The way doctors analyze your DNA is stuck in the past, and researchers are finally fixing it. For over two decades, scientists have compared every person's genome to a single reference standard created from just a handful of anonymous genomes. Now, a growing movement in precision medicine is pushing for a complete rethinking of how we sequence, interpret, and use genomic data to diagnose disease and guide treatment.

Why Is the Current Genome Reference Outdated?

When the first human genome was sequenced in 2003, it was a historic achievement. But the reference genome that emerged from that effort has become a bottleneck for modern medicine. The reference is incomplete, capturing only the coding regions of DNA while leaving out significant portions of the genome. It also fails to represent known genetic variations that are common in real populations, making rare but clinically important mutations nearly impossible to detect.

Perhaps most critically, the reference genome is haploid, meaning it represents only one copy of each chromosome. Humans, however, are diploid; we inherit two copies of every chromosome, one from each parent. This fundamental mismatch means the current standard misses important genetic information that could affect diagnosis and treatment decisions.

"The reference genome was made by cobbling together a few anonymous genomes about 20 years ago. And while gene sequencing technology has moved on since then, our standards, the way we analyze and interpret genomes, which largely depends on the reference genome, have not," explained Euan Ashley, Arthur L. Bloomfield Professor in Medicine at Stanford University.

Euan Ashley, Arthur L. Bloomfield Professor in Medicine, Stanford University

What Is the Pangenome, and How Does It Fix These Problems?

To address these limitations, researchers are developing what's called a pangenome, an atlas of human genetic diversity rather than a single map. Instead of comparing every patient's genome to one outdated reference, the pangenome is a collection of real genomes from real people, capturing the full spectrum of human genetic variation.

The most advanced pangenomes include what scientists call "telomere to telomere" sequencing, meaning they capture the entire genome from end to end, including regions that were previously impossible to sequence. This comprehensive approach reveals genetic variations that the old reference genome simply could not detect, opening new possibilities for early disease detection and personalized treatment.

How Is AI Driving Precision Oncology Forward?

While genome sequencing standards are being reimagined, AI and machine learning are already transforming how doctors use genomic data to treat cancer. At the upcoming International Association for the Study of Lung Cancer World Conference on Lung Cancer in Seoul, researchers from Caris Life Sciences will present seven studies demonstrating how AI-driven analysis of large genomic datasets is revealing new insights into cancer biology and treatment response.

One key finding involves the relationship between two genetic markers in lung cancer. Analysis of 39,124 non-small cell lung cancer samples found that tumors with high PRMT5 expression combined with MTAP deletion showed the shortest overall survival and fewer immune cells that could fight the cancer. Among patients treated with immunotherapy, these tumors performed particularly poorly, suggesting that AI can now identify which patients need different treatment strategies.

Another study examining 43,261 lung cancer samples revealed significant differences in tumor genetics and immune response across racial and ethnic groups. Non-Hispanic Asian Pacific Islander patients showed longer survival and fewer genetic alterations associated with immunotherapy resistance compared to non-Hispanic White patients. Notably, non-Hispanic Black patients achieved longer survival despite having a higher prevalence of immunotherapy-resistant genetic markers, suggesting that genomic data alone does not tell the complete story of treatment response.

Steps to Harmonize Genomic Medicine Across Medical Fields

• Establish Common Standards: Different medical specialties currently use different approaches to interpret genomes for rare disease diagnosis, cancer treatment, and prenatal screening. Creating unified standards ensures patients receive consistent, accurate interpretations regardless of their condition or where they receive care.
• Expand Pangenome Representation: Building pangenomes that include diverse populations ensures that genetic variations are accurately captured across all ancestry groups, reducing the risk that rare but important variants are missed in underrepresented populations.
• Integrate AI Quality Control: Before gene editing tools like CRISPR are used in patients, AI-powered sequencing must verify that edits target the correct DNA locations and do not accidentally alter other parts of the genome, a problem called off-target effects.

What Real-World Impact Are These Changes Having?

The shift toward more comprehensive genomic analysis is already improving patient outcomes in specific cancers. In mesothelioma, a rare and aggressive cancer, researchers identified three distinct molecular subtypes using transcriptomic analysis of 386 samples. The findings suggest that chemotherapy with or without bevacizumab works better than immunotherapy in two of the three subtypes, pointing toward a more personalized approach to treatment selection.

Germline testing, which identifies inherited genetic mutations that increase disease risk, is also benefiting from larger-scale analysis. Among 3,609 non-small cell lung cancer patients, clinically relevant inherited mutations were found in 12.2 percent of cases, with frequently altered genes including MUTYH, CHEK2, ATM, BRCA2, and MITF. When tumor sequencing was also available, 45.8 percent of pathogenic or likely pathogenic inherited variants had a matching mutation in the tumor itself, suggesting these inherited mutations may have driven cancer development.

Genome sequencing is now being applied across multiple precision medicine applications, from detecting fetal abnormalities early in pregnancy to identifying rare diseases that might otherwise go undiagnosed for years. The technology also enables personalized cancer vaccines designed around a patient's unique tumor mutations and earlier detection of cancer recurrence.

"This year's presentations underscore the power of large-scale molecular and clinico-genomic datasets in revealing meaningful insights into cancer biology and therapeutic response," said George W. Sledge, Jr., Chief Medical Officer of Caris Life Sciences.

George W. Sledge, Jr., Chief Medical Officer, Caris Life Sciences

Why Does Harmonizing Genomic Standards Matter for Your Health?

As genome sequencing becomes more affordable and widespread, millions of people worldwide now have their DNA sequenced. Yet without unified standards for interpretation, a patient's genomic results could be analyzed differently depending on which hospital, clinic, or research program processes the data. This inconsistency could lead to missed diagnoses, unnecessary treatments, or delayed care.

The convergence of improved sequencing technology, AI-driven analysis, and large-scale genomic databases is creating the foundation for truly personalized medicine. By moving beyond a single outdated reference genome and embracing the pangenome, researchers can ensure that every patient benefits from the most accurate interpretation of their genetic blueprint, regardless of their ancestry or the condition they face.