Page 27 - GTM-3-3

Global Translational Medicine Computational advances in cancer liquid biopsy

R package called SIGNALS allowed users to go beyond and patterns of fragmentation in cfDNA can provide
55
total CNV information to identify cell heterogeneity additional information beyond the genetic analysis of
phasing copy number events to individual homologs in somatic mutations and CNV abnormalities, offering hope
single-cell genomes, quantifying haplotype-specific CNV for patients affected by cancers with limited genetic lesions,
distributions, and thereby validating copy number events such as Ewing sarcoma. The fragmentation patterns of
64
using variant allele fractions (Table 1). DNA released by tumor cells do not appear to be random,
Alongside genomics, it is now possible to produce nor entirely determined by the DNA primary sequence.
proteomic, 56-58 transcriptomic, 56,59,60 and epigenomic data 61-63 Instead, chromatin structure and the epigenetic states of
from a single CTC, providing a more complete picture of the cells from which the DNA fragments originated play
the molecular mechanisms underlying the disease. a significant role in shaping them. 21,65,66 As a result, ctDNA
typically has shorter and more heterogeneous fragment
3. Epigenomics lengths compared to normal cfDNA. Normal cfDNA
Extensive DNA methylation perturbations have been fragments are usually around 160 – 200 base pairs (bp)
widely explored in human cancer, but very few studies in length, reflecting the size of nucleosomal units from
have focused on analyzing this phenomenon in CTCs, 61-63 apoptotic cells. In contrast, ctDNA fragments are often
probably due to the technical difficulties of isolating shorter, with a significant proportion of fragments being
and performing epigenetic analysis on such rare cells. <150 bp.
Nevertheless, significant results have been obtained in the As several authors have pointed out, integrating
field, shedding light on the molecular mechanism of cancer multiple sequencing modalities – including epigenomics,
cell stemness and dissemination. For instance, Gkountela fragmentomics, and genomics – would provide the
et al. performed a genome-wide analysis of CTC’s DNA- best performance for early cancer detection, improving
62
methylome, demonstrating how CTC clusters differ from sensitivity, specificity, and robustness of prediction.
67
single CTCs in their methylation profiles. They associated Future efforts should consider the high variability in
CTC clustering with defined changes in DNA methylation fragmentation between different individuals depending
that enhance stemness and promote metastasis. on genetics, comorbidities, and lifestyle factors, to build

Significant technical progress has been made with methodologies that consider intersample variability and
cfDNA. Recent research suggests that chromatin signatures define patient-specific thresholds.

Table 1. Summary table of CNV analysis tools
Tool name Developed for Normal Programming Coverage Sample References
required language interface
IchorCNA LP-WGS Yes R Low cfDNA 19
Ginkgo LP-WGS No GUI-cloud-based Low Single cell DNA 44
AneuFinder LP-WGS No R Low Single cell DNA 53
SCOPE LP-WGS Yes R Low Single cell DNA 54
SIGNALS LP-WGS Yes R Low Single cell DNA 55
SCNV LP-WGS Yes Command line/R Low Single cell DNA 45
baseqCNV LP-WGS No Python Low Single cell DNA 51
SCCNV LP-WGS Yes Python Low Single cell DNA 125
SCICoNE LP-WGS No Python Low Single cell DNA 52
CHISEL LP-WGS Yes Command line/ Low Single cell DNA 126
Python
Sequenza WES and WGS Yes Command line/R High Bulk DNA 127
CNVkit WES, WGS, and TAS Yes Command line High Bulk DNA 46
ASCAT WES, WGS, and TAS Yes R High Bulk DNA 47
Control-FREEC WGS, WGS, and TAS No Command line High Bulk DNA 128
InferCNV scRNASeq Yes R High Single cell RNA https://github.com/broadinstitute/inferCNV
Numbat scRNASeq No R High Single cell RNA 129
Abbreviations: CNV: Copy number variation; LP-WGS: Low-pass whole-genome sequencing; cfDNA: Cell-free DNA.

Volume 3 Issue 3 (2024) 4 doi: 10.36922/gtm.3063