MrBam will give numbers of unique reads for a combination of
following conditions: supporting reference or alternative, clustered
by single or multiple reads, and locating in overlapped or
non-overlapped region.
The result of MrBam can be used to filter variants called from
ctDNA sequencing data. According to our experience, to report a
mutation, we need at least two unique read pairs supporting it, and
each pair should either have this mutation in its overlapped region
or be a consensus pair generated by multiple pairs. Due to the high
ratio sequencing error and extreme depth of ctDNA sequencing
data, the mutations only supported by a few single reads at their
non-overlapped regions are usually false positive.
MrBam is an open source project. It is developed in Python
with its source available at: http://githubs.com/OpenGene/
MrBam.2.6 Methylation
Analysis of Cell-
Free DNA
Methylation changes are common for different cancer types and
usually occur early in cancer development, typically repressing the
expression of tumor suppressor genes [38]. Aberrant DNA meth-
ylation may offer a more consistent and hence broadly applicable
marker of tumor DNA in blood than mutations [39].
There is a very large amount of published information describ-
ing DNA methylation patterns in tumor tissue and their impact on
patient prognosis. When tumor DNA is shed into the blood stream,
these patterns are also detectable in plasma and serum [40].
Tumor-specific ctDNA methylation can be used to quantitate
tumor DNA, providing information about the level of tumor bur-
den, as well as revealing the methylation patterns in the tumor.
DNA methylation-based biomarkers could be incorporated into
patient care and management with only very minor changes to
clinical practice, such as recent applications of methylated ctDNA
in determining cancer prognosis and in disease monitoring follow-
ing surgery or during chemotherapy treatment. Methylated ctDNA
assays are also developed to meet the stringent criteria required for
cancer screening.
Next-generation sequencing platforms allow the construction
of genomic maps of DNA methylation at a single-base resolution
[41]. Treating genomic DNA with sodium bisulfite deaminates
unmethylated cytosine (C) to uracil (U), while methylated C resi-
dues remain unaffected [42]. The U eventually converts to thymine
(T) in a subsequent polymerase chain reaction (PCR). Whole-
genome bisulfite sequencing (WGBS) and reduced representation
bisulfite sequencing (RRBS) are two classic methods for genome-
wide methylation study.
WGBS (BS-seq; MethylC-seq) theoretically covers all the C
information [43]. In this method, genomic DNA is purified and
sheared into fragments. The fragmented DNAs are end-repaired;
adenine bases are added to the 3^0 end (A-tailing) of the DNABioinformatics Analysis for Cell-Free Tumor DNA Sequencing Data 85