Fetal fraction calculation
Three different methods to calculate fetal fractions were used in this
study for different purposes, including an artificial neural network
model named the FF-QuantSC [18], a Y-chromosome method [19], and
the relative coverage of the trisomy chromosome. FF-QuantSC was used to
calculate the overall fetal fraction of both co-twins. In pregnancies
with one male co-twin, the Y-chromosome method was used to calculate the
fetal fraction of the male fetus. In pregnancies with two male fetuses,
the Y-chromosome method calculated the fetal fraction equal to an
overall fetal fraction, and results were compared with
the FF-QuantSC method. The
relative coverage of the trisomy chromosome was specifically used to
calculate the fetal fraction of the trisomy co-twin who was terminated
by the selective reduction.
FF-QuantSC exploits an artificial neural network to calculate the fetal
fraction of both male and female co-twins [18]. Briefly, a fully
connected neural network model with a single hidden layer and 128
neurons was trained by the Vapnik-Chervonenkis dimension [20] and
trained with over 100 thousand of male pregnancies. Sequencing reads
were partitioned into continuous genomic windows of 60kb in length for
feature selection. Standardization by within-sample z-score
transformation was then used to generate a final feature matrix. After
training, the FF-QuantSC model was testified in six testing groups of
about 240 thousand pregnancies, including 36 thousand twin pregnancies
and 80 thousand female pregnancies [18].
Y-chromosome method calculated fetal fractions using unique sequence
reads from the Y chromosome (eq.1).
Fetal fraction =
2*\(\frac{\overset{\overline{}}{\text{UR}_{Y}}}{\overset{\overline{}}{\text{UR}}}\)(1)
where \(\overset{\overline{}}{\text{UR}_{Y}}\) represents the unique
reads on chromosome Y, and \(\overset{\overline{}}{\text{UR}}\)represents the unique reads on autosome chromosomes.
In pregnancies with a trisomy co-twin, fetal fraction was calculated
using the sequencing reads of the trisomic chromosome, typically
chromosome 13, 18, or 21 (eq.2).
\(\varepsilon_{i}=2\times\frac{\text{UR}_{i}-\overset{\overline{}}{\text{UR}}}{\overset{\overline{}}{\text{UR}}}\)i =chromosome 13, 18, or 21 (2)
Where \(\varepsilon_{i,Y}\) represents the fetal fraction estimate by
chromosome \(i\); URi represents the unique reads
of chromosome \(i\); \(\overset{\overline{}}{\text{UR}}\) represents the
mean of unique reads on all autosomal chromosomes.