PNAS PLUS
Second generation noninvasive fetal genome analysis
reveals de novo mutations, single-base parental
inheritance, and preferred DNA ends
K. C. Allen Chana,b,1, Peiyong Jianga,b,1, Kun Suna,b,1, Yvonne K. Y. Chengc, Yu K. Tonga,b, Suk Hang Chenga,b,
Ada I. C. Wonga,b, Irena Hudecovaa,b, Tak Y. Leungc, Rossa W. K. Chiua,b,2, and Yuk Ming Dennis Loa,b,2
a
Li Ka Shing Institute of Health Sciences, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong; bDepartment of Chemical Pathology,
The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong; and cDepartment of Obstetrics and Gynaecology, The
Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong
Contributed by Yuk Ming Dennis Lo, October 4, 2016 (sent for review August 24, 2016; reviewed by Mary E. Norton and Erik A. Sistermans)
Plasma DNA obtained from a pregnant woman was sequenced to a
depth of 270× haploid genome coverage. Comparing the maternal
plasma DNA sequencing data with the parental genomic DNA data
and using a series of bioinformatics filters, fetal de novo mutations
were detected at a sensitivity of 85% and a positive predictive value
of 74%. These results represent a 169-fold improvement in the pos-
itive predictive value over previous attempts. Improvements in the
interpretation of the sequence information of every base position
in the genome allowed us to interrogate the maternal inheritance
of the fetus for 618,271 of 656,676 (94.2%) heterozygous SNPs within
the maternal genome. The fetal genotype at each of these sites was
deduced individually, unlike previously, where the inheritance was
determined for a collection of sites within a haplotype. These results
represent a 90-fold enhancement in the resolution in determining the
fetus’s maternal inheritance. Selected genomic locations were more
likely to be found at the ends of plasma DNA molecules. We found
that a subset of such preferred ends exhibited selectivity for fetal- or
maternal-derived DNA in maternal plasma. The ratio of the number of
maternal plasma DNA molecules with fetal preferred ends to those
with maternal preferred ends showed a correlation with the fetal DNA
fraction. Finally, this second generation approach for noninvasive fetal
whole-genome analysis was validated in a pregnancy diagnosed with
cardiofaciocutaneous syndrome with maternal plasma DNA sequenced
to 195× coverage. The causative de novo BRAF mutation was success-
fully detected through the maternal plasma DNA analysis.
|
noninvasive prenatal testing massively parallel sequencing | use of maternal haplotypes, hence greatly increasing the reso-
DNA fragmentation patterns
T he discovery of cell-free fetal DNA in maternal plasma has
enabled the development of noninvasive prenatal testing
(NIPT) (1). Over the last few years, NIPT has been implemented
globally for the noninvasive prenatal investigation of fetal chro-
mosomal aneuploidies (2–5). With higher depth of sequencing and
improved bioinformatics analyses, NIPT has now been extended to
the detection of a variety of subchromosomal aberrations (6, 7).
Further expanding the applications of NIPT, we showed in 2010
that it was possible to deduce the fetal genome by deep sequencing
of maternal plasma (8). Work by Fan et al. (9) and Kitzman et al.
(10) confirmed these results. In these previous efforts, the depths
of maternal plasma DNA sequencing ranged from 52.7× to 78×
haploid human genome coverages (8–10).
There are a number of limitations in these previous studies. For
example, Kitzman et al. (10) explored the possibility of detecting
fetal de novo mutations on a genome-wide level from the maternal
plasma DNA sequencing data. In one variation of bioinformatics
analysis, they found 2.5 × 107 candidate fetal de novo mutation
sites in the plasma DNA sequencing data. Only 39 of these were
true fetal de novo mutations. Because the studied fetus had a total
of 44 de novo mutations, the positive predictive value (PPV)
was 0.000156%, and the sensitivity was 88.6%. With additional
Downloaded by guest on July 8, 2021
refinement in bioinformatics analysis, Kitzman et al. (10) improved
www.pnas.org/cgi/doi/10.1073/pnas.1615800113
SEE COMMENTARY
the PPV to 0.438%, although the sensitivity was reduced to 38.6%.
These data, thus, indicate the enormous challenge of detecting fetal
de novo mutations on a genome-wide scale using NIPT. In par-
ticular, dramatic improvement in the PPV would be needed for
such an approach to be clinically practical.
A second area that needs improvement concerns the detection
of sequences that the fetus has inherited from its mother. Previous
efforts in elucidating the maternal inheritance of the fetus on a
genome-wide scale have generally used a haplotype-based strat-
egy, which has been referred to as the relative haplotype dosage
(RHDO) approach (8–10). Hence, for a pregnant woman who has
two haplotypes in a particular chromosomal region, she would
pass one of these onto her fetus. Because her plasma contains a
Significance
We explored the limit of noninvasive prenatal testing by per-
forming genome-wide sequencing of maternal plasma DNA at
195× and 270× haploid genome coverages. Combined with the
use of a series of bioinformatics filters, fetal de novo mutations
could be detected with a positive predictive value that was two
orders of magnitude higher than previously reported. A de novo
MEDICAL SCIENCES
BRAF mutation was noninvasively detected in a case with car-
diofaciocutaneous syndrome. The maternal inheritance of the
fetus could be ascertained on a genome-wide level without the
lution of such analysis. Finally, we showed that certain genomic
locations were overrepresented at the ends of plasma DNA
fragments with fetal or maternal selectivity.
Author contributions: K.C.A.C., R.W.K.C., and Y.M.D.L. designed research; K.C.A.C., P.J., K.S.,
Y.K.Y.C., Y.K.T., S.H.C., A.I.C.W., I.H., and T.Y.L. performed research; K.C.A.C., P.J., K.S.,
R.W.K.C., and Y.M.D.L. analyzed data; Y.K.Y.C. and T.Y.L. recruited subjects and ana-
lyzed clinical data; and K.C.A.C., P.J., R.W.K.C., and Y.M.D.L. wrote the paper.
Reviewers: M.E.N., University of California, San Francisco; and E.A.S., VU University
Medical Center.
Conflict of interest statement: R.W.K.C. and Y.M.D.L. received research support from
Sequenom, Inc. R.W.K.C. and Y.M.D.L. were consultants to Sequenom, Inc. K.C.A.C., R.W.K.C.,
and Y.M.D.L. hold equities in Sequenom, Inc. K.C.A.C., R.W.K.C., and Y.M.D.L. are founders
of Xcelom and Cirina. K.C.A.C., P.J., and R.W.K.C. are consultants to Xcelom. P.J. is a
consultant to Cirina. K.C.A.C., P.J., R.W.K.C., and Y.M.D.L. have filed patent applications
(PCT/CN2016/073753 and PCT/CN2016/091531) based on the data generated from this
work, which have been licensed to Cirina.
Freely available online through the PNAS open access option.
Data deposition: The sequence data for the subjects studied in this work who had con-
sented to data archiving have been deposited in the European Genome-Phenome Archive
(EGA), https://www.ebi.ac.uk/ega/, hosted by the European Bioinformatics Institute (EBI;
accession no. EGAS00001001882).
See Commentary on page 14173.
1
K.C.A.C., P.J., and K.S. contributed equally to this work.
2
To whom correspondence may be addressed. Email: rossachiu@cuhk.edu.hk or loym@
cuhk.edu.hk.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1615800113/-/DCSupplemental.
PNAS | Published online October 31, 2016 | E8159–E8168
 mixture of her own DNA and that from the fetus, there would be a
slight overrepresentation of the haplotype shared by both the
pregnant mother and her fetus. This haplotype-based approach
has imposed two limitations to NIPT. First, it requires the eluci-
dation of the maternal haplotype using a direct haplotyping ap-
proach (11–13), via pedigree analysis (8, 14), or via founder
haplotype analysis in selected populations (15). Second, this
haplotype-based approach has limited the resolution in which the
maternal inheritance of the fetus can be determined. In this
regard, the mean length of the haplotype blocks that had been
used in previous efforts ranged from 300 kb to over 1 Mb (8–10).
Recently, there is a lot of interest in the fragmentation patterns
of plasma DNA (16–18). Studies showed that plasma DNA frag-
mentation sites are located in clusters across the genome that bore
relationships with positions of nucleosome arrays and open chro-
matin domains. Based on these data, researchers have concluded
that the fragmentation process of plasma DNA is nonrandom. One
interpretation of such observations is that plasma DNA fragments
are cleaved at the accessible parts of the genome. We have gone
further in this study to explore whether the actual ending sites of
plasma DNA were nonrandom down to a single-base level. In
other words, are plasma DNA molecules repeatedly cut at a se-
lected set of genome coordinates? Previous studies could not ad-
dress this question, because the sequencing depths of individual
samples were not high enough, and PCR duplicates were discarded
before additional data analysis. We hypothesized that some of such
“duplicates” were not created by PCR. Instead, they were plasma
DNA molecules that had originated from different cells and were
cleaved at the same genome coordinates. To investigate this hy-
pothesis, we processed some of the studied plasma samples using a
PCR-free library preparation protocol.
In this study, we aimed to develop a second generation approach
to noninvasively decipher the fetal genome from maternal plasma
using sequencing depth that had hitherto not been achieved in
previous studies (Fig. 1). We hypothesized that the use of ultra-
deep genome-wide plasma DNA sequencing and multiparametric
Noninvasive fetal whole genome analysis
Sequencing Depth
De Novo Mutaon
Sensivity
PPV
Maternal Inheritance
Principle
Resoluon
1. Nucleosomal paern
2. Size difference between maternal-
and fetal-derived fragments
Frequency (%)
Plasma DNA
Fragmentaon
Paern
Fig. 1. Summary of the key differences between the first and second generation approaches for noninvasive fetal whole-genome analysis. aThe per-
Downloaded by guest on July 8, 2021
formance characteristics quoted for first generation fetal whole-genome analysis were from refs. 8–10. *RHDO represents RHDO analysis, and GRAD
represents GRAD analysis.
E8160 | www.pnas.org/cgi/doi/10.1073/pnas.1615800113
bioinformatics analysis that takes into account the sequence,
concentration, and size of the circulating DNA fragments would
allow one to significantly improve the detection of fetal de novo
mutations, interrogate fetal inheritance at single-base resolution,
and study the end characteristics of plasma DNA on a molecule by
molecule manner. We further illustrated how such detailed level of
fetal genomic information would be of clinical utility.
Results
Clinical Samples and DNA Sequencing. Pregnant women were
recruited from the Department of Obstetrics and Gynecology,
Prince of Wales Hospital with informed consent. Hypotheses were
first explored and methodologies were developed based on the
analysis of a plasma sample from a pregnant woman at 38 wk of
gestation. These “second generation” fetal genome methodologies
were then applied to the analysis of plasma collected from a sec-
ond trimester pregnancy affected by cardiofaciocutaneous syn-
drome. Plasma samples from 26 first trimester pregnant women
were analyzed to further investigate the phenomenon of preferred
ending sites among plasma DNA molecules.
Key Maternal Plasma DNA Sequencing Metrics. Maternal plasma
DNA of the third trimester case was sequenced using the Illumina
TruSeq PCR-free library preparation protocol to a depth of 270×
coverage of a haploid human genome. The maternal blood cells,
paternal blood cells, and umbilical cord blood cells were se-
quenced to 40×, 45×, and 50× haploid human genome coverages,
respectively, using the same sequencing protocol. We analyzed
SNPs where both parents were homozygous but for different al-
leles and identified 239,816 of such SNPs. For these SNPs, the
fetus would be an obligatory heterozygote for the paternal and
maternal alleles. Among these SNPs, the paternal alleles were
detectable in the maternal plasma of all of 239,816 SNPs. The
fetal DNA fraction (F) in maternal plasma was deduced as 31.3%
from the aggregated numbers of paternally inherited fetal alleles
1st generaona 2nd generaon
53× − 78× 195× − 270×
38.6% 81% − 85%
0.438% 62% − 74%
Haplotype-based by RHDO* SNP-based by GRAD*
300k − over 1M Single nucleode
Plasma DNA preferred end sites
Genomic coordinates
Fetal-
derived
DNA
166 fragments
Maternal-
143 derived
DNA
fragments
Maternal preferred
end site
Size (bp) Fetal preferred end site
Chan et al.

(p) and the maternal alleles (m) at these SNPs using the formula
F = 2p/(p + m).
We next measured the difference in the sizes of fetally and
maternally derived DNA in the maternal plasma at individual
SNPs. We compared the difference in size distribution of fragments
carrying the fetal-specific allele and the allele shared by the fetus
and the mother at a particular SNP (Fig. S1). Such size information
would be useful for determining the likelihood of any variant allele
detected in maternal plasma being a fetal de novo mutation. The
median difference between fragments carrying the fetal-specific
alleles and the shared alleles was 22 bp (Fig. S2A). At 90% of these
SNP loci, the difference was larger than 10 bp (Fig. S2A).
Detection of Fetal de Novo Mutations from Maternal Plasma. Fifty-six
single-nucleotide variants were found to be present in the umbilical
cord blood cells and absent in the parents. These 56 single-
nucleotide variants were thus deemed de novo mutations possessed
by the fetus. Deep genome-wide sequencing is needed for the
screening of fetal de novo mutations from maternal plasma. At this
level of sequencing, a vast number of sequencing errors would be
generated and hinder the sensitive and specific detection of true
fetal de novo mutations. To tackle the challenge, we designed a
bioinformatics protocol that used a series of steps to filter out the
false-positive mutations (Fig. 2). The bioinformatics protocol made
use of steps that identified high-quality base sequences in combi-
nation with steps that took into account the known biological
characteristics of cell-free fetal DNA.
First, we identified genomic locations where the maternal
plasma DNA sequencing showed at least one sequence variant,
whereas the parental DNA analysis showed that both parents
shared the same sequence. Second, we applied a dynamic cutoff
algorithm to determine if the variant detected in the maternal
Dynamic cutoff
Variant frequency higher than sequencing error rate
60,111
Detection rate: 54/56 (96%); PPV: 0.089%
Realignment
Realign all reads covering putative mutation sites
148
Detection rate: 52/56(93%); PPV: 35%
Fractional concentration
Variant frequency compatible with fetal DNA fraction
85
Detection rate: 51/56 (91%); PPV: 60%
Plasma DNA fragment size
Mean size of fragments carrying the variant 10 bp shorter than
the mean size of fragments carrying the parental alleles
65
Detection rate: 48/56 (85%); PPV: 74%
Fig. 2. Detection of the fetal de novo mutations through the analysis of the
sequencing data of maternal plasma DNA. The number in bold in each step
Downloaded by guest on July 8, 2021
represents the number of candidate de novo mutations identified after the
particular process.
Chan et al.
PNAS PLUS
plasma was more likely to have originated from the fetus or se-
quencing errors. In this algorithm, a variant would need to be
observed in a threshold number of sequenced reads so as to qualify
as a candidate de novo mutation (Table S1). This threshold was
determined from a binomial probability function based on the total
SEE COMMENTARY
number of times that a particular nucleotide was actually se-
quenced and the sequencing error rate, so that the theoretical
probability of a putative de novo mutation being a sequencing
error would be less than 1 in 3 × 109. Using this dynamic cutoff
algorithm, 60,111 candidate de novo mutations were identified,
and these candidate de novo mutations included 54 of 56 (i.e.,
96%) de novo mutations present in the cord blood (Fig. 2).
Then, we realigned the sequenced reads covering the 60,111
candidate de novo mutation sites to the reference human genome
using the BOWTIE2 software (19). The initial alignment software,
Short Oligonucleotide Alignment Program 2 (SOAP2) (20), and
the realignment software, BOWTIE2, were based on two different
matching algorithms, namely the Burrows–Wheeler transform al-
gorithm and Smith–Waterman-like algorithm, respectively (19, 20).
A variant would be removed from being considered as a candidate
de novo mutation if the two alignment algorithms do not map the
variant to the same genomic location. This realignment procedure
significantly reduced the number of false-positive results caused by
alignment errors. After this realignment process, the number of
candidate de novo mutations was reduced to 148 and contained
52 (93%) of the de novo mutations detected in the cord blood
(Fig. 2). This result is equivalent to a PPV of 35%.
To further reduce the number of false positives, we applied a
filter based on the fractional concentration of the variants. Because
the fetal DNA fraction in the sample was 31.3%, 95% of the var-
iants would be expected to be present at a level of ≥10% of the
sequenced reads covering a particular nucleotide based on the
Poisson distribution. Hence, we filtered out the candidate muta-
tions that were present at a fraction of <10%. Eighty-five candidate
de novo mutations were present at ≥10% of the sequenced reads
covering the nucleotide and included 51 (91%) of the mutations
detected in the cord blood (Fig. 2). This result represents a PPV
of 60%.
Next, we included a filter based on the size of the plasma DNA
MEDICAL SCIENCES
molecules carrying the candidate de novo mutations. As illustrated
in Fig. S2A, the difference between the mean fragment size for
fragments carrying the fetal-specific alleles and the shared alleles
was larger than 10 bp for 90% of loci at which the fetus was
heterozygous and the mother was homozygous. Thus, we applied a
size filter to the candidate de novo mutations, requiring the mean
size of the plasma DNA fragments carrying the mutants to be at
least 10 bp shorter than that of the fragments carrying the parental
alleles. Sixty-five candidates passed this size-filtering criterion that
included 48 (85%) of the mutations detected in the cord blood
(Fig. 2). This result represents a PPV of 74%.
Paternal Inheritance Analysis. For paternal inheritance analysis, we
focused on SNPs that were heterozygous in the father and homo-
zygous in the mother. In principle, the presence or absence of the
paternal-specific allele in the maternal plasma would indicate the
inheritance of paternal-specific allele or the shared parental allele
by the fetus, respectively. To enhance the accuracy of deducing the
paternal inheritance, we used a binomial mixture model, which
took into account the fetal DNA concentration, the sequencing
error rate, and the number of times that the paternal-specific allele
was observed in the maternal plasma (details are in Materials and
Methods) (21). For 667,586 SNPs for which the mother was ho-
mozygous and the father was heterozygous, the overall accuracy
was 99.99% for the deduction of the paternal inheritance.
Genome-Wide Relative Allele Dosage Analysis. Previous efforts in
prenatally elucidating the maternal inheritance of the fetus on a
genome-wide level had all been based on the RHDO approach
PNAS | Published online October 31, 2016 | E8161
 (8–10). Enabled by the high-quality genome-wide base information,
we showed that the maternal inheritance of these heterozygous
SNPs could be resolved without the use of the haplotype in-
formation of the mother. The dosage of the two maternal alleles in
maternal plasma would be compared to determine, statistically,
whether the two alleles are present at the same or different con-
centrations. For SNP loci at which one allele was present at signif-
icantly higher concentration and at which the two alleles were
present at statistically nondifferentiable concentrations, the fetus
would be deduced to be homozygous and heterozygous, respectively.
We denote this method as genome-wide relative allelic dosage
(GRAD) analysis. This method is a genome-wide and sequencing-
based version of our previously described relative mutation dosage
method (22). There were a total of 656,676 heterozygous SNPs in
the mother’s genome. Using GRAD analysis, the maternal in-
heritance of the fetus was resolved with statistically significant results
in 618,271 SNPs (i.e., 94.2%). The allelic counts were insufficient to
give statistically significant results in 38,405 SNPs (i.e., 5.8%). The
fetus was deduced to be homozygous at 335,988 (54%) SNPs and
heterozygous at 282,283 (46%) SNPs. Among 618,271 SNPs with
statistically significant results, the deduction of maternal inheritance
was correct for 610,084 (98.7%) SNPs compared with the genotypes
from the analysis of the cord blood cells.
The resolution of GRAD analysis would be affected by the se-
quencing depth and the fetal DNA fraction in the maternal plasma
sample. We performed computer simulation analyses to determine
the number of SNPs required for achieving one classification of
maternal inheritance with an overall accuracy of over 95% (Fig.
S3). With a sequencing depth of 250×, we would be able to achieve
close to single-nucleotide resolution for the deduction of maternal
inheritance when the fetal DNA fraction was above 20%. When
the fetal DNA fraction was 10% and a sequencing depth was at
300×, maternal inheritance could be deduced in 1 of every 5 SNPs
with an accuracy of over 95%, but the resolution of maternal in-
heritance would drop to 1 classification per 30 SNPs with a se-
quencing depth of 150×.
Preferred End Sequences in Maternal Plasma DNA. Facilitated by the
high-sequencing depth of a non-PCR–amplified library of the
maternal plasma DNA sample, we investigated if certain base
positions in the genome would be preferentially represented at the
ends of plasma DNA fragments. In this regard, we showed that
plasma DNA fragments carrying the most prevalent 0.5% of
plasma DNA ends represented 3.5% of all DNA fragments. Across
the genome, 25% of the fragments had at least one identical
fragment that shared the same ending sites at both ends. If the
cleavage or breakage of plasma DNA was completely random, only
1.45% of the DNA molecules would be expected to have at least
one counterpart sharing common ends (details are in Materials and
Methods). Had the sequencing been performed using a PCR-
amplified library, then 14% of the fragments would have been wrongly
assumed to be PCR duplicates and would have been filtered off.
We further investigated if there would be differences in the
preferred ending sites for maternal and fetal DNA in the same cell-
free DNA sample. To show this phenomenon, informative SNP
loci where the mother was homozygous (genotype denoted as AA)
and the fetus was heterozygous (genotype denoted as AB) were
identified. In this illustrative example, the B allele would be fetal-
specific, and the A allele would be shared by the mother and the
fetus. The fetal-specific and shared reads covering one such in-
formative SNP are shown in Fig. 3. For comparison, the se-
quencing results of a DNA sample obtained from blood cells and
artificially fragmented using sonication are also shown. Clusters of
preferred ending positions could be observed among the plasma
DNA data. For the plot of the probability of a genomic location
being an end of DNA fragments (also referred to as the fragment
end probability), three peaks were observed for each of two groups
Downloaded by guest on July 8, 2021
of fragments carrying the fetal-specific allele and the allele shared
E8162 | www.pnas.org/cgi/doi/10.1073/pnas.1615800113
by the mother. These peaks represent the hotspots for the end
positions of fetal- and maternal-derived DNA in maternal plasma,
respectively. In contrast, the fragmentation pattern of the soni-
cated DNA seems to be random without such clusters, and the
fragment end probability is similar across the region (Fig. 3).
We further studied the coordinates that had an increased
probability of being an ending position for plasma DNA fragments.
We focused our search based on fragments covering the in-
formative SNPs, so that the fragments carrying fetal-specific alleles
and alleles shared by the mother and the fetus could be evaluated
separately. We determined if certain locations within the human
genome had a significantly increased probability of being an ending
position of plasma DNA fragments using a Poisson probability
function (Materials and Methods). A P value of <0.01 had been
chosen to indicate statistical significance. Statistically significant
ending positions were determined for DNA fragments carrying the
shared allele and the fetal-specific allele independently (Fig. 4A).
We identified a total of 8,242 (Set A) and 23,857 (Set B) nu-
cleotide positions with a significantly increased chance of being an
end for plasma DNA fragments carrying fetal-specific alleles and
shared alleles, respectively, covering 10,233 SNPs; 8,909 of the
nucleotide positions were observed to be overrepresented among
the plasma DNA molecules with the fetal-specific allele (Set A) as
well as molecules with the shared allele (Set B). We called this
overlapping set of ends Set C (Fig. 5A). There were 48,707, 53,901,
and 65,205 plasma DNA fragments carrying fetal-specific alleles
ending on Set A, Set B, and Set C positions, respectively. In other
words, multiple fetal-specific DNA molecules showed identical
ending positions. A median of 6 (range = 6–41) plasma DNA
fragments carrying fetal-specific alleles terminated at a Set A
ending position. Based on a sequencing depth of 270×, fetal DNA
fraction of 31.3%, and the size distribution of fetal DNA frag-
ments, it was expected that only 0.29 fragment would end at each
site if the fragmentation of plasma DNA had been random. Thus,
the probability of ending at these sites was 20 times higher than
expected. There were 54,541, 376,343, and 182,791 plasma DNA
fragments carrying shared alleles ending at Set A, Set B, and Set C
positions, respectively. The genomic coordinates of Set A, Set B,
and Set C positions are shown in Dataset S1.
Using the same principle, we analyzed the ending positions for
maternal-specific plasma DNA fragments. SNPs that were het-
erozygous in the mother (genotype AB) and homozygous in the
fetus (genotype AA) were noted, and plasma DNA molecules that
carried any one of the maternal-specific alleles (B allele) were
deemed to be maternally derived. Among the maternal-specific
plasma DNA molecules, 7,527 (Set X) and 18,829 (Set Y) nucle-
otide positions showed increased occurrence as an ending position
for plasma DNA fragments carrying maternal-specific alleles and
shared alleles, respectively, covering 9,489 SNPs (Fig. 4B); 10,534
positions were observed to be overrepresented among the plasma
DNA molecules with the maternal-specific allele (Set X) as well as
molecules with the shared allele (Set Y). We called this over-
lapping set of ends Set Z (Fig. 5B). There were 69,136, 82,413, and
121,607 plasma DNA fragments carrying maternal-specific alleles
ending at Set X, Set Y, and Set Z positions, respectively. A median
of 10 (range = 9–54) plasma DNA fragments carrying maternal-
specific alleles terminated at a Set X base position. If the frag-
mentation of plasma DNA had been random, it was expected that
only 0.56 fragment would terminate at each site. Thus, the prob-
ability for ending at these sites was 18 times higher than expected.
There were 46,554, 245,037, and 181,709 plasma DNA fragments
carrying shared alleles ending on Set X, Set Y, and Set Z positions,
respectively. The genomic coordinates of the Set X, Set Y, and Set
Z positions of this case are shown in Dataset S1.
Using Preferred End Positions to Deduce Fetal DNA Fractions in Maternal
Plasma in First Trimester Pregnancies. After having identified sets of
maternal plasma DNA preferred end positions from one third
Chan et al.
 PNAS PLUS
Maternal plasma DNA Sonicated maternal blood cell DNA
Informative SNP Informative SNP

SEE COMMENTARY
Shared allele
Fetal-specific allele

8
4

Fraction of
fragments ending
8

on the nucleotide

8
position (%)
4

4
-300 -200 -100 0 100 200 300 -300 -200 -100 0 100 200 300
Relative genomic coordinates (bases) Relative genomic coordinates (bases)

Fig. 3. Left shows an illustrative example of the nonrandom fragmentation patterns of plasma DNA carrying a fetal-specific allele and an allele shared by the
mother and the fetus. In Upper, each horizontal line represents one sequenced DNA fragment. The ends of the DNA fragments represent the ending position of
the sequenced read. The fragments are sorted according to the coordinate of the left outermost nucleotide (genomic coordinate with the lower numerical value).
In Lower, the percentage of fragments ending on a particular position is shown. Upper Right shows the span of all of the sequenced DNA fragments aligned to
the same SNP obtained from the sonicated blood cell DNA of the same pregnant woman. Lower Right shows the percentage of fragments ending on a particular
position. The x axes show the relative genomic coordinates. The SNP of interest is located in the middle (marked by a dotted line).

trimester case, we explored if such ends would be detectable in fetus. Sequencing libraries were prepared from these maternal
samples of other pregnancies and whether the relative abundance plasma samples using the KAPA Library Preparation Kits (Kapa
of plasma DNA ending at these sets of nucleotide positions would Biosystems), which included a PCR amplification step to enrich the
reflect the fetal DNA fraction. We sequenced the plasma DNA of adaptor-ligated molecules. The median number of mapped reads
26 first trimester (10–13 wk) pregnancies, each involving a male per sample was 16 million (range = 12–22 million). The proportion

A Informative SNP that mother was homozygous B Informative SNP that mother was heterozygous
and fetus was heterozygous and fetus was homozygous

MEDICAL SCIENCES
Sequencing reads carrying Sequencing reads carrying
8

shared allele
8

Significant ending shared allele

Proportion of plasma DNA ending at

Proportion of plasma DNA ending at

Significant ending
position specific for
Significant ending position specific for
fragments carrying Significant ending
position common for fragments carrying
a genomic coordinate (%)

shared alleles
a genomic coordinate (%)

shared alleles position common for

fragments carrying
fragments carrying
shared and fetal-specific
4

shared and maternal-

alleles
specific alleles
0

0
4

Significant ending Significant ending

position specific for position specific for
fragments carrying fragments carrying
fetal-specific alleles maternal-specific alleles
Sequencing reads carrying Sequencing reads carrying
8

fetal-specific allele maternal-specific allele

18878200 18878400 18878600 18878800

17053400 17053600 17053800 17054000 17054200
Genomic coordinates (chr22) (bases) Genomic coordinates (chr1) (bases)
Fig. 4. Plot of probability of a genomic coordinate being an ending position of maternal plasma DNA fragments across a region with an informative SNP at
which (A) the mother was homozygous and the fetus was heterozygous and (B) the mother was heterozygous and the fetus was homozygous. (A) Results for
nucleotide positions with a significantly increased probability of being an end of plasma DNA fragments carrying a shared allele and a fetal-specific allele are
shown in red and blue, respectively. (B) Results for nucleotide positions with a significantly increased probability of being an end of plasma DNA fragments
Downloaded by guest on July 8, 2021

carrying a shared allele and a maternal-specific allele are shown in red and blue, respectively. The x axes of the graphs show the relative genomic coordinates.
The position of the SNP of interest is marked by a dotted line.

Chan et al. PNAS | Published online October 31, 2016 | E8163

 A Ending positions for plasma DNA fragments To investigate the specificity of these positions, we analyzed the
across a SNP that was homozygous in the mother size distribution of DNA fragments ending on the Set A or Set X
and heterozygous in the fetus positions when the genomic coordinates of these positions were
shifted by one to five nucleotides. To quantify the size difference,
Preferred ending positions Preferred ending positions cumulative frequencies for fragments ending at the two sets of
for fragments carrying for fragments carrying
shared alleles
positions were plotted against size. The difference in the two cu-
fetal-specific alleles
mulative frequency curves (denoted as ΔS) would reflect the
magnitude of the size difference (Figs. 7B and 8B). For the third
trimester sample, the difference in size distribution diminished
8,242 8,909 23,857 when the number of shifted nucleotides was increased and no
longer observable when the coordinates were shifted by four nu-
Set A Set C Set B cleotides (Fig. 7 C and D). For the pooled sequenced reads from 26
first trimester cases, a reduction in the difference in size distribu-
tion with nucleotide shift was similarly observed (Fig. 8 C and D).

Fetal Genome Analysis of a Second Trimester Pregnancy Affected by

Ending positions for plasma DNA fragments
B across a SNP that was heterozygous in the mother Cardiofaciocutaneous Syndrome. To validate the above-mentioned
and homozygous in the fetus methodologies, we sequenced a maternal plasma sample collected
at 18 wk of gestation to 195× coverage. This case presented with
Preferred ending positions Preferred ending positions increased nuchal translucency in the first trimester. A chorionic
for fragments carrying for fragments carrying villus sample was obtained at 11 wk of gestation, which showed a
maternal-specific alleles shared alleles normal karyotype. At 14 wk of gestation, a cystic hygroma and
mild club foot deformity were detected on ultrasound scan, and
hence, microarray analysis was performed on the chorionic villus
sample to detect gene mutations that were associated with
7,527 10,534 18,829 Noonan syndrome and related conditions. The microarray analysis
Set X Set Z Set Y detected a mutation (c770A > G) on the BRAF (B-Raf proto-
oncogene, serine/threonine kinase) gene that resulted in car-
diofaciocutaneous syndrome. Because the mutation was absent in
the mother’s and father’s genomes, it was deemed a de novo
mutation. The fetus was aborted, and placental tissues were col-
Fig. 5. Analysis of ending positions for plasma DNA fragments across
(A) SNPs that were homozygous in the mother and heterozygous in the fetus
lected after the procedure.
and (B) SNPs that were homozygous in the fetus and heterozygous in the DNA from the maternal buffy coat, paternal buffy coat, and
mother. (A) Set A included preferred ending positions for fragments carrying placental tissues collected after termination was sequenced to 40×,
fetal-specific alleles. Set B included preferred ending positions for fragments 60×, and 60× human genome coverages, respectively. The fetal
carrying shared alleles. Set C included preferred ending positions for both DNA fraction in the maternal plasma was 24%. A difference of at
types of plasma DNA fragments. (B) Set X included preferred ending posi- least 8 bp was observed between the size of fragments carrying
tions for fragments carrying maternal-specific alleles. Set Y included pre- fetal-specific alleles and alleles shared by the fetus and the mother
ferred ending positions for fragments carrying shared alleles. Set Z included at over 90% of informative SNPs (Fig. S2B). Because the size
preferred ending positions for both types of plasma DNA fragments.

of sequenced reads aligning to chromosome Y was used to calcu- 1.1

late the actual fetal DNA fraction in each plasma sample. A pos-
itive correlation could be observed between the relative abundance Pearson’s r = 0.66
[denoted as the fetal/maternal (F/M) ratio] of plasma DNA with P-value < 0.001
recurrent fetal (Set A) and maternal (Set X) ends and the fetal
1.0
DNA fraction (R = 0.66, P < 0.001, Pearson correlation) (Fig. 6). A
Ratio (F/M)

median of 248 Set A ends was observed among 26 samples. A

median of 286 Set X ends was observed among those samples.

Size Distribution of Plasma DNA Fragments Terminating on the Fetal- 0.9

Specific End Positions. Because fetal DNA in maternal plasma is
shorter than the maternal-derived DNA in maternal plasma, we
compared the size distributions of plasma DNA ending on the Set
A and Set X positions. For the deeply sequenced third trimester 0.8
case, the size distribution for fragments ending at Set A positions
was shorter than those ending at Set X positions (Fig. 7 A and B).
Then, we analyzed the pooled sequenced reads from 26 first tri-
mester plasma samples used for fetal DNA fraction analysis. A 0.7
shorter size distribution was observed for fragments ending 0 5 10 15 20 25
at fetal-preferred Set A positions compared with those ending at
maternal-preferred Set X positions (Fig. 8 A and B). These results
were consistent with the fact that fetal DNA was shorter than Fetal DNA fraction (%)
maternal-derived DNA and further support the hypothesis that Fig. 6. Correlation between the relative abundance (ratio F/M) of plasma
the end positions derived from one pregnant woman can poten-
Downloaded by guest on July 8, 2021

DNA molecules with preferred fetal (Set A) and maternal (Set X) ends, and
tially be generalized to other pregnant cases. fetal DNA fraction.

E8164 | www.pnas.org/cgi/doi/10.1073/pnas.1615800113 Chan et al.


A B
Cumulave frequency (%)
Fetal- Fetal-
preferred preferred
Frequency (%)
Maternal- Maternal-
preferred preferred
ΔS
Size (bp) Size (bp)
C D
∆S (%)
∆S (%)
Size (bp) Size (bp)
Fig. 7. (A) Plasma DNA size distributions for fragments ending on fetal-pre-
ferred ending positions (Set A; blue) and fragments ending on maternal-pre-
ferred ending positions (Set X; red). A shorter size distribution was observed
for fragments ending on Set A positions compared with those ending on Set X
positions. (B) Cumulative plot for the size distributions for the two sets of
fragments. The difference in the cumulative frequencies of the two sets of
fragments (ΔS) against fragment size is shown. (C) A plot of ΔS against size
when the end coordinates are shifted by 0 to +5 bp with regard to the ge-
nomic coordinates of the preferred end. (D) A plot of ΔS against size when the
end coordinates are shifted by 0 to −5 bp with regard to the genomic coor-
dinates of the preferred end.
distribution of plasma DNA fragments carrying the alleles shared
by the fetus and the mother is dependent on the fractional fetal
DNA concentration in plasma, the size difference required for
filtering de novo mutation would need to be determined on a case
by case basis. In this case, at least 8-bp size difference between the
variant and the parental alleles was used as the criterion for
qualifying the variant as a candidate de novo mutation; 75 can-
didate de novo mutations were identified in the plasma, of which
47 were confirmed to be present in the placenta. Because a total of
58 de novo mutations were detected in the placenta, the plasma
analysis had a detection rate of 81% and a PPV of 62% (Fig. S4).
The de novo BRAF mutation was among 47 variants detected by
the plasma DNA analysis. Using GRAD analysis, the maternal
inheritance of the fetus was deduced in 528,008 (68% of 775,456
SNPs where the mother was heterozygous) SNPs with an accuracy
of 96.8%. Regarding preferred end sites, plasma DNA fragments
carrying the most prevalent 0.5% of plasma DNA ends repre-
sented 2.4% of all DNA fragments. Strikingly, 59% of the most
prevalent 0.5% plasma DNA end sites overlapped with those of
the third trimester case. Based on the analysis of informative
SNPs, 10,401 fetal-preferred (Set A) (Fig. S5A) and 6,562 ma-
ternal-preferred (Set X) (Fig. S5B) end sites were identified. The
preferred end sites for fragments carrying alleles shared between
the mother and the fetus were also identified (Set B, Set C, Set Y,
and Set Z) (Fig. S5). The genomic coordinates of each set of
preferred positions of this case are shown in Dataset S2. In 26 first
trimester maternal plasma samples, a positive correlation could be
observed between the relative abundance (i.e., the F/M ratio) of
Downloaded by guest on July 8, 2021
the plasma DNA molecules with fetal-preferred (Set A) and
Chan et al.
PNAS PLUS
maternal-preferred (Set X) ends and the fetal DNA fraction (R =
0.66, P < 0.001, Pearson correlation) (Fig. S6).
Discussion
In this work, we performed ultradeep genome-wide sequencing of
SEE COMMENTARY
the plasma DNA obtained from two pregnant women to 195× and
270× haploid genome coverages. We believe that these are the
deepest genome-wide sequencings yet reported for plasma DNA
obtained from any one pregnancy. Previous efforts in the genome-
wide sequencing of plasma DNA obtained from pregnant women
had attempted sequencing depths from 52.7× to 78× haploid ge-
nome coverages (8–10). The depth of sequencing achieved in our
analyzed case has allowed us to push forward the limit and expand
the applications of NIPT.
First, we investigated the possibility of searching for fetal de
novo mutations on a genome-wide scale using the maternal
plasma DNA sequencing data. It has been reported that there are
∼74 de novo point mutations per person and that the mutation
rate increases with paternal age (23). It is now increasingly rec-
ognized that de novo mutations play an important role in human
genetic diseases and that diseases associated with de novo muta-
tions are not rare collectively (23). Thus, it is clinically relevant to
be able to screen for de novo mutations prenatally. Efforts in the
noninvasive prenatal detection of fetal de novo mutations are
limited to diseases associated with structural abnormalities de-
tectable on ultrasound and where those diseases are known to be
associated with hotspot de novo mutations (24). For example,
A B
Cumulative frequency (%)
Fetal- Fetal-
preferred preferred
Frequency (%)
Maternal- Maternal-
preferred preferred
MEDICAL SCIENCES
ΔS
Size (bp) Size (bp)
C D
ΔS (%)
ΔS (%)
Size (bp) Size (bp)
Fig. 8. (A) Plasma DNA size distributions in a pooled plasma DNA sample from
26 first trimester pregnant women for fragments ending on fetal-preferred
ending positions (Set A; blue) and fragments ending on maternal-preferred
ending positions (Set X; red). A shorter size distribution was observed for
fragments ending on Set A positions compared with those ending on Set X
positions. (B) Cumulative plot for the size distributions for the two sets of
fragments. The difference in the cumulative frequencies of the two sets of
fragments (ΔS) against fragment size is shown. (C) A plot of ΔS against size
when the end coordinates are shifted by 0 to +5 bp with regard to the genomic
coordinates of the preferred end. (D) A plot of ΔS against size when the end
coordinates are shifted by 0 to −5 bp with regard to the genomic coordinates of
the preferred end.
PNAS | Published online October 31, 2016 | E8165
 >90% of achondroplasia cases are caused by a de novo mutation
in the FGFR3 (fibroblast growth factor receptor 3) gene with no
prior familial history of skeletal dysplasia (25). A clinical suspicion
for achondroplasia could, therefore, be followed by maternal
plasma DNA analysis with the aim to specifically detect fetal
FGFR3 mutations (26).
Genome-wide scanning for fetal de novo mutations without a
priori information on potentially affected genes or diseases is a
huge technological challenge. Previous efforts in this area had
reported a PPV of 0.438% at a sensitivity of 38.6% (10). Hence,
using the depth of sequencing as a foundation, we used a series of
bioinformatics filters taking into account the sequencing error rate,
minimizing the realignment errors, and considering the plasma
fraction of a putative mutation as well as the size difference be-
tween fetal and maternal DNA in maternal plasma. Using such
filters, we were able to detect fetal de novo mutations with a PPV
of 74% at a sensitivity of 85% for the third trimester case. For the
second trimester case, we achieved a PPV of 62% and a sensitivity
of 81%. The PPVs, in particular, represent a significant step for-
ward over previous efforts and are at a level that suggests that, with
additional improvements, such an approach might eventually have
clinical impact. It is noteworthy that the approach successfully
detected the de novo BRAF mutation that caused the craniofa-
ciocutaneous syndrome in the second trimester case. For the
analysis of cases with earlier gestational ages and lower fetal DNA
fractions, higher sequencing depths would be required to achieve
similar levels of performance. We should perhaps add a cautionary
note that, although we have shown the technological feasibility of
detecting fetal de novo mutations from maternal plasma, the in-
terpretation of the clinical implications of each of these mutations
is a great challenge that would require a well-supported clinical
genetics infrastructure and continual improvements in our un-
derstanding of the human genome.
Second, the depth of maternal plasma DNA sequencing and
refined bioinformatics strategies used in this study allow one to
determine the maternal inheritance of the fetus on a genome-wide
level without resorting to maternal haplotype dosage analysis (8).
Using GRAD analysis, the maternal inheritance of the fetus could
be resolved in 94.2% of 656,676 heterozygous SNPs in the ma-
ternal genome of the third trimester pregnancy and 68% of
775,456 heterozygous SNPs in the genome of the second trimester
pregnant woman. In our previous study on noninvasive fetal ge-
nome analysis, the maternal inheritance was deduced using
RHDO analysis (8). The mean size of the haplotype blocks was
409 kb. For other studies involving the noninvasive fetal genome
analysis, the mean size of haplotype blocks with maternal in-
heritance determined ranged from 300 kb to over 1 Mb (9, 10, 27).
For our previous study (8), the maternal inheritance of 7,332
haplotype blocks was successfully determined with an accuracy of
99.1%. The present work, thus, represents an ∼90-fold (i.e.,
656,676/7,332 and 716,646/7,332 for the third and second trimester
cases, respectively) increase in resolution in elucidating the ma-
ternal inheritance of the fetus.
Third, the deep sequencing of maternal plasma DNA samples
prepared using non–PCR-amplified libraries has allowed us to
investigate if there might be genomic locations that would be
preferentially represented at the ends of plasma DNA fragments
and whether such ends would exhibit differences depending on
their tissue of origin (i.e., from the placenta of the fetus or the
mother). Our data indicate that such recurrent ends are indeed
present in plasma DNA. For example, our data from the third
trimester case revealed that plasma DNA fragments carrying the
most prevalent 0.5% of plasma DNA ends represented 3.5% of all
DNA fragments and that 25% of the fragments had at least one
additional fragment sharing identical ends at both ends. In-
terestingly, the preferred ending positions identified in the third
trimester case were also detectable in 26 first trimester cases, even
Downloaded by guest on July 8, 2021
when conventional PCR-amplified DNA libraries were sequenced
E8166 | www.pnas.org/cgi/doi/10.1073/pnas.1615800113
to a relatively shallow depth. These observations indicate that the
ending sites that we have identified are indeed preferential plasma
DNA fragmentation sites that are detectable recurrently across
different samples at different gestational ages and with more than
one library preparation protocol.
Furthermore, we have found that there were subsets of such
recurrent plasma DNA ends that were preferentially associated
with fetal or maternal DNA (Figs. 3, 4, and 5). Using such fetal- or
maternal-preferred ending sites, one could estimate the fetal DNA
fraction in a particular sample (Fig. 6 and Fig. S6). This latter
correlation is particularly striking, because the fetal- and maternal-
preferred ending sites were originally identified in the third and
second trimester cases. However, the abundance of such fetal-
specific ending sites correlated with the fetal DNA fractions in the
series of independent first trimester cases.
For two cases in the study that were subjected to ultradeep
sequencing (270× and 195× haploid genome coverages), it was
desirable to use PCR-free library preparations, because the
probability of encountering PCR duplicates increased with the
depth of sequencing. Furthermore, the use of PCR-free libraries
was also the most convincing approach to answer the fundamental
question as to whether there were preferred ending sites for
plasma DNA, because one could positively identify them without
the worry that one was seeing an artefactual result (i.e., PCR
duplicates).
For using the recurrent ends to deduce the fetal DNA fractions
for the first trimester samples, we had used PCR-amplified se-
quencing libraries. However, the depth of sequencing was very
shallow (median of 15 million reads per case; i.e., 0.8× haploid
genome coverage). Hence, one would not expect any of the pre-
ferred end sites to appear more than once in a particular sample.
Consequently, the filtering step that one normally performed to
remove PCR duplicates, leaving one read for subsequent bio-
informatic analysis, would not be expected to bias the results. Our
demonstration of a good correlation between the F/M ratio (based
on preferred ends) and the fetal DNA fraction is a testimony to
our belief.
The plasma DNA fragments with the fetal-preferred ending
sites had a shorter size distribution than those with the maternal-
preferred ending sites (Fig. 7). Most importantly, this size differ-
ence was observed in not only the third trimester case in which the
recurrent ends were identified but also, 26 independent first tri-
mester cases (Fig. 8). It is also interesting to note that, as we
shifted the end coordinates by one to five nucleotides on both
sides of the fetal- and maternal-preferred end sites, the size dif-
ference rapidly disappeared (Fig. 7 C and D). A similar finding
was also observed from 26 first trimester cases (Fig. 8 C and D).
These data suggest that there is a high level of specificity in the
process of plasma DNA fragmentation to the extent where, in
some parts of the genome, the cleavage or breakage occurs pref-
erentially at specific base locations.
There are active recent discussions on the nucleosomal origin of
plasma DNA. One clue to such an origin came from work on the
high-resolution size profiling of maternal plasma DNA (8, 28).
Such work has revealed that the size profile of the total DNA in
maternal plasma, in which a predominant proportion is from the
pregnant mother, contains a dominant population of plasma DNA
molecules with a size of 166 bp. Such data have also shown that,
when one focuses on the placentally derived fetal DNA in ma-
ternal plasma, one would observe a predominant population with
a size of 143 bp. The 166-bp molecules have been postulated to
represent molecules containing the nucleosome core plus the
linker (8). The 143-bp molecules, however, have been postulated
to represent molecules containing the nucleosomal core without
the linker. Straver et al. (17) created an artificial “nucleosome
track” by pooling maternal plasma DNA sequencing data from
multiple cases. They found that the frequency of reads with
starting sequences within regions 73 bp upstream and downstream
Chan et al.

of the deduced nucleosome center showed a positive correlation
with the fetal DNA fraction. Outside of the context of pregnancy,
Snyder et al. (16) also explored the nucleosomal footprint of
plasma DNA. They developed a metric that they called the win-
dowed protection score, which has been defined as the number of
molecules spanning a particular genomic window minus those
molecules with an end point within the window. Conclusions
drawn from those studies are that plasma DNA cleavages or
breakages cluster around locations in the genome in relation to
positions of nucleosome arrays and open chromatin domains.
However, as shown in Figs. 7 and 8, the preferred plasma DNA
ending sites identified by this study are exquisitely sensitive to the
genomic location. We have, therefore, shown that there is another
layer of complexity associated with the nonrandom nature of
plasma DNA fragmentation that is beyond factors just related to
genome architecture or structural domains. Additional studies are
needed to understand the relationship between the various factors
and mechanisms resulting in the highly orchestrated and location-
specific patterns of plasma DNA fragmentation. We also believe
that a catalog of such tissue-specific end sites would provide an
alternative approach for elucidating the origin of plasma DNA in
addition to methods based on methylation analysis (29) and nu-
cleosome foot printing (16).
In summary, we have developed a second generation approach
that produces noninvasive fetal genomes at high resolution using
maternal plasma DNA sequencing. We believe that this work has
significantly pushed forward the limit of NIPT by showing the
feasibility of detecting fetal de novo mutations on a genome-wide
level from maternal plasma. We have also substantially increased
the resolution in which one could determine the maternal in-
heritance of the fetus using a nonhaplotype-based approach.
Currently, the potential clinical implementation of this approach is
limited by the costs involved in sequencing maternal plasma DNA
to the depth reported here (195× and above; e.g., over US $40,000
in our center). However, with the continual reduction in the costs
of massively parallel sequencing, it is envisioned that this cost
barrier would eventually be removed. Finally, our ΔS demonstra-
tion of preferred end sites in plasma DNA that exhibit specificity of
their origin (e.g., the placenta of the fetus) has opened up nu-
merous avenues of investigation. In the context of NIPT, it would
allow a relatively simple approach for estimating the fetal DNA
fraction. Additional work would need to determine the relative
merit between this approach and others previously reported [e.g.,
using plasma DNA size (28) and nucleosomal profiles (17)]. Ac-
tually, these methods could potentially be used in combination for
NIPT. It would also be of great interest to explore the presence of
such recurrent end positions in clinical and pathological scenarios
outside of the pregnancy context (e.g., cancer).
Materials and Methods
Clinical Samples. The study was approved by the Joint Chinese University of
Hong Kong and Hospital Authority New Territories East Cluster Clinical Re-
search Ethics Committee. For each of the two cases whose plasma DNA samples
were sequenced to high depths, 20 mL maternal venous peripheral blood was
collected; 10 mL paternal venous peripheral blood was also collected. For the
third trimester pregnant case, 3 mL cord blood was collected after delivery. For
the case with cardiofaciocutaneous syndrome, the placenta was collected
after therapeutic abortion. Twenty-six first trimester pregnant women were
recruited for fetal DNA fraction analysis.
Sample Processing. Blood samples were centrifuged at 1,600 × g for 10 min at
4 °C. The plasma portion was harvested and recentrifuged at 16,000 × g for
10 min at 4 °C to remove the blood cells. The blood cell portion was recentri-
fuged at 2,500 × g, and any residual plasma was removed. DNA from the blood
cells and that from maternal plasma were extracted with the blood and body
fluid protocol of the QIAamp DNA Blood Mini Kit and the QIAamp DSP DNA
Blood Mini Kit (Qiagen), respectively. DNA from the placenta was extracted
Downloaded by guest on July 8, 2021
with the QIAamp DNA Mini Kit (Qiagen) according to the manufacturer’s
tissue protocol.
Chan et al.
PNAS PLUS
DNA Library Construction. For the two cases subjected to deep sequencing, DNA
libraries for the genomic DNA samples and the maternal plasma sample were
constructed with the TruSeq DNA PCR-Free Library Preparation Kit (Illumina)
according to the manufacturer’s protocol, except that one-fifth of the indexed
adapter was used for plasma DNA library construction. For the construction of
plasma DNA libraries, DNA extracted from 8 mL plasma was used. For the
SEE COMMENTARY
genomic DNA samples, including the mother’s buffy coat DNA, the father’s
buffy coat DNA, and the cord blood buffy coat DNA, 1 μg sonicated DNA was
used for library construction. The library concentrations ranged from 34 to
58 nM in a 20-μL library. For the first trimester, plasma DNA sequencing libraries
were constructed using a KAPA Library Preparation Kit (Kapa Biosystems).
Sequencing of DNA Libraries. The libraries for maternal plasma DNA and ge-
nomic DNA were sequenced using the HiSeq2500 and the HiSeq1500 Se-
quencing Systems (Illumina), respectively, using a paired end sequencing
protocol. Seventy-five nucleotides were sequenced from each end.
Alignment of Sequencing Data. The paired end sequencing data were analyzed
by means of the SOAP2 in the paired end mode (20). The paired end reads were
aligned to the nonrepeat-masked reference human genome (hg19). Up to two
nucleotide mismatches were allowed for the alignment of each end. The ge-
nomic coordinates of these potential alignments for the two ends were then
analyzed to determine whether any combination would allow the two ends to
be aligned to the same chromosome with the correct orientation spanning an
insert size ≤600 bp and mapping to a single location in the reference human
genome. On average, 86% of the reads were aligned to the reference human
genome (hg19). For the detection of fetal de novo mutations from maternal
plasma, an additional step of realignment using the BOWTIE2 software was
performed (19). All sequenced reads covering the candidate mutation sites
and exhibiting the variant alleles were realigned using BOWTIE2.
Binomial Mixture Model Analysis for Paternal Inheritance. To deduce the pa-
ternal inheritance of the fetus using maternal plasma DNA, we focused on SNPs
at which the mother was homozygous (genotype AA) and the father was
heterozygous (genotype AB). At these SNPs, the genotype of the fetus would be
either AA or AB. Hence, we used a two-component binomial mixture model to
fit the observed allelic counts at each SNP locus (21). For each SNP locus i, the
counts for the A and B alleles are denoted as ai and bi, respectively, and Ni is
the total number of reads covering the locus. Binomial mixture model was
then used to determine if the fetus would be homozygous (AA) or hetero-
zygous (AB) for each locus i based on the observed allelic counts ai and bi, the
sequencing error rate, and the fetal DNA fraction (21).
MEDICAL SCIENCES
GRAD Analysis for Maternal Inheritance. For SNPs in which the mother was
heterozygous, we tested whether the two alleles were present in maternal
plasma at the same or different concentrations using the sequential probability
ratio test (SPRT). An odds ratio of 20 was used for the calculation of the
threshold for accepting the null or the alternative hypothesis. The null hy-
pothesis for each SPRT analysis was the absence of dosage imbalance between
the read counts for the two maternal alleles. The alternative hypothesis was the
overrepresentation of one allele. The calculation of the upper and lower
boundaries of the SPRT curves was as previously described (8).
Dynamic Cutoff Analysis for de Novo Mutation. Dynamic cutoff filtering criteria
were developed to distinguish de novo mutations of the fetus from sequencing
errors. The sequencing error rate was assumed to be 0.3% (30). The probability
that the same variant would be observed in a number of sequenced reads
because of sequencing errors (Perr) would be calculated as

  e ðbi −1Þ
Ni
Perr = ×e× ,
bi 3
 
Ni
where represents a mathematical combination function Ni !=bi !ðNi − bi Þ!, e
bi
represents the sequencing error rate, Ni represents the total number of se-
quenced reads covering the locus i, and bi represents the number of sequenced
reads carrying the variant. The dynamic cutoff values were determined based
on a theoretical false-positive rate of <1 in 3 × 109. In other words, the the-
oretical number of false-positive sites caused by sequencing errors would be
less than one for a haploid genome.
Plasma DNA End Position Analysis. To screen for genomic coordinates that were
preferred ending positions for maternally and fetally derived DNA fragments,
fetal-specific informative SNP sites (i.e., at which the mother was homozygous
and the fetus was heterozygous) and maternal-specific informative SNP sites
(i.e., at which the fetus was homozygous and the mother was heterozygous)
PNAS | Published online October 31, 2016 | E8167
 were identified. For each informative SNP site, all of the sequencing reads
covering the informative SNP sites were analyzed.
For the analysis of SNPs where the mother was homozygous (genotype AA)
and the fetus was heterozygous (genotype AB), the A allele would be the
“shared allele,” and the B allele would be the fetal-specific allele. The number
of sequenced reads carrying the shared allele and the fetal-specific allele
would be counted. In the size distribution of plasma DNA, a peak would be
observed at 166 bp for both the fetally and maternally derived DNA. If the
fragmentation of the plasma DNA had been random, the two ends would be
evenly distributed across a region 166 bp upstream and 166 bp downstream of
the informative SNP. A P value would be calculated to determine if a particular
position has a significantly increased probability for being an end for the reads
carrying the shared allele or the fetal-specific allele based on a Poisson prob-
ability function:
 
P  value = Poisson Nactual , Npredict ,
where Poisson() is the Poisson probability function, Nactual is the actual number
of reads ending at the particular nucleotide, and Npredict is the total number of
reads divided by 166.
A P value of <0.01 was used as a cutoff to define preferred ending positions
for the reads carrying the fetal-specific allele or the shared allele.
For the SNPs where the mother was heterozygous (AB) and the fetus was
homozygous, the reads carrying the maternal-specific allele (B allele) and the
shared alleles were analyzed.
1. Lo YM, et al. (1997) Presence of fetal DNA in maternal plasma and serum. Lancet
350(9076):485–487.
2. Chiu RW, et al. (2008) Noninvasive prenatal diagnosis of fetal chromosomal aneu-
ploidy by massively parallel genomic sequencing of DNA in maternal plasma. Proc
Natl Acad Sci USA 105(51):20458–20463.
3. Chiu RW, et al. (2011) Non-invasive prenatal assessment of trisomy 21 by multiplexed
maternal plasma DNA sequencing: Large scale validity study. BMJ 342:c7401.
4. Bianchi DW, et al.; CARE Study Group (2014) DNA sequencing versus standard pre-
natal aneuploidy screening. N Engl J Med 370(9):799–808.
5. Norton ME, et al. (2015) Cell-free DNA analysis for noninvasive examination of tri-
somy. N Engl J Med 372(17):1589–1597.
6. Srinivasan A, Bianchi DW, Huang H, Sehnert AJ, Rava RP (2013) Noninvasive detection
of fetal subchromosome abnormalities via deep sequencing of maternal plasma. Am J
Hum Genet 92(2):167–176.
7. Yu SC, et al. (2013) Noninvasive prenatal molecular karyotyping from maternal
plasma. PLoS One 8(4):e60968.
8. Lo YM, et al. (2010) Maternal plasma DNA sequencing reveals the genome-wide
genetic and mutational profile of the fetus. Sci Transl Med 2(61):61ra91.
9. Fan HC, et al. (2012) Non-invasive prenatal measurement of the fetal genome. Nature
487(7407):320–324.
10. Kitzman JO, et al. (2012) Noninvasive whole-genome sequencing of a human fetus.
Sci Transl Med 4(137):137ra76.
11. Fan HC, Wang J, Potanina A, Quake SR (2011) Whole-genome molecular haplotyping
of single cells. Nat Biotechnol 29(1):51–57.
12. Kitzman JO, et al. (2011) Haplotype-resolved genome sequencing of a Gujarati Indian
individual. Nat Biotechnol 29(1):59–63.
13. Lam KW, et al. (2012) Noninvasive prenatal diagnosis of monogenic diseases by targeted
massively parallel sequencing of maternal plasma: Application to β-thalassemia. Clin
Chem 58(10):1467–1475.
14. New MI, et al. (2014) Noninvasive prenatal diagnosis of congenital adrenal hyper-
plasia using cell-free fetal DNA in maternal plasma. J Clin Endocrinol Metab 99(6):
E1022–E1030.
15. Zeevi DA, et al. (2015) Proof-of-principle rapid noninvasive prenatal diagnosis of
autosomal recessive founder mutations. J Clin Invest 125(10):3757–3765.
16. Snyder MW, Kircher M, Hill AJ, Daza RM, Shendure J (2016) Cell-free DNA Comprises
an in vivo nucleosome footprint that informs its tissues-of-origin. Cell 164(1-2):57–68.
Downloaded by guest on July 8, 2021
E8168 | www.pnas.org/cgi/doi/10.1073/pnas.1615800113
Expected Proportion of Fragments Sharing Two Identical Ends. The probability
of having n fragments sharing the same ending position on one side, P(n), was
calculated using a Poisson probability function assuming that the average size
of plasma DNA fragments was 166 bp and that the average sequencing depth
was 270×:


270
PðnÞ = Poisson n, .
166
The probability of two fragments having the same size (Psize) can be calcu-
lated from the size distribution of the plasma DNA fragments:
X
600
Psize = ½%ðsizeÞ2 ,
size=20
where %(size) is the proportion of fragments having the particular size.
Therefore, the overall probability of finding two fragments with identical
ends on both sides (Pboth) can be calculated as
X
∞

n
Pboth = PðnÞ × × Psize ,
2
n=2
 
n
where represents a mathematical combination function n!=2!ðn − 2Þ!.
2
ACKNOWLEDGMENTS. This work was supported by the Hong Kong Research
Grants Council Theme-Based Research Scheme T12-403/15. Y.M.D.L. was
supported by an endowed chair from the Li Ka Shing Foundation.
17. Straver R, Oudejans CB, Sistermans EA, Reinders MJ (2016) Calculating the fetal
fraction for noninvasive prenatal testing based on genome-wide nucleosome profiles.
Prenat Diagn 36(7):614–621.
18. Chandrananda D, Thorne NP, Bahlo M (2015) High-resolution characterization of
sequence signatures due to non-random cleavage of cell-free DNA. BMC Med
Genomics 8:29.
19. Langmead B, Salzberg SL (2012) Fast gapped-read alignment with Bowtie 2. Nat
Methods 9(4):357–359.
20. Li R, et al. (2009) SOAP2: An improved ultrafast tool for short read alignment.
Bioinformatics 25(15):1966–1967.
21. Jiang P, et al. (2012) FetalQuant: Deducing fractional fetal DNA concentration from
massively parallel sequencing of DNA in maternal plasma. Bioinformatics 28(22):
2883–2890.
22. Lun FM, et al. (2008) Noninvasive prenatal diagnosis of monogenic diseases by digital
size selection and relative mutation dosage on DNA in maternal plasma. Proc Natl
Acad Sci USA 105(50):19920–19925.
23. Veltman JA, Brunner HG (2012) De novo mutations in human genetic disease. Nat Rev
Genet 13(8):565–575.
24. Chitty LS, et al. (2015) Non-invasive prenatal diagnosis of achondroplasia and tha-
natophoric dysplasia: Next-generation sequencing allows for a safer, more accurate,
and comprehensive approach. Prenat Diagn 35(7):656–662.
25. Vajo Z, Francomano CA, Wilkin DJ (2000) The molecular and genetic basis of fibro-
blast growth factor receptor 3 disorders: The achondroplasia family of skeletal dys-
plasias, Muenke craniosynostosis, and Crouzon syndrome with acanthosis nigricans.
Endocr Rev 21(1):23–39.
26. Saito H, Sekizawa A, Morimoto T, Suzuki M, Yanaihara T (2000) Prenatal DNA di-
agnosis of a single-gene disorder from maternal plasma. Lancet 356(9236):1170.
27. Chen S, et al. (2013) Haplotype-assisted accurate non-invasive fetal whole genome
recovery through maternal plasma sequencing. Genome Med 5(2):18.
28. Yu SC, et al. (2014) Size-based molecular diagnostics using plasma DNA for non-
invasive prenatal testing. Proc Natl Acad Sci USA 111(23):8583–8588.
29. Sun K, et al. (2015) Plasma DNA tissue mapping by genome-wide methylation se-
quencing for noninvasive prenatal, cancer, and transplantation assessments. Proc Natl
Acad Sci USA 112(40):E5503–E5512.
30. Ross MG, et al. (2013) Characterizing and measuring bias in sequence data.
Genome Biol 14(5):R51.
Chan et al.