How Advanced hWGS/WES Analysis Supports Precision Oncology Research

Tumor genomes are uniquely complex. Unlike germline samples, cancer tissues carry somatic mutations, copy number alterations, structural rearrangements, clonal heterogeneity, and distinct mutational signatures resulting from historical DNA damage and repair processes. A basic variant-calling workflow can detect some single-nucleotide changes, but it often misses the broader genomic context that drives tumor evolution, treatment resistance, and metastatic potential.

At BMKGENE, we perform advanced analysis of human whole-genome (hWGS) and whole-exome sequencing (hWES) and deliver ready-to-interpret, publication-quality results. By analyzing paired tumor and matched normal samples, it becomes possible to integrate somatic and germline variant calling and perform a series of specialized tumor-focused analyses.

Below are examples of key insights that can be obtained through BMKGENE’s advanced tumor analysis workflow.

Loss of Heterozygosity (LOH) and Copy Number Variants (CNV)

Loss of heterozygosity (LOH) refers to the transition of specific genomic sites from a heterozygous state in normal tissue to a homozygous state in tumor tissue, caused by chromosome copy-number changes, gene conversion, somatic recombination, or mitotic nondisjunction. LOH is often associated with tumor suppressor genes and can directly contribute to tumorigenesis: when one allele is lost or inactivated, and the remaining allele is already impaired, the gene’s protective function is eliminated, allowing malignant transformation to occur. Complementing LOH analysis, exploring copy-number variation (CNV) distribution across samples helps identify significant CNVs and LOH that recur across similar tumor types and are likely key drivers of initiation and progression. 

Image caption:
CNVs and LOH: The figure shows a segmented CNV heatmap, where red indicates copy-number gains and blue indicates copy-number losses. The vertical axis represents genomic coordinates, while each column along the horizontal axis corresponds to an individual sample.

Tumor Signatures

By evaluating the trinucleotide context of each mutation (one base upstream and downstream), SNVs can be classified into 96 mutational types (4 × 6 × 4). Analyzing the distribution of these 96 classes enables us to define a distinct mutational signature for each tumor. By comparing these signatures to the COSMIC database, researchers can identify the underlying biological processes causing this mutational activity, such as APOBEC activity, UV exposure, tobacco-related damage, or defective DNA repair, and understand how these processes differ across samples within a dataset.

Image caption:
Tumor Signatures: The rows and columns represent the mutational signatures of the samples and the known COSMIC signatures, respectively. The color gradient (shifting toward red) indicates increasing similarity between signatures, with cosine values closer to 1 reflecting stronger correspondence.

Driver and Predisposition genes

Somatic driver mutations, whether SNVs, small InDels, or focal CNAs, provide tumor cells with a selective growth advantage, while predisposition (germline) variants increase an individual’s inherited cancer risk (for example, BRCA1/2, MLH1, ATM). Using paired tumor–normal analysis, we detect somatic drivers by combining recurrence and hotspot detection with statistical background-mutation models to predict novel drivers, and by cross-referencing known drivers against curated cancer databases. At the same time, we screen the normal sample for germline variants, annotate them against cancer databases. In this way, we produce a prioritized, evidence-backed set of somatic and germline findings that supports robust biological interpretation.

Heterogeneity analysis

Tumor heterogeneity refers to the presence of multiple genetically distinct cell  populations within a tumor. These subpopulations can differ in growth rate, invasiveness, drug sensitivity, and overall behavior — making heterogeneity an essential factor to characterize.

• Purity is the proportion of cancer cells in a sample. Low purity can dilute tumor-specific signals, making somatic variants appear at lower frequencies and harder to detect.
• Ploidy represents the average copy-number state of the tumor genome. Accurate ploidy estimation is necessary to correctly interpret copy-number changes and to calculate cancer-cell fractions.

Because purity and ploidy directly shape how mutations and copy-number changes appear in sequencing data, they form the basis for reliable clonal analysis. Using these parameters, we can group variants into clonal (early, shared) or subclonal (later, emerging) populations and reconstruct the tumor’s evolutionary history.

Why this matters: Clonal analysis helps distinguish early driver events from later resistance-associated alterations, improves interpretation of driver significance, guides sample selection for longitudinal studies, and enables researchers to compare evolutionary patterns across tumors.

Image caption:
Tumor Evolution Tree: The chart depicts the evolutionary relationships among tumor subclones within each patient.

Yichao Bu and colleagues^[1] used whole-exome sequencing performed at BMKGENE on matched specimens of hepatocellular adenoma (HCA), co-existing hepatocellular carcinoma (HCC) components, and adjacent normal liver to investigate the genomic events underlying malignant transformation. Their WES-based analysis, including somatic variant calling, CNV profiling, mutational-signature extraction (NMF), and phylogenetic reconstruction, revealed that HCA and HCC tissues within each case shared a largely similar mutational landscape and a common monoclonal origin, despite measurable genomic heterogeneity between cases. Phylogenetic trees and signature analysis highlighted both shared and distinct mutational features associated with the HCA→HCC transition, and the authors also observed elevated immune-related markers in these tumors, suggesting potential sensitivity to immunotherapy.

At BMKGENE, we provide advanced analysis of paired tumor–normal WGS and WES datasets, including the approaches described in this post. If you’re exploring tumor genomics, reach out to us to discover how we can help you gain deeper insights and advance your research.

Reference:

[1] Y. Bu et al., “Whole-exome sequencing-based mutational profiling of hepatocellular adenoma malignant transformation to hepatocellular carcinoma,” Clin Exp Pharmacol Physiol, vol. 51, no. 7, p. e13901, Jul. 2024, doi: 10.1111/1440-1681.13901