Genomic surveillance acts as the "eyes and ears" of public health, allowing researchers to track how pathogens evolve, spread, and develop resistance to drugs. However, this vision is often tunneled. When sequencing efforts are concentrated in specific geographic hubs or high-income nations, the resulting data creates a distorted map of global biodiversity. This geographic sampling bias does more than just leave gaps in a database; it fundamentally skews our understanding of evolutionary trajectories, transmission chains, and the efficacy of medical interventions.
The Blind Spot Problem
The core intuition behind the danger of geographic bias is the "blind spot." If a virus emerges or evolves in a region with low sequencing capacity, it remains invisible to global monitoring until it crosses into a well-monitored region. By then, the opportunity for early intervention is lost.
Anderson F. Brito and colleagues in Global disparities in SARS-CoV-2 genomic surveillance quantified this, noting that over 75% of low- and middle-income countries sequence less than 0.5% of their cases. This disparity, as Sergey Knyazev argues in Unlocking capacities of viral genomics for the COVID-19 pandemic response, means that emerging variants in under-represented regions can spread globally long before they are detected by international networks.
Breaking the Evolutionary Chain
Phylogenetic trees—the "family trees" of pathogens—rely on a continuous chain of samples to reconstruct how a virus or bacterium has changed over time. When geographic sampling is sparse, these chains break.
Dhanasekaran Vijaykrishna and Gavin J. D. Smith demonstrated this in their study of swine influenza A virus, showing that surveillance from a single regional hub (like Hong Kong) fails to capture the erratic and diverse evolution occurring in other global production centers. Similarly, Yousong Peng found that spatiotemporal gaps in H5N1 data can obscure the true pathways of viral mutation, making it difficult to predict which strains will dominate next (Continual Antigenic Diversification in China).
Without representative global data, researchers may misidentify the "hubs" of evolution. For instance, Shaofu Qiu and Jinpeng Guo highlighted how public databases are often overrepresented by resource-rich countries, potentially obscuring the role of the Indian subcontinent and Southeast Asia as critical hubs for Chikungunya virus expansion (Source-tracking of the Chinese Chikungunya viruses).
Microbial Dark Matter and Western Bias
Geographic bias also affects our understanding of the human microbiome and "last-resort" antibiotic resistance. Most genomic reference databases are built on samples from Western populations. When researchers study non-Western cohorts using these biased tools, they encounter "microbial dark matter"—sequences that cannot be identified because they don't exist in the reference set.
To combat this, Beining Ma constructed the Human Gut Microbiome Reference (GMR), adding a quarter-million genomes from Chinese populations to reveal 1,700 potential new species. Similarly, Xiuchao Wang found that 90% of the viruses in Chinese gut viromes were missing from global databases (The genetic diversity and populational specificity of the human gut virome). This lack of representation means that diagnostic and therapeutic targets—like those for malaria vaccines discussed in Distinct evolutionary regimes across domains of the Plasmodium falciparum CSP gene—may fail to account for significant regional genetic variations.
Artifacts of Sparse Sampling
A subtle but dangerous effect of geographic bias is the creation of "false breaks." In phylogeography, researchers look for genetic divides that suggest a species has been separated by a physical barrier. However, as noted in Bats of a feather, sparse or uneven sampling can make continuous genetic variation look like a discrete population break. These artifacts of data collection can lead scientists to incorrectly conclude that a species is fragmented when it is actually moving freely across a landscape.
Toward Equitable Surveillance
The consensus across these 68 papers is that the utility of genomic data is inseparable from its geographic representativeness. Whether tracking SARS-CoV-2 in Switzerland or antimicrobial resistance in aquaculture, the reliance on voluntary or centralized hospital data (Chuanning Tang) introduces systematic errors.
Building a truly global surveillance system requires more than just more sequencing; it requires a deliberate effort to fill the geographic gaps that currently leave the world vulnerable to the next emerging threat. As Wei Zhou argues in the context of the Global Biobank Meta-analysis Initiative, ancestral and geographic diversity are not just ethical goals—they are the technical requirements for the next generation of precision medicine and public health.