
Austria concerning
the first genetic and phylogeographic evidence for Jewish Holocaust victims
found at the
former Nazi death camp in Sobobor.
For the past six years, Marta was a team member of the Polish genetic database
of victims
of Tolitarianisms, a DVI project with the purpose of employing forensic methods
to identify
victims of communist and Nazi regimes in Poland.
Currently, her target is the practical use of state-of-the-art forensic methods
in routine
investigations and presenting them to law enforcement.
Her work focuses on the newest MPS-based solutions.
We'll also hear from Jesse Cox, assistant professor at the University of
Nebraska Medical
Center.
Jesse will describe the underpinnings of an approach to de-combolute mixtures
of two individuals
to provide ancestry and phenotypic information via NGS for both a major and
minor contributor.
Jesse completed his AP/CP Pathology Residency and Molecular Genetic Pathology
Fellowship
training at the University of Nebraska Medical Center in Omaha.
He currently serves as director of the Human DNA Identification Laboratory and
HLA Tissue
Typing Laboratories at UNMC.
Thank you for joining us for the final webinar in the 2021 Future Trends in
Forensic DNA
Technology Series, and we hope you have enjoyed this series as much as we have.
Be sure to check your email earlier in the new year, because we will be
continuing this
series.
Now, without further ado, let's get underway.
Welcome, everyone.
My name is Matha Deepindrok.
I currently work as a product researcher in forensic genomics and the Institute
of Human
Disease at LMEU.
Today, it is my pleasure to present the results of a collaborative work I would
call it from
Poland and Austria.
And during my presentation, I will discuss how our genetic and phylogenographic
analysis
shed light on the historical events and the solvable death camp, because I
believe that
as forensic scientists for the death and the living, we must bear witness.
According to the last year's survey, conducted among young Americans, almost 70
% of them
are not aware how many Jews were killed during the Holocaust and what is more
some stated
that it is actually Jews who caused the Holocaust.
Almost 50% of the respondents were not able to name a single concentration camp
and almost
a quarter of them called the Holocaust a myth.
The Holocaust was the systematic bureaucratic state sponsored by the
persecution and murder
of 6 million Jews by the Nazi regime and its collaborators.
It was practiced by locking Jews in gaztos, by executing them in 12th ground,
by forcing
them to work in labor camps and by mass killing them in death camps.
The last were the most secret yet the deadliest form of the Holocaust.
In contrast to the better known concentration camps, like for example Auschwitz
, dead camps
served only one purpose.
They were created to carry mass scale mergers.
People who arrived at them were sent direct people gaztrenders and the
existence of these
places was kept secret even within the Nazis.
Therefore, so Ivar is a place that many never heard about.
It was built in spring 1942 as one of the three camps operating as a part of a
secretive
opposition, Reinhardt, which was a plan to kill most Polish Jews in the general
government
district.
This single mass murder campaign took lives of almost 2 million people.
In Sobibor itself almost 200,000 people were killed.
A group of resistance organized a revolt which led around 300 prisoners to
freedom.
As an aftermath the camp was closed and completely demolished, leaving no trace
of its existence.
Only 58 of those who escaped survived the world.
Due to the fact that the Nazis destroyed any evidence of Sobibor's presence,
the only
sources that could testify about the place where the aforementioned survivors
and the
guards who walked at the camp.
However, both sources were unfortunately not fully reliable.
The prisoners who escaped from Sobibor never saw with their eyes the deadliest
part of
it, so-called "laga trite" or in English "camp 3".
Namely the place where the gas chambers were located and where the bodies were
burnt and
burnt.
During the revolt, prisoners operating this part of the camp would not be
informed about
the upcoming events and therefore none of them survived the escape.
And as for the guards, the sincerity of their testimony could naturally be
caught in the
question.
This slide shows an aerial photo taken by Luftwaffe almost a year after the
camp was destroyed.
As you can see, except over a few buildings that could not be linked to any
criminal activities,
the place was torn to the ground.
Based on the available testimonies, the potential layout of the camp was
reconstructed.
However, new evidence still keeps coming to the light.
Just last year, some private photos from one of the Nazi guards working at the
camp were
found hidden at the attic.
The pictures were taken mostly during everyday activities and they show only
the Nazi residential
buildings, but even there, the historians were able to spot important evidence.
Like for example, here in the background we can see an excavator which was used
to dig
out the already buried decaying bodies in order to burn them as it was required
by their new
regulations.
Therefore, in order to learn more about the camp's layout, archaeological work
was carried
out at the place, starting in the year 2000.
It led to the discovery of mass graves with cork wax and human ashes and
archaeological
signs with the potential relics of the gas chambers.
The excavations continued as non-invasive workers and later in 2007, systematic
excavations
began which led to the discovery of the lagatite relics.
Due to a planned commemoration of the victims of the camp, further work at the
camp 3 was
carried out.
And this led to the discovery of foundations that undoubtedly belonged to the
former gas
chambers.
However, surprisingly, also intact human skeletons were revealed.
Here is a picture taken by the archaeologists working in the camp showing the
view on the
lagat and rice prepared for the commemoration of the victims.
And here are all the mass graves previously located at the camp 3.
However, unexpected skeletal burials were also found on the outskirts.
And expected as according to the available testimonies, all the Jewish victims
killed
at the camp were cremated.
The pictures showed two of the four skeletal graves recovered at the camp.
On the left, on the picture seen on the right is present at Vojik Manzurak, an
archaeologist
who led the excavations and with whom we collaborated during this project.
Later with his colleagues, when analyzing the discovery, they observed that the
layers
of the burials were crossing the camp ground players, which suggested that the
victims were
buried after the end of the criminal activity of the camp.
Additionally, the historians working on the case found a report on a criminal
activity
in the camp already after the war.
The remains of the villages in the area witnessed shops which were connected to
the crimes committed
by communists urging.
Therefore, it was assumed that the remains belonged to Polish partisans, killed
secretly
in the 1950s by communists.
The investigation was taken over by the Institute of National Remembrance and
based on their
decision, the remains recollected and sent to the Polish genetic database of
totalitarian
victims, a humanitarian project led by the stretching and forensic institute.
According to the procedures used in the project, first the remains were
submitted to anthropological
and medicalical assessments.
All 10 excavated skeletons were attributed to men who were estimated to be
between 20
and 60 years old.
Our scars showed traces of injuries caused by a firearm and once teleported on
the
plate signs of injuries caused by a gunshot to the chest.
All entry ones had a diameter of between 7 and 8 millimeter.
The localization of the entry ones pointed to a systematic way of execution of
the victims.
Shorts were aimed to the back of the head or near the neck and this pattern
indicated
that the bullets were fired from a shot weapon to a victim lying down or
kneeling with the
head strongly inclined downwards.
Altogether, more than 100 other parts were recovered during the field work in
Soviet
work.
Most of these included pieces of metal and fabric, which did not carry any
specific information
about the victims, however, some objects appeared to be some personal
belongings like, for
example, a pocket knife or a leather shoe.
And three of the artifacts found during the archaeological work were submitted
to ballistic
analysis.
Those were to grasp pistol shells and one iron made pistol projectile.
Two shells were assigned to a German company producing weapons during the
Second World
War and both shells were of caliber 7.65 millimeter and could be assigned the
brownie
system, which was difficult for self-loading pistols.
These hand guns were standard equipment of German guards in concentration and
death guns.
However, they were also used by the partisans after the war.
Bon samples were submitted to their lab workflow used in the project, which
after the bon preparation
focused on the golden standard of forensic DNA typing done with capillary elect
rophoresis.
This included also haploid marker analysis, NDB1 and NDB2 of the micro-hondrel
DNA were
sequenced with sandal and because of a high degradation of the samples, hapl
otypes were
obtained for half of the individuals.
But rather these results look interesting as three of the haplotypes were
assigned by
m-pop to the K haploid group, which is not typically expected in Poland,
especially not
the particular K1A damage.
This obviously literature search, which pointed at the K1A limit, is one of the
four most common
micro-hondrel haploid groups found among modern Ashken algorithms.
Based on those findings, the white chromosome on the data was also analyzed
from phylogenetic
perspective.
We started with only heads, namely, YSCR profiles, and we used them to perform
haploid
group estimation using a Neptune predictor.
Again, for eight out of 10 individuals, the predicted haploid groups are not
typically
found in Poland.
In fact, they are rather infrequent among all Europeans and found frequent big
among the
Ashkenazi, especially the J1-h
among Ashkenazi.
And looking at the haplotypes predicted as J1A, we observed that they are
almost identical
for all four individuals.
An observation concerning this particular similarity was done by other
scientists already in the
late '90s, when they studied genetics of Jewish priests.
The group of interests were Jewish Kohanim, who are believed to be of direct
patriena
are destined from the De Bicah Aron.
Study of their white chromosome revealed that the most frequent haploid group
observed among
the group is J3-58, and that almost all of them show even the same haplotype
called later
their co-hen model haploid types.
Using all of these preliminary observations left to our decision, that our
analysis should
be expanded.
For this matter, we established a collaboration with a group from INSCROC led
by Professor
Viter Tarzinn.
We plan to extend the mitochondrial analysis with performing the whole
mitochondrial genome
sequencing using the next generation sequencing.
And for the white chromosome, we decided to design primers for the white SNPs
suggested
by Nefter and to analyze them with Sanga sequences.
And the whole mitochondrial genome sequencing was done using their precision ID
mitochondrial
DNA whole genome panel and the ion-s system.
We used multiple extracts for each individual, as they were all highly bed-
related, with
some showing less than 100 micro copies per microliter.
Nevertheless, we were successful and we obtained whole microgenomes for all 10
individuals,
which were assigned to AMPOP to eight different haplogroups.
Since the largest published, Ashkenazi mitochondrial DNA datasets are based on
controlled region
sequences, we used those to estimate relative frequencies of their haplogroups
observed
among Ashkenazi groups of different European origin and we contrasted them to
the distribution
of the same haplogroups among corresponding European populations.
The data showed that all of haplogroups established for the study individuals
and corresponded
to a lineage found in Ashkenazi while being rare or missing in other European
populations.
To do a more specific search, we queried the obtained control region haplotypes
against
AMPOP database, which we called almost 40,000 worldwide control region
sequences.
And the comparative search was also performed for the microgenome data querying
a dataset
of almost 30,000 worldwide microgenomes, downloaded from gene bank.
All individuals got matches, which consisted of or at least included samples of
non-Ashkenazi
origins.
Since the original haploteases assume that the remains belong to Polish
partisans, further
phylogenographic analysis were performed on a selected set of Polish microtypes
and on
available Ashkenazi microgenomes.
For each haplogroup observed among the study individuals, we prepared a
separate dataset
which consisted of haplotypes and interests, their Polish and the Ashkenazi
microgenomes
and we performed multiple sequence alignments in the DIGS.
This analysis revealed that four of the established haplogroups clustered with
known Ashkenazi
images and the mitochondrial DNA results of finding the study had special
meaning since
in Judaism, materialina and descent is of a special importance.
As for the paternal lineage, while the Sunday is deciding on which SNP sequence
we looked
at both national predictions and at the markers mentioned by literature.
As the data is evolving and the phylogenetic number structure is changing.
The standard sequencing revealed and derived are used at all the analyzed signs
thereby
confirming that most of the lineages observed among the study individuals are
infrequent
among Europeans, especially the J358 lineage.
Two white chromosomes were assigned to haplogroup R, which is frequently
observed in both
Ashkenazi and other European populations.
However, studies show that a particular L1A lineage, namely M at 5H2 that we
observed
now study, is extremely rare in non-Jewish populations.
As it was confirmed that four individuals assigned to J
haplogroup belong to their J58 lineage, we continued our work concerning the C
auhan
model haplotype.
The first published haplotype was based on only six YSTR markers, presented
here on the
slide, and our STR data included five of these markers, and all the observed
values were fully
recorded with the Cauhan model haplotype.
A later study in the topic resulted in extending the markers set to 12 YSTRs,
and when comparing
our data, we observed that two of four individuals still show a match with the
haplotype and interest.
But in order to perform a full comparison, we were missing a Lele data for two
YSTRs,
which are not included in any modern commercial forensic technique.
And therefore, we sequenced the markers in interest with Sangha, and as a
result, we obtained
a full match with the extended Cauhan model haplotype for two of the study
individuals,
and the remaining ones were close neighbors.
Our genetic and phylogenographic analysis were submitted to the prosecutor's
office from
the Institute of National Remembrance, who led them in the investigation
concerning the
remains.
Based on that, the responsible authorities ordered every burial of the remains.
As seen on the photo, the ceremony was led by a rabbi, and it was held
following George
Wright.
The victims were buried in separate graves at the place of their discovery.
Earlier this year, the results of this collaborative interdisciplinary study,
including specialists
from history, archaeology, anthropology, statistics, and parenthesis, were
published in genome
biology.
Since then, we continuously endeavored to promote our findings, as we agreed
with one of the
sub-uberance survivors' sets, education is the key to a better world.
I had an honor to present the results of the work of many great specialists
from different
fields, and I would like to thank all of them for collaborating on this project
Thank you for your attention.
I am happy to answer any questions.
Hello.
My name is Jesse Conchsen.
I am the director of the Human DNA Identification Laboratory at the University
of Nebraska Medical
Center in Omaha, Nebraska.
Today I'm going to discuss our recent development and work towards using next-
generation sequencing
technology to analyze a two-person mixture of ancestry and phenotype polymorph
isms, and
their decombulation to provide investigative leads.
I have no competing financial disclosures to report.
In this talk, I will discuss the reasoning for addressing this problem, the
theory and
logic behind our approach and the development of our corresponding tool, the
process we
used in validation, and lastly, how we can use this tool in order to assign
ancestry and
phenotypes to two individuals that have contributed to a single specimen.
The central question our laboratory has had was whether we could use ancestry
and phenotype
testing to help investigators bolster or refute potential investigative leads.
Using this information, could we provide our partners with potential physical
characteristics
that a person of interest may have?
Welcome to this question, and underpinning biochemistry and molecular biology
is the principle that
DNA is transcribed into RNA that is then translated into proteins.
Further, this information encoded in the standard nucleotides is both
simultaneously unique to
an individual but is passed from generation to generation.
As we are all aware, this genetic material is progressively packaged within
each nucleated
cell and inherited from mother and father.
Speaking toward the single nucleotide polymorphism and other genetic
alterations that are inherited
and at the same time make us unique, certain SNPs have been identified in which
particular
alleles are more commonly found in individuals of certain ethnicities.
In the pictured reference SNP example, the reference allele C is seen
approximately 99%
of persons of Asian descent, whereas nucleotide C is seen only about 4% of the
time in persons
of European descent.
Thus, if a C is seen at this particular SNP, the individual will be much more
likely to
be of Asian descent as compared to European descent.
With this principle in mind, assessment of SNPs across the genome with varying
frequency
and peoples of differing geographic origins may be used to imply where a
particular person
may have descended from.
In fact, this type of approach is how many of the ancestry services go about
providing
detailed ancestry data to their patrons.
To bring this type of testing in-house, we assessed at the time the various NGS
platforms
available to determine ancestry.
We identified the ion torrent S5 platform as being suitable for our purposes,
running
the Precision ID ancestry panel.
Our human DNA identification laboratory works alongside our molecular diagn
ostics laboratory
at UNMC, which was using ion torrent technology as well, giving us some
familiarity with the
strengths and weaknesses of the platform.
Further, the ion chef, Lipid Handeler, enabled West Hands-On Time for our techn
ologists so
we could better receive our workflows.
The panel we chose to assess was the Precision ID ancestry panel based off
ancestry-focused
SNPs identified and characterized by doctors' kid and selden.
From this panel, the broad regions identified and ancestry testing included
those in the
figure on screen, including America, Europe, North Africa, Middle East, Sub-
Saharan Africa,
South Asia, East Asia, as well as Oceana.
These data would then allow us to intentionally interpret the hands-sustery of
an individual
from the geographically distinct areas.
As discussed earlier, genes encoded in genomic DNA is transcribed into RNA and
subsequently
translated into protein, to building blocks that give rise to our physiological
and physical
characteristics.
To respect the question, could certain SNPs be identified which could predict
certain
traits, such as hair color and eye color?
And the obvious answer to the prior question is an unequivocal yes, and the
tools are freely
available online to perform such predictions.
Screen captures from one such tool, the Kaiser 6 blocks are shown on the screen
now.
Using six different SNPs, eye color, blue versus brown versus a mixture of the
two, can
be predicted dependent on the genotype input into the tool.
Broadening this notion out, as applicable to the platform we chose for our NGS
testing,
we identified and validated the ion, amplicity, DNA phenotyping panel to
provide phenotype
information in addition to our ancestry panel.
Based upon the work out of the Kaiser and Walsh Laboratories, the set of 17 SNP
targets
can be used to predict eye color, hair color and hair tone.
To validate NGS for use in our laboratory, 83 samples were tested over the
course of approximately
2,000 runs.
As part of our validation, we included members of our institution with varying
ethnicities,
reference materials, as well as various specimen sources.
These data were then analyzed to assess for precision and accuracy, reproduc
ibility, as
well as sensitivity.
We were also very interested in assessing whether we could deconvolute mixtures
as we
oftentimes encounter samples with more than one contributor.
To specifically assess precision and accuracy, five individuals were run
multiple times.
Resulting genotypes were at worst 97% identical between the individual runs.
Phenotypes of the ancestry is found to be 96.8% concordant with self-reported
ancestry,
with the most pronounced deviation occurring in those of Hispanic descent.
Phenotype was concordant in 93% of runs.
We next assessed sensitivity of the assay through the use of serial dilution of
input
material.
We were able to determine the optimal DNA input range to be in the 0.2-2 nan
ogram range.
Further, we assessed whether no template controls could be run to both examine
for contamination
as part of our standard workflows, as well as to help optimize our overall
workflow in
the laboratory.
Because the sequencing used, because the chips which are used for sequencing
are designed
to accommodate multiple individual samples, a balance between batching samples,
as well
as to accommodate reasonable turnaround times was an important consideration
for us.
We were pleased to determine that no template controls or blanks could be run
without affecting
overall results on a partially filled chip, which can therefore bolster both
our workflows
as well as our turnaround time.
We also wanted to confirm that a variety of specimen sources would be suitable
for this
type of testing.
Because of this, we assessed standard samples such as buckle swabs, blood, as
well as hair
roots.
We were also able to confirm that other sources would work as well, including
saliva, semen,
degraded tissues, as well as touch DNA.
These data gave us confidence that this platform would work for sources
commonly encountered
in the molecular forensics laboratory.
The last major area we were very interested in assessing and validating
concerned the
assessment of ancestry from mixtures of two individuals.
At the very least, we were interested in identifying ancestry of a major
contributor
with identification of the minor contributor being a bonus.
During our validation, we prepared mixtures of individuals at different ratios
between
one to one down to ten to one.
Using the assay supplied HID SNP genotype route, we could generally assign
ancestry to a major
contributor, especially at ratios between three to one and ten to one.
Importantly, however, the minor contributor was never able to be determined.
To illustrate why assignment of a minor contributor could not occur, we can
list how a reference
in alternate fields should be observed given various scenarios concerning the
homozygous
heterozygous status, both a major and minor contributor.
As you can see in the table to the right, we can, depending on the scenario,
determine
whether or not an individual will have either 50% between reference and
alternate if their
heterozygous or 100% for reference or 100% for alternate depending on the
status.
Now if we expand this table, and we take this scenario list and expand it, let
's suppose
in ratio of a major and minor, in this case, say an 80 to 20 ratio, we can
develop predicted
varying allele frequencies for a reference and an alternate allele to occur
given a particular
scenario.
So for example, in this 80 to 20 mixture of major and minor, where the minor
contributor
is heterozygous and the minor contributor is homozygous for the reference, we
would predict
that the varying allele frequency of the reference allele should be
approximately 60%.
We play a similar game, and we assume that the major is homozygous for the
alternate
allele, whereas the minor is homozygous for the reference allele, we would
expect the
reference allele to be present at 20% at the time.
Taking a step in the other direction, we can use the observed variant of the
allele frequency
of a reference allele determined by the equations in the rightmost column where
R equals percentage
of major to predict which scenario was in place with regards to the genotype of
the major
and minor contributor.
Using these predictions, we developed a tool that given a percentage of major
to minor
can ingest coverage data produced on our S5 and output corresponding genotypes
for a
major contributor and the minor contributor.
In the screen captures presented, a 5 to 1 ratio is analyzed, and we were able
to determine
a number of genotypes which differ between the major and the minor.
To verify our genotype deconvolution tool could reasonably perform and compare
deconvolution
results of major and minor to genotype results, and the contributors were run
as our gold
standard of a single source specimen.
We looked at the accuracy of ratios ranging from 2 to 1 out to 10 to 1.
Results are graphically represented on the left with the top line representing
the major
and the lower representing minor contributors.
The table on the right shows the number of correct alleles with the maximum
correct being
100 and maybe 7.
Also presented as the average number of alleles which were collected and I
correctly call
them along with standard deviation.
Lastly, the percentages reflect the number of alleles with a denominator of 107
Overall the decongolition tool performs best at ratios between 4 to 1 and 9 to
1 with up
to 92.5% of the minor genotype being correctly assigned.
With these data, we next want to determine whether we could ascertain useful
phenotype
and ancestry interpretations from the decongulated data.
To do so, we developed tables in order to ingest our genotype data to assign an
ancestry.
Our first goal was to recreate the Bioinformatics tools provided in the HID SNP
genotype route.
At the same time, we sought to further delineate some of the larger geographic
regions such
as the America regions in order to give information which was a little more
granular.
To do this, we collected SNP frequency data for different geographic regions
from the
alpha database, the HGDP database, the page database and from Alfred.
The frequency tables from these databases were then used in conjunction with
the Hardy-Wine
group formulas to assess likelihood of a given SNP occurring in a person
descended from a
particular region.
Snips were then combined using the product rule and the geographic regions were
ranked
as displayed on the screen now.
Overall, the results of these analyses are then presented in the following form
with
the results using each database presented individually.
Dependent on the results, some databases are excluded.
For instance, assignment of a person under the European umbrella will lead to
the exclusion
of results from the page database as the page database does not have to be
represented.
If these data are presented most of the time, people of European descent will
be presented
as being of human descent.
Databases are then combined into an overall summary panel which includes 11
geographic
regions again to increase the granularity over their original genotype results.
In particular, the original America group has been expanded into Latin America
East which
includes places like the Caribbean, Latin America West which includes places
like Mexico, North
America as well as South America.
Finally, results are presented in a radar plot to better help the reader
visualize the output
for analysis.
In the example presented on screen, the major contributor is a Latin America
West descent
whereas the minor contributor is a European descent.
To assess how the ancestry calling performed from our tool, we compared the
results using
the former HID SNP genotype tool to our developed tool and found no significant
differences
when looking at single source specimens and those data are not shown.
We then assessed how to mix your deconvolution compared with assignment using
single source
specimen and self-reporting and found that the major was conchordant
approximately 91%
of the time with the minor contributor being conchordant approximately 50% of
the time.
Because we plan to use this tool to help with investigative leads, we looked at
whether
our misses were close geographically as people's of neighboring geographic
regions may share
physical characteristics.
So for example, if we incorrectly called somebody as being from Latin America
West when they
should have in fact been North American or vice versa.
Because we planned when loosening our astringency, accuracy increased 96% per
major and 82% for
the minor and also keep in mind this includes all data examined at ratios from
3.1 out to
10.
And especially at that 10.1 ratio, our data did not perform.
Our deconvolution tool did not perform very well.
We next wanted to look more specifically at ratios where the genotype deconv
olution tool
performed best.
Our accuracy rose substantially at the 6.1 level where we saw 100% concordance
in the major
ancestry and 83% in the minor.
Allowing again for our less stringent interpretation, accuracy was up to 100%
at the 91 and 61 levels
for both major and minor contributors.
In keeping with assessing for phenotype in addition to ancestry, we also
adopted informatics
tools to assign eye and hair color for both our major and minor contributor and
the results
into output performed similarly as our ancestry tool.
So overall, the tool we have put together to deconvolute mixtures of two
individuals
to provide ancestry and phenotype data for both a major and minor looks like
the image
on the screen.
As we've worked with this tool during its development, we have identified a
number of
rules and caveats for interpretation, but overall, we think that this will be
useful
for our investigative partners.
Specific rules, for example, include identifying a major contributor of African
descent, which
may skew results for a minor contributor.
In summary, next generation sequencing allows for the assessment of many
individual genomic
locations and assessment of SNP status of these locations and can be used to
assign ancestry
and phenotype information.
Using variant allele frequency of a reference allele allows us to predict which
genotype
scenario is occurring between a major and minor contributor and applying this
over many
SNPs can help us to assign corresponding interpretive data for both ancestry
and phenotype.
Lastly, we believe that this will be a valuable tool for our investigative
partners.
And with that, I'd like to thank you for your time and attention.
I would be happy to answer any questions you may have.
A big thank you to Martha and Jessie for those incredibly interesting
presentations.
If you have a question for either of them or even both of them, now would be
the time
to submit it through the Q&A panel that you see on your screen.
And while we wait for everyone to submit their questions, why don't we do a few
polling questions
really quickly.
Our first polling question is, does your agency utilize Y chromosome analysis?
Yes, YSTRs for sexual assault, YSTRs for homicide cases, YSTRs for mixture
analysis,
YSTRs for kinship familial analysis, Y screen for sample screening, or no?
Go ahead and get your answer in there.
The largest percentage is no, but then followed by yes, YSTRs for sexual
assault cases.
So that's really interesting.
All right, let's go to our next question.
Does your agency utilize NGS next-generation sequencing analysis?
Yes, for MTDNA analysis.
Yes, for SMP analysis.
Yes, for STR analysis.
No, but we do plan to do so in the future or just no.
Let's see what everyone said.
That's awesome.
It's no, but you do plan to do so in the future, which is really interesting
because
as we know, NGS is really taking off.
So that's interesting.
All right, our last one before we get into Q&A, would you like to receive more
information
on today's webinar topics?
Whether that be about YSTRs analysis, MTDNA analysis, using NGS or SMP analysis
using
NGS.
All right, thank you everybody so much for answering those questions.
Now we will get to the fun part, the attendee questions.
So Jesse, Martha, you guys still with us?
I think so.
Awesome.
All right, Jesse then, we will give you your first question.
Jesse, can you elaborate how you subdivided the Americas and how that made for
better
predictions in your research?
Yes, so when we were going through initially kind of setting up our frequency
tables and
whatnot, the first thing that we were trying to do was more or less recreate
what the
Bioinformatics platform we had already been using for the process of going
through that
whole de-contribution process.
Since we had the opportunity to kind of expand and look at some of these other
databases,
many databases especially like Alfred, for example, have a lot of granularity
when it
comes to the different ethnicities and peoples that they've done sequencing for
And so to kind of address some of the shortcomings, especially those folks
which were of Latin
American descent, we kind of took advantage of some of that granularity to help
kind of
generate and build out some of these more subdivided groups especially over in
the Americas.
So hopefully that's helpful.
Absolutely.
Now, how would you adapt this to other panels?
And so for other panels, a lot of it is just determining which SNPs are
interrogated and
looked at there.
Dependent upon which SNPs are there, what you'd have to go to is identify the
frequency
tables for those particular SNPs.
So in order to adapt the current tool that we have right now to say a new up-to
-date
or a different platform or something, there would be a lot of kind of heavy
lifting towards
addressing that and kind of getting to that.
But it's definitely doable, it would just take some time.
Okay.
Now another one if you, Jesse, Marta, before we move on to a few questions for
you.
So Jesse, how do you focus on the Y SNPs for a certain contributor, whether
that's major
or minor?
Do you have to perform STR analysis first and then focus on the certain alleles
from that
contributor?
So the way that we're envisioning that this workflow will work is we're going
to do traditional
STR analysis to establish a proportion of major to minor.
We're not currently using YSTRs or Y SNPs to kind of delineate between major
versus minor.
So it would just be more looking at those autosomal STR contributions and what
not to
establish that we have a two-person mixture and then from there what is the
ratio between
the major and the minor.
So hopefully that's helpful.
Absolutely.
Thank you.
I'm sure we'll come back to you.
But Marta, we want to ask you a few questions that we have from the audience.
The first one is, can you tell us in your study what were the biggest technical
issues
that you faced?
The biggest technical issue was actually with the bone material itself.
In a project we collected almost 800 bone samples and the ones collected in
probably
Boudoir.
One of the most liquid databases we had to work with, we implemented many
different DNA
extractions methods.
So we started with PrEPFiler BTA, then we continued with the organic funnel
chloroform
and still we were not happy with the results as you saw, the first results from
Sanga
were definitely incomplete.
So the next step was introducing the in-house methods from in-group for the DNA
extraction
that was lately published and luckily we had enough extract for each individual
to continue
with their next generation sequencing.
And we got home microgenomes and the Y chromosome data, but the SDR, the Y SDR
sequencing had
to be repeated many times and the mitochondrial and DNA was also extremely
degradation.
So it was a big technical challenge to even get the satisfying results.
Okay.
Can you comment and talk to us about how adding whole genome MIDO analysis by N
GS actually
aided in the identification that you were able to make?
Yes.
So without my Tohonjua, the homo-tohonjua genome, we would not get phylogenetic
resolution
that we wanted.
So the slides of Sanga showed that for three individuals, we got the prediction
of K-HAPLO
group, even though for two of them, the control regions of HV1 and HV2 allowed
only to predict
the HAPLO group SK, which was definitely not informative enough.
So thanks to the whole genome sequencing, we got a higher phylogenetic
resolution, first
of all, and second of all, when we were comparing the homo-tohonjua genome
sequences with the
reference data from Polish and Ashkenazi, Michael Genome, we definitely got
also a better
understanding of where the results fall in phylogenetically.
So we were able to cluster the micro-hondyl genome, which was not able only
based on the
control-avision sequence.
Okay, gotcha.
Now, what was the biggest challenge of publishing your results?
I think that the biggest challenge was how complex our study is, and even
though we, like
the major focus was on DNA, obviously, and I'm a forensic scientist, and all of
the results
come from a forensic scientist, we could not ignore how complex the history
behind our
findings.
So when writing the paper, I had to go very deep into the history of everything
, and we
had to work with archeologists and anthropologists, because we could not
present the results without
exploring the history of the findings.
We also wanted people to understand how important the results are and how big
the discovery
is, and this would not be possible without putting all of the historical
background to
it.
And this is something we just have a very typical forensics.
So introducing so much of history and archeology into our study.
Gotcha.
Okay, and that sounds like a fascinating article.
Can you remind the audience again what's the name of it is, and it's published
in genome
biology, is that correct?
Yes, it's published in genome biology in August this year.
It's genetic and phylogenetic evidence for Jewish Holocaust victims, a found in
Saudi
border, that can, so I hope that the paper includes everything I said and
beyond.
So I think it's great if people can really learn about the story, because I
think that
this is a part of history that many really don't know much about, and I think
that our
paper just tries to shed some light on it.
Absolutely.
And I have one more question for you before we throw it back to Jessie to round
this up.
You know, what's next for you?
Are there current efforts to identify the remains?
Are you yourself planning any more humanitarian projects concerning the remains
of the Second
World War?
This would be of course an outstanding achievement to be able to identify
people.
And this is something that of course we hope for you.
We don't know yet exactly if this would be possible.
Right now when we have the phylogenetic data and we know that the people of
Jewish origin,
we have to consider who the victims could be.
And we already know that there were 200,000 people killed in Saudi border, but
prisoners
who actually destroyed the camp were the prisoners from the other camps, which
was Treblin, our
800,000 people were killed.
So we are looking potentially at 1 million victims.
So I know that there are amazing tools in the forensics right now, in the
foreign industry,
geology, we have all of the extended analysis.
So I hope that one day this study moves forward because giving names back to
those victims
would be just something absolutely great.
And since this project concerns my previous workplace and my previous team, I
right now
moves to Munich and I work at the Institute of the Kalmani-sen in Germany.
And even though I focused more on implementing NGS into routine investigation,
I never stopped
thinking and talking about bonds.
So everyone who knows me also hear about this from me almost daily because this
is just
my greatest session in forensics.
And for a while already I'm working on another humanitarian project that was
starting my
institute with other colleagues joining in that was introduced in the
scientific identification
of the foreign German soldiers from the second forward.
Oh wow, that's interesting.
We'll definitely have to look out for that.
Thanks, Marta.
Thanks.
Alright, Jesse, we have another one for you.
When do you anticipate that you can offer this to our investigative partners?
So more or less our validation is completed.
So we just have to write up a few more documents.
Obviously, so the kind of a null part of this is being able to take this NGS
data, parse
out a major contributor and a minor contributor and then obviously offer some
sort of, you
know, type and ancestry analysis on those.
So we'd like to have, you know, maybe a publication or something like that here
in the coming
months or whatnot.
But more or less, we're pretty close to being able to offer this to
investigative agencies.
Wow, that's fantastic.
Alright, and one more question for actually both of you before we call it a day
on this
webinar.
Jesse, we'll start with you.
Do you focus more on autosomal SMPs or YSMPs?
So for the system that we're using now, I mean, assay that we're using now, all
of
these ancestry SNPs and phenotypes SNPs are located on the autosomal
chromosomes.
You know, certainly if you're getting into more of the genealogy types of
studies, you're
going to broaden the number of SNPs that you're looking at.
And then when you're kind of getting into that, you can start looking at SNPs,
which
are located on X and Y respectively.
So at the current moment in time, everything that we're doing right now is more
focused
on the autosomal stuff, but certainly YSMPs are an area of interest.
It's to be kind of expand our testing more.
Marta, what about you?
Definitely on Y, which was always my biggest interest in the autosomal one.
And in everyday work, I also work with phenotypes, I think, as Jesse.
So Y chromosomes are of interest for me, also from the phylogenetic perspective
So definitely Y chromosomes.
Gotcha.
Okay.
Thank you guys.
All right.
Audience, that about wraps up all the time we have for today.
I'd like to thank our fantastic presenters, as well as thermofuscious
scientific, who's
a sponsor for not only today's webinar, but the entire 2021 Future Trends and
DNA Technology
series.
In 24 hours or less, this webinar will be available on demand if you'd like to
watch
it again or share it with friends and colleagues.
In fact, the entire 2021 series is available on demand on the forensic website,
www.ferensicmag.com.
Additionally, later today, you will receive an email with information on how to
obtain
CE credit documentation for your participation in this webinar.
And while this unfortunately concludes the 2021 webinar series, forensics and
thermofuscious
scientific are thrilled to continue this informative series next year.
So be on the lookout for our notice in early 2022.
Thank you so much, everyone, and have a great day.