After the Human Genome Project — From a Single Sequence to the Pangenome
Starting Point — A Common Map Was Needed
At the end of the 20th century, researchers knew some genes and markers in the human genome but lacked a common map to compare them at the same coordinates. The Human Genome Project (HGP), launched in 1990, aimed to systematically construct the sequence and gene map of human DNA. It was a large-scale collaborative effort in which multiple countries and research institutions jointly designed principles for technology, samples, and data sharing.
The declaration of completion in 2003 signified that the core objectives set at that time had been achieved. It did not mean that every base and structural variant of every individual had been fully read. Gaps remained in complex regions such as repetitive sequences and centromeres, and the primary reference was a linear sequence assembled from a limited number of individuals.
The First Change — The Emergence of Coordinates
The primary role of the reference sequence was not to provide “the correct answer for a single normal human,” but rather to establish a common coordinate system. Researchers can now map gene locations, variants, and assay targets onto the same map, enabling direct comparison of results. This foundation underpins modern analysis, including the alignment of sequencing reads to the reference and the recording of variants in the format chromosome–position–allele.
However, the reference allele is not guaranteed to be the most common or healthy allele in a population. It is also incorrect to treat all sequences differing from the reference as aberrant. Coordinate systems must be distinguished from biological standards.
Technology Expands the Question
Early large-scale sequencing was time- and cost-intensive, but subsequent technological advances have enabled faster interrogation of individual genomes and multiple populations. Short-read sequencing has facilitated large-scale variant discovery, while long-read and assembly technologies have enabled the resolution of repetitive sequences, structural variants, and haplotypes over longer distances.
Technological improvement has not implied automated interpretation. As more variants are identified, pathogenicity, phenotype, ancestry, and data governance must be addressed concurrently. On the map established by the HGP, new uncertainties have become more distinct.
T2T — The Challenge of Closing Gaps
The Telomere-to-Telomere Consortium has presented an assembly that substantially fills the complex regions remaining in existing references by combining long-read and complementary technologies. This has expanded research opportunities for regions including centromeres and segmental duplications. The term “complete” should be interpreted within the scope of the cell line, haplotype, and assembly version employed.
While a more complete assembly reduces sequence gaps, it does not alone resolve the issue of representing diversity in human populations.
Pangenome — From One Line to Multiple Paths
The Human Pangenome Reference Consortium aims to represent sequences and structures missed by a single linear reference by incorporating diverse haplotype assemblies. Utilizing graphs or multiple assemblies may reduce bias caused by the absence of specific population sequences in the reference during mapping and better capture complex variants.
However, pangenomes do not constitute a simple event of replacing one file with another. Coordinate systems, tool compatibility, variant normalization, data storage, and clinical reporting frameworks must all evolve accordingly. Provenance regarding which release and path were used becomes increasingly critical.
The Remaining Question of Representativeness
Including diverse ancestries is an issue of equity as well as technical accuracy. Groups that are underrepresented in reference and study cohorts may experience greater uncertainty in variant frequency, imputation, and risk prediction. Rather than simply increasing sample sizes, community engagement, consent, benefit sharing, and controlled data access must be co-designed.
Representativeness does not end with filling in a few ancestry labels. Intra-population diversity and social context should not be reduced to simple biological categories.
What is understood
The legacy of the HGP is not a completed book that provides all answers to disease, but rather infrastructure that enables the formulation of comparable questions. The T2T has reduced gaps in a single assembly, and the pangenome has opened directions for representing multiple haplotypes. These three stages do not discard each other; instead, they reveal and extend different limitations of the reference.
Current Revalidation and Limitations
A superior reference can improve mapping and variant discovery, but it does not automatically determine the disease causality of variants, clinical actionability, or individual treatment choices. Changes in assemblies and tool releases necessitate coordinate conversion and reanalysis. Research results must always include the reference name and version.
Conceptual Connections
The meaning of coordinates and the reference allele leads to the Reference genome, diverse assembly and structural variant observation to PacBio SMRT·HiFi, and population representation to the UK genomic medicine and Broad MPG concepts.