Telomeres have long been measured as an average length across all chromosomes, but this approach hides important differences: in humans, the ends 17p, 20q, and 12p are consistently shorter than the rest, and they are the ones linked to genome instability. Neighboring works in our archive also look for hidden heterogeneity: gene maps predict breast cancer outcomes from individual profiles, and methylation analysis shows how each cell ages differently. In this series, the present paper stands apart because it transfers the idea of individuality to the level of individual chromosome ends. Previously, such analysis required labor-intensive experiments with limited coverage, but now long reads and complete genome assemblies allow it to be done on a mass scale.
The authors developed TeloXplorer, a modular program that works with long-read sequencing data (Oxford Nanopore and PacBio). First, it finds reads with telomeric repeats, then determines the boundary between telomere and subtelomere for each chromosome end, estimates telomere length, deciphers variant repeats (TVRs), and, if modification data are available, profiles cytosine methylation. The tool can operate in three modes: with the genome assembly only, with reads only, or with both. To test accuracy, reads were simulated with different coverage depths and error rates, and then the results were compared with four existing methods: TeloBP, Topsicle, Telometer, and Telogator2. On real data, they used the HG002 cell line with a fully phased assembly, family trios, and 232 samples from the second release of the human pangenome (HPRC2).
TeloXplorer accurately determined telomere boundaries in humans, Arabidopsis, and yeast. On simulated data, it showed high precision and recall in identifying telomeric reads, outperforming competitors on non-human genomes. For HG002, telomere lengths ranged from 0.66 to 9.61 thousand base pairs depending on the chromosome end and parental haplotype. The main discovery: each chromosome end has a characteristic profile of three parameters: length, variant repeat diversity, and subtelomere methylation level. These profiles are stable across populations but do not correlate with each other, meaning they are governed by different mechanisms. Family trios showed that variant repeats are inherited in a Mendelian fashion, and methylation is tied to TAR1 repeats. Analysis of 232 people revealed that African populations have higher TVR diversity than non-African ones, consistent with a founder effect upon leaving Africa.
The work differs from its neighbors in the archive by combining three dimensions of telomeres in one tool and applying it to hundreds of people rather than individual samples. If in the cell aging project methylation is considered a marker of age, here it becomes a local feature of a specific chromosome end, linked to TAR1 repeats. This is not just a new method but a paradigm shift: telomeres can no longer be described by a single average length. The results show that the inherited architecture of chromosome ends sets the framework within which changes during aging and disease occur. For aging and cancer researchers, this means the need to move from aggregate measurements to per-end ones.
The next step is to apply TeloXplorer to large cohorts with known age and clinical data to test which chromosome ends predict disease risk. So far, the HPRC2 analysis is limited by the lack of donor age and possible telomere drift in cell culture. Experiments with targeted deletion or insertion of TAR1 will be needed to prove a causal link between these repeats and methylation or TVR diversity. It is also important to expand the sample beyond the five continental groups, especially for rare haplotypes, such as the variant on 11p found only in East Asian populations.
The tool will be useful in medical genetics for finding links between the length of specific chromosomes' telomeres and predisposition to cancer or age-related diseases. In evolutionary biology, it will allow tracking the history of telomere rearrangements between populations and species.
Validation on primary tissues, not just cell lines, and analysis of large pedigrees to confirm telomere length inheritance. In parallel, TeloXplorer should be integrated with TERRA transcription data to link telomere structure with their activity.
For telomere biology, a long-standing mystery is why some chromosome ends shorten faster than others and how this relates to cellular aging. The work shows that the answer lies not in average length but in local architecture: variant repeats and subtelomere methylation create a unique context for each end. This connects the unresolved question of the mechanisms of selective shortening with specific genomic elements, such as TAR1, and opens the way to testing causal hypotheses.
Human telomeres contain TTAGGG repeats, but in Arabidopsis they are slightly longer: TTTAGGG. And in yeast, the repeats are so irregular that they are described as TG₁₋₃, that is, from one to three guanines in a row.
In Douglas Adams's novel The Hitchhiker's Guide to the Galaxy, the characters search for the question to the answer "42". Telomeres also store encoded history, but instead of the number 42, they contain repeat variants that can only be read with long reads and the right tool.
Not published on arXiv. Here are its page, the PDF and the formalization: you can check it yourself, not only read our account.