In biological systems, the information encoded within the DNA duplex is not a mere linear script but a complex network of interactions where each strand influences and coordinates with its complement. This physical and functional coupling is essential, orchestrating genomic processes such as transcription regulation, replication, and DNA repair. Hence, the ability to model these cross-strand relationships dynamically offers a powerful and nuanced way to decode genomic function with unprecedented fidelity. CrossDNA takes a bold step forward by explicitly modeling these relationships rather than relying on implicit or static approximations.

The architecture of CrossDNA is notably distinctive. It employs a dual-branch framework in which the model alternates between processing forward and reverse-complement segments of DNA sequences. This design emulates the natural duplex structure of DNA, providing the model with both “views” of the genomic code and forcing it to learn the interplay across strands actively. Moreover, CrossDNA facilitates explicit interstrand communication through a lightweight but highly effective cross-strand communication module. This feature enables real-time information sharing between the forward and reverse branches during the learning process, ensuring that context-dependent interactions are captured dynamically rather than treating strands as isolated entities.

A significant technical challenge when working with genomic sequences is accommodating their length and contextual dependencies. Genomic regulatory elements can span thousands of base pairs and require models to attend to vast, complex sequence contexts. To address this, the developers of CrossDNA ingeniously combined a recurrent long-context backbone with sliding-window attention mechanisms. This hybrid approach allows the model to maintain an extensive memory of the sequence context while efficiently focusing on local relevant regions. As a result, CrossDNA achieves a new level of long-range genomic understanding that surpasses existing models’ capabilities.

The performance gains offered by CrossDNA are not merely theoretical. When benchmarked across a variety of genomics prediction tasks—including enhancer element identification, transcription factor binding prediction, and non-coding variant prioritization—the model consistently outperformed traditional DNA language models. Particularly notable is its performance on enhancer prediction, where the ability to model cross-strand interactions directly correlates with improved robustness to sequence orientation changes. These findings underscore the functional relevance of explicitly modeling DNA as a duplex rather than as a one-dimensional sequence or a simplistic double-complement symmetry.

One of the most striking aspects of CrossDNA is its parameter efficiency. While many state-of-the-art DNA foundation models contain hundreds of millions of parameters, CrossDNA achieves comparable—and often superior—predictive performance with only a fraction of the parameter count, in the million-parameter scale. This streamlined design not only accelerates training and inference but also enhances the model’s accessibility for broader scientific use, especially where computational resources might be constrained. It represents a paradigm shift towards more biologically-grounded and computationally sustainable genomic AI models.

Beyond improving model metrics, CrossDNA opens new avenues for interpretation and discovery within genomics. By capturing explicit dynamic cross-strand interactions, it provides a framework to better understand the regulatory logic underlying gene expression and chromatin organization. This could lead to the identification of novel regulatory elements that have been elusive to previous models and experimental assays. Additionally, CrossDNA’s capability to prioritize disease-associated non-coding variants promises to accelerate the interpretation of human genetic variation, facilitating advances in personalized medicine and genomic diagnostics.

The design principles behind CrossDNA also highlight the importance of mimicking biological reality within computational models. Many earlier efforts tried to impose reverse-complement symmetry or strand equivalence through data augmentation or static equivariant transformations. While useful, these approaches inevitably gloss over the dynamic, context-dependent nuance of real DNA strand interactions. CrossDNA’s approach to explicitly and iteratively learning cross-strand dependencies reflects an important conceptual leap, treating DNA as a fundamentally duplex molecular entity rather than as two separate strands.

In terms of technical implementation, CrossDNA’s cross-strand communication module is a lightweight yet powerful component that acts as a bridge transmitting information between the dual branches. This module dynamically integrates contextual signals during training, allowing each branch to incorporate what the other learns in a way that mirrors physical strand interactions. The synergy derived from this interbranch communication is essential for the model’s superior performance and ability to understand complex genomic structures.

The recurrent long-context architecture embedded into CrossDNA deserves special mention as well. Long-range dependencies in DNA sequences pose profound challenges due to the sheer length and complexity of genetic material. The combination of a recurrent backbone with sliding-window attention ensures that the model can both hold onto historical context and prioritize immediate, biologically relevant sequence patterns. This architecture mitigates the memory bottlenecks and computational inefficiencies that typically plague large sequence models, charting a path forward in genomic deep learning.

Perhaps most excitingly, CrossDNA transforms the concept of DNA language modeling into a more faithful analog of biological reality. By explicitly modeling cross-strand interactions dynamically, it transcends previous approximations that reduced the duplex DNA to unidirectional strings or symmetrical pairs. This leap forwards will not only enhance computational genomics but may also deepen our fundamental understanding of DNA’s role as an information carrier within the cell.

In practical terms, the advent of CrossDNA paves the way for more reliable and interpretable genomic prediction tools. Researchers investigating regulatory element functions, epigenetic markers, and mutation impacts will benefit from this improved modeling fidelity. Clinical geneticists tasked with identifying pathogenic variants in non-coding regions—an area historically challenging due to data complexity—now have a powerful computational ally that integrates contextual nuances from both DNA strands simultaneously.

Looking ahead, the interdisciplinary team behind CrossDNA has laid a foundation that could extend beyond human genomics. The principles underpinning cross-strand modeling may find applications in broader biological sequence analysis, including RNA duplexes, protein-DNA interactions, and even synthetic biology. This creates exciting possibilities for AI-driven innovation rooted in biomolecular structure and function.

Moreover, CrossDNA exemplifies a successful marriage of biological insight with AI techniques, showcasing how domain expertise can guide architectural decisions for transformative results. The model’s ability to efficiently leverage cross-strand information without ballooning parameter counts sets a precedent for future genomics models that balance complexity and interpretability with resource constraints.

In sum, CrossDNA represents a paradigm shift in the functional interpretation of genomic sequences by embracing the inherently duplex nature of DNA. Its explicit, dynamic modeling of cross-strand interactions, combined with innovative architecture for long-context handling and parameter efficiency, establishes new benchmarks in the field. This breakthrough has profound implications for genetics, molecular biology, and precision medicine, positioning CrossDNA as a pioneering tool in the new age of genome interpretation fueled by artificial intelligence.

Subject of Research:
Innovative DNA sequence language modeling focusing on explicit, dynamic cross-strand interactions within the DNA duplex to enhance genomic function interpretation and prediction accuracy.

Article Title:
Explicit dynamic cross-strand interactions for DNA sequence language modelling.

Article References:
Yang, C., Liu, Y., Ling, L. et al. Explicit dynamic cross-strand interactions for DNA sequence language modelling. Nat Mach Intell (2026). https://doi.org/10.1038/s42256-026-01249-1

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s42256-026-01249-1

The focus of this investigation was on the protein PAF15 and its influence on PCNA, a pivotal player in DNA replication. PCNA acts as a sliding clamp, coordinating the activities of DNA polymerases at the replication fork. Previous work established PAF15’s role in stabilizing PCNA on chromatin; however, its capacity to modulate PCNA behavior under conditions of overexpression or loss was not fully understood. This latest work reveals that while PAF15 is essential for normal replication processes, its overabundance can severely disrupt replisome dynamics, highlighting a previously unappreciated layer of regulation.

Using inducible overexpression systems in PAF15 knockout cells, the researchers found that short-term increases in PAF15 effectively stabilized chromatin-bound PCNA. However, this stabilization depends on PAF15’s PIP-box motif, a key domain facilitating direct PCNA binding. Mutations in this motif abrogated the stabilization effect, underscoring the specificity of the interaction. Moreover, the N-terminal basic region of PAF15, reminiscent of histone H3, was shown to be essential for PAF15 stability and its capacity to retain PCNA on chromatin, indicating cooperative functional domains within PAF15 that maintain replication fork integrity.

Remarkably, though initial PCNA stabilization appeared protective against replication catastrophe in the short term, sustained overexpression of PAF15 led to profound replication impediments and eventual cell death. This toxicity was linked specifically to the PAF15-PCNA interaction since PAF15 variants lacking either a functional PIP motif or a stable N-terminal domain did not elicit these harmful effects. This finding suggests that the cellular system naturally restricts PAF15 levels to prevent interference with the intricate choreography of the replisome.

Further mechanistic insights were gained by examining the roles of leading and lagging strand polymerases under conditions of heightened PAF15. Despite enhanced PCNA retention on chromatin, levels of the lagging strand polymerase POLδ1 decreased markedly, whereas the leading strand polymerase POLε1 remained stably bound. This imbalance hints at disrupted coordination between the two strands, manifested as accumulation of poly(ADP-ribose) (PAR) chains indicative of replication stress and aberrant Okazaki fragment processing. Notably, excessive PAF15 appeared to hinder PCNA’s interaction with POLε, thereby slowing replication fork progression and culminating in genomic instability.

The molecular underpinnings of this phenomenon were illuminated through structural analyses. POLε forms a stable trimeric contact with PCNA by occupying all three PIP-box binding sites on the PCNA ring, effectively saturating the trimer and occluding binding by other factors such as PAF15. Structural alignment demonstrated steric clashes between PCNA bound simultaneously to PAF15 and POLε1, offering a clear mechanistic explanation for how PAF15 overexpression disrupts replisome function and replication fork progression.

To understand how normal cells prevent such detrimental PAF15 access to leading strand PCNA, attention turned to replisome components Timeless and Claspin. These proteins, part of the replication progression complex (RPC), link POLε to the CMG helicase and are critical for fork stability. The study revealed that Timeless and Claspin act as protective elements, shielding leading-strand PCNA from promiscuous PAF15 binding. Loss of Timeless leads to mislocalization of PAF15 onto POLε1, exacerbating fork slowdown and triggering cell cycle arrest and death. Strikingly, concurrent depletion of PAF15 rescues these phenotypes, confirming that inappropriate PAF15 binding is the toxic event.

Timeless–Claspin mediated protection is modulated further by PCNA itself, as shown by the reversal of increased Timeless–Claspin chromatin binding in PAF15-deficient cells upon treatment with the PCNA inhibitor T2AA. This suggests a feedback mechanism where PCNA-bound complexes adjust dynamically in response to PAF15 levels, maintaining replication fidelity. Interestingly, human Claspin also harbors a PIP motif, hinting at additional layers of PCNA regulation that remain to be fully characterized.

The findings also support a model whereby Timeless–Claspin transiently shield PCNA during its loading on the leading strand, preventing inappropriate factor access. The authors propose that temporary displacement of Timeless and Claspin under oxidative stress or replication impediments may expose PCNA sites to PAF15, enabling adaptive replisome remodeling during genome surveillance processes such as topological stress response and DNA damage repair.

This pioneering study fundamentally advances the concept that maintaining an optimal and restricted pool of PAF15 is essential for balanced PCNA dynamics and proper replication fork progression. Any disruption, especially PAF15 overabundance, leads to replication collapse and cell death, emphasizing the importance of tight regulatory control. Moreover, it positions Timeless and Claspin as critical guardians of replication strand specificity by limiting PAF15 access to leading strand PCNA.

Beyond unraveling the replisome architecture and its regulation, these findings have broad implications for understanding replication stress responses and genome instability commonly associated with cancer and aging. The delicate interplay between PAF15, PCNA, and replisome components represents a promising target for therapeutic intervention designed to modulate DNA replication in disease contexts.

Given the importance of PAF15 dosage control, the study also examined transcriptional pathways governing its expression. They identified a conserved regulatory mechanism mediated by E2F transcription factors, with E2F4 acting as a transcriptional repressor for PAF15. Loss of E2F4 led to increased PAF15 expression and replicative toxicity, further stressing the critical need for tight transcriptional control to avoid destabilizing replication forks.

The comprehensive integration of biochemical, cellular, structural, and genomic analyses established in this work provides a new framework to understand DNA replication dynamics with strand-specific nuances. It highlights how protein dosage and chromatin-based protective complexes coevolve to safeguard genomic integrity during cell proliferation.

In conclusion, the exhaustion model of PAF15–PCNA interaction offers a paradigm for strand-specific replication control, where regulated access and displacement ensure efficient DNA synthesis without eliciting catastrophic replication stress. This study not only deciphers the molecular determinants governing replication fork fidelity but also opens avenues for exploring therapeutic strategies aimed at replication vulnerabilities in cancer and other proliferative diseases.

Subject of Research:
PAF15’s role in regulating PCNA dynamics and strand-specific control of DNA replication.

Article Title:
PAF15–PCNA exhaustion governs the strand-specific control of DNA replication.

Article References:
Chhetri, G., Badugu, S.B., Petriman, N.A. et al. PAF15–PCNA exhaustion governs the strand-specific control of DNA replication. Nature (2026). https://doi.org/10.1038/s41586-025-10011-3

Image Credits:
AI Generated

DOI:
https://doi.org/10.1038/s41586-025-10011-3

Polyploidization—a state in which cells harbor more than two complete sets of chromosomes—has long been recognized as a critical driver of tumor evolution and heterogeneity. It introduces chromosomal instability and aneuploidy, conditions frequently observed in cancer, which correlate with increased therapy resistance and poor patient prognoses. Until now, the precise cellular routes leading to polyploid genomes and their consequences on genome integrity remained understudied.

Using a sophisticated approach that tracks DNA replication dynamics in single cells, researchers identified two mechanistically distinct routes leading to polyploidy: rereplication and endoreplication. Rereplication is characterized by the replication of already duplicated DNA within the same cell cycle, resulting in replication bubbles and genome amplification. Endoreplication, in contrast, involves multiple rounds of DNA synthesis without subsequent cell division, also termed endocycling or endoreduplication. These pathways were observed distinctly in U-2 OS cancer cells treated with pevonedistat, a NEDD8-activating enzyme inhibitor known to stabilize the DNA replication licensing factor CDT1.

Pevonedistat treatment induced significant nuclear enlargement and replication stress, hallmarks of polyploidy. Single-cell lineage analyses demonstrated continuous replication events consistent with both rereplication and endoreplication. Intriguingly, sister cells often exhibited asymmetric replication patterns, highlighting the heterogeneity even within genetically identical cellular lineages. This asymmetry potentially fuels intratumoral diversity by generating cells with different replication histories and genomic contents.

Further investigations revealed that the route by which polyploidy is induced has profound implications for genome integrity. Although cells that underwent rereplication and endoreplication achieved similar DNA ploidies by the end, rereplicating cells accumulated higher levels of DNA damage markers such as γH2AX, suggesting elevated genomic stress. This distinction was affirmed through sequential staining protocols and DAPI-based DNA quantifications, underscoring the differential genome stability consequences based on the underlying polyploidization mechanism.

To elucidate whether oncogenic signaling mimics these polyploidization routes, the team engineered cells to overexpress common cancer-associated oncogenes—H-RAS V12 and cyclin E1. Overexpression of these oncogenes reproduced replication stress phenotypes, including reduced replication fork velocity and increased formation of micronuclei. Importantly, single-cell tracking over multiple generations revealed elevated heterogeneity and polyploidization in these cells, recapitulating the dual pathways observed with pevonedistat treatment. These findings affirm that oncogenic signals can trigger both rereplication and endoreplication routes converging on polyploidy.

Assessing the timing of DNA damage exposure revealed nuanced controls over pathway choice. Cells irradiated in G2 phase exhibited a predilection for endoreplication, whereas G1 phase irradiation skewed replication aberrations toward rereplication in the subsequent S phase. Hence, the cell cycle stage during which DNA damage occurs biases cells toward distinct mechanisms of aberrant genome duplication, emphasizing the dynamic interplay between genotoxic stress and replication control.

To decipher the molecular underpinnings differentiating these polyploid populations, the authors performed fluorescence-activated cell sorting followed by single-cell RNA sequencing of normal and polyploid cells under pevonedistat treatment or HRAS overexpression. Polyploid cells consistently clustered apart from non-polyploid controls, exhibiting dysregulated expression of gene sets enriched for cell cycle regulation, chromosome segregation, DNA replication, and DNA damage response pathways.

Focused subcluster analyses identified a core network of genes strongly associated with mitotic progression and genome maintenance, including key regulators such as CDK1, cyclin A, aurora kinase B (AURKB), and topoisomerase II alpha (TOP2A). This gene signature appears central to orchestrating the balance between rereplication and endoreplication, potentially dictating cellular decisions to undergo one pathway over the other.

Protein-level validation by quantitative imaging confirmed heterogeneity in nuclear cyclin A levels among polyploid cells, aligning with gene expression data. Pharmacological inhibition experiments further supported these functional insights: inhibition of TOP2A preferentially fostered rereplication, whereas CDK1 inhibition primarily induced endoreplication phenotypes. These findings illuminate how targeted interference with key cell cycle regulators can modulate the mode of polyploidization.

Moreover, chromatin binding analysis of replicative helicase components (MCM2, MCM4, and MCM7) revealed distinct loading patterns dependent on the kind of induced replication stress, reflecting the mechanistic divergence underlying rereplication and endoreplication. Such regulation of helicase access may serve as a critical checkpoint to prevent genome overduplication or ensure faithful replication cycles.

These landmark insights articulate a compelling paradigm in which cancer cells exploit multiple routes to increase genomic content, with significant consequences for genome stability and therapeutic resistance. Understanding the differential pathways to polyploidy adds a crucial layer to our knowledge of tumor evolution dynamics and may unearth exploitable vulnerabilities for novel anticancer strategies.

Future research will need to explore how these polyploid states influence tumor aggressiveness and how tumor microenvironmental factors might steer cells towards one polyploidization route or another. Ultimately, targeting the molecular circuits that govern rereplication and endoreplication may offer avenues to curtail cancer cell adaptability, potentially enhancing the efficacy of existing treatments.

By integrating advanced live-cell imaging, genetic manipulation, and transcriptomic profiling, this study sets a new standard for dissecting complex cellular behaviors at single-cell resolution across generations. The revelation of two distinct yet interrelated pathways to oncogenic polyploidy heralds a new understanding of genome instability in cancer and opens the door to innovative diagnostic and therapeutic approaches.

Subject of Research: Mechanisms of polyploidization and genome instability in cancer cells through DNA replication dynamics and oncogene overexpression.

Article Title: Multigenerational cell tracking of DNA replication and heritable DNA damage.

Article References:
Panagopoulos, A., Stout, M., Kilic, S. et al. Multigenerational cell tracking of DNA replication and heritable DNA damage. Nature (2025). https://doi.org/10.1038/s41586-025-08986-0

Image Credits: AI Generated