Genomic Data: The Foundation of Molecular Testing

Normal (Germline) vs. Tumor Tissue Genomic Data

Genomic data examines DNA sequences, but the source matters enormously. Germline genomic data comes from normal tissue (blood, saliva) and represents your inherited genetic blueprint - the DNA you were born with. When we test for BRCA1/BRCA2 mutations in blood, we're looking for inherited predispositions passed down through families.

Tumor genomic data, by contrast, examines DNA from cancer cells themselves. These contain both your original inherited DNA plus additional somatic mutations - genetic changes that occurred during cancer development. A person might have normal germline BRCA1/BRCA2 genes but develop somatic BRCA mutations specifically in their tumor tissue.

Expression Data: From Proteins to Transcriptomes

Expression data measures what genes are actually doing rather than just what the DNA sequence says. This represents the dynamic, functional output of the genome.

Standard Expression Markers (ER/PR/HER2)

These classic biomarkers use immunohistochemistry (IHC) to measure specific proteins directly in tissue. When we say a tumor is "ER+" we mean estrogen receptor proteins are detectable on cancer cell surfaces. This is targeted, focused expression analysis of known important proteins.

Whole Transcriptome Expression

RNA sequencing measures the expression levels of all ~23,000 human genes simultaneously by quantifying messenger RNA (mRNA) molecules. This provides a comprehensive snapshot of cellular activity - which genes are turned "on" or "off" and by how much.

Immune Profiling Through Expression Data

Immune profiling typically uses RNA sequencing to measure expression of immune-related genes. Tests like NanoString IO360 examine 770 immune genes to characterize the tumor microenvironment - are T-cells infiltrating? Are immune checkpoint pathways active? This helps predict immunotherapy response by understanding the immune landscape around the tumor.

Circulating Tumor DNA (ctDNA): The Liquid Biopsy Revolution

ctDNA represents fragments of tumor DNA that circulate freely in blood. When cancer cells die, they release their DNA into the bloodstream. Advanced sequencing can detect these tumor-derived DNA fragments among the much larger background of normal cell-free DNA.

Functional Testing: Beyond Genetics to Actual Behavior

Functional testing moves beyond "what mutations does the tumor have?" to "how does the tumor actually behave?" This involves growing tumor cells in laboratory conditions and directly testing their response to various treatments.

Methods

• Organoids: 3D tissue cultures that mimic tumor structure
• Drug screening: Testing multiple chemotherapy agents directly on patient cells
• Sensitivity profiling: Measuring which drugs kill tumor cells most effectively

SNP Data: Population Genetics and Common Variants

Single Nucleotide Polymorphisms (SNPs) represent common genetic variants found throughout human populations. Unlike rare disease-causing mutations, SNPs are normal genetic differences that occur in >1% of people.

Emerging and Advanced Data Types

Epigenetic Data: The Genome's Control Layer

Epigenetic modifications control gene expression without changing DNA sequence itself. DNA methylation patterns can silence tumor suppressor genes even when the underlying DNA is normal. Methylation-based tests like Galleri detect cancer by identifying abnormal methylation patterns across thousands of genes.

Proteomic Data: The Functional Output

While genomics tells us what could happen and transcriptomics what should happen, proteomics measures what actually is happening. Mass spectrometry can quantify thousands of proteins simultaneously, revealing post-translational modifications and protein interactions that don't show up in DNA or RNA data.

Metabolomic Data: Cellular Chemistry in Action

Metabolomics measures small molecules (metabolites) that represent the end products of cellular processes. Cancer cells often have altered metabolism, producing unique metabolic signatures detectable in blood or urine.

Cutting-Edge Multimodal AI Analysis

Genomics + Imaging Integration

Emerging AI approaches combine multiple data types for more powerful analysis. Radiogenomics uses machine learning to predict genetic mutations directly from medical images (CT, MRI, pathology slides). For example, AI can predict EGFR mutations in lung cancer from CT scans, or identify microsatellite instability from H&E pathology slides without additional molecular testing.

Multimodal Pathology AI

Advanced systems analyze digitized pathology slides alongside genomic data to predict treatment responses, identify missed diagnoses, or discover new biomarkers. The AI learns patterns invisible to human pathologists by processing millions of image features simultaneously.

Secondary Analysis: Making Comprehensive Data Work Harder

The Power of Comprehensive Data Collection

When a test performs whole exome sequencing (20,000+ genes) but only reports results for a specific gene panel, the unreported data doesn't disappear - it remains available for future analysis. This enables "secondary analysis" where comprehensive datasets can answer new questions without additional testing.

Examples

• Signatera: Performs whole exome sequencing on tumor and normal tissue but initially reports only 16 personalized variants for monitoring. Patients can later request analysis of cancer predisposition genes, pharmacogenomic variants, or other findings from the same dataset.
• Tempus xR: Sequences the entire transcriptome but initially focuses on clinically actionable findings. The same data can later be reanalyzed for research signatures, pathway analysis, or newly discovered biomarkers.

Current Limitations and the Research-to-Clinical Gap

Insurance and Standard of Care Barriers

Advanced multimodal analyses and many secondary analyses are not covered by insurance because they are not yet considered "standard of care." Insurance systems typically lag years behind scientific developments, requiring extensive clinical trial validation before coverage approval.

The Clinical Translation Lag

A significant gap exists between research developments and clinical application. Promising biomarkers and AI-based analyses may show strong research evidence but require 5-15 years to become clinically adopted. During this period, patients may benefit from research-grade information that their healthcare systems cannot yet provide.

Access Limitations

Most cutting-edge analyses remain confined to academic medical centers or research institutions. Standard community oncology practices may lack access to the latest developments, creating disparities in available information. Organizations like CEtHI aim to bridge this gap by providing research-grade analyses directly to patients, empowering informed decision-making without waiting for traditional clinical adoption timelines.