Donor-specific genomes reveal 1,790 chromatin hotspots

Lingbin Ni and co-authors said in a bioRxiv preprint posted Oct. 11 that donor-specific genome assemblies recovered 2.8% more usable Hi-C contact pairs than standard references.

Evidence grade: Early (preprint; not peer reviewed). In the Oct. 11 preprint, the team analyzed 354 human and 12 nonhuman ape haplotypes and found 1,790 recurring hotspots where 3D genome contacts differed.

What the study found

The authors studied how structural variants, which are large changes in DNA sequence or arrangement, relate to chromatin organization. Chromatin is DNA packaged with proteins, and its three-dimensional folding can influence which genes are accessible to regulatory machinery.

The study used haplotype-resolved chromatin interaction maps, meaning the researchers tried to separate signals from inherited chromosome copies. They paired those maps with matched genome assemblies and long-read transcriptomes, which track RNA output using longer sequencing reads.

Hi-C is a laboratory method that estimates which DNA regions sit near one another inside cells. When the authors used donor-specific assemblies rather than a standard reference genome, they said 13.2% of same-chromosome Hi-C links moved by more than 25 kilobase pairs in estimated genomic distance. One kilobase pair is 1,000 DNA base pairs.

The team also reported 2,495 structural variants associated with chromatin structure. Those variants were enriched in regulatory genomic contexts and were linked to variation in gene expression, according to the preprint.

Why it matters

Most genome analyses compare sequencing data with a reference genome. That can simplify analysis, but it can also obscure DNA segments that differ substantially across individuals or populations. A donor-specific assembly is built from an individual donor’s own genome data, rather than relying only on one standard template.

The preprint suggests that some differences in 3D genome maps may come from the way genomes are assembled and aligned, not only from biology captured cleanly by a reference genome. For researchers, that matters because chromatin contacts are often used to connect noncoding DNA regions with possible effects on gene regulation.

The authors said the contact changes were selective, not spread evenly across the genome. They concentrated in 1,790 hotspots associated with a restricted subset of structural variants. That pattern matters because it points researchers toward specific regions for follow-up rather than implying that all structural variation has the same effect on genome folding.

The team also said more than 90% of the chromatin-structure-associated structural variants were derived in the human lineage. The authors interpreted that as consistent with recent remodeling of regulatory architecture at the population level.

Limits and context

The finding is not clinical evidence. It does not show that any one variant causes a disease, predicts a health outcome or should be used in medical decision-making. For ordinary readers, the study is best understood as a methods and biology result, not as a test, treatment or prevention finding.

The evidence is also early because the paper is a bioRxiv preprint and has not been certified by peer review. The supplied abstract does not give details on cell types, experimental replication, statistical thresholds or validation experiments beyond the headline results, so those elements would need review in the full paper and by outside experts.

The Human Pangenome Reference Consortium was among the author groups listed. Pangenome work aims to represent more human genetic diversity than a single reference genome can capture, which is central to the study’s argument that donor-specific assemblies can improve interpretation of 3D genome data.

The authors declared funding from the US National Human Genome Research Institute of the National Institutes of Health through four grants: U01HG010971, 1U41HG007497, R01HG002385 and R01HG010169. Evan Eichler disclosed serving on the scientific advisory board of Variant Bio; the other authors declared no competing interests.

For bioinformatics teams, the practical issue is whether donor-specific assembly pipelines become standard in 3D genomics workflows. The preprint’s reported gain of 2.8% more usable Hi-C pairings is modest in percentage terms, but the authors argue that it helps expose recurrent regulatory-architecture differences that standard references can miss.

Source: academic preprint, bioRxiv, Oct. 11, 2026