WUT researchers contribute to the 4D human nucleome atlas

The 4DN project has for years been developing and integrating complementary experimental and computational methods to study genome architecture ‘in space and time’. The aim of the 4D Nucleome (4DN) programme is to understand how DNA (the genome) is organised within the cell nucleus in 3D and how this organisation changes over time (the fourth dimension), and how this translates into genome function.

"We aim to understand the molecular mechanisms responsible for the structure and dynamics of chromatin in human cells. The key challenges addressed by 4DN and our publication include: comparing and integrating complementary 3D genome mapping methods, linking genome structure with function (transcription, DNA replication, enhancer-promoter regulation) and its positioning relative to nuclear components – lamina, nucleolus, speckles – capturing cell-to-cell variability in the dynamic and heterogeneous organisation of the genome, and using these maps to predict the impact of genetic variants in a medical context,” says Prof. Dariusz Plewczyński from the Faculty of Mathematics and Information Science at WUT.

In the published study, the authors combined multiple types of 3D/4D genomics and imaging data to describe genome organisation in two reference human cell types: H1 embryonic stem cells and immortalised HFFc6 fibroblasts.

"We generated and integrated diverse genomic and biological imaging datasets, each contributing unique observations, which enabled us to compile extensive catalogues including over 140,000 chromatin loops per cell type, produce detailed classifications and annotations of genomic domain types and their positions within nuclear space, and obtain three-dimensional models of single cells showing the nuclear environment of all genes, including their long-range interactions with distant regulatory elements," explains Prof. Dariusz Plewczyński.

The publication also includes a broad benchmarking of methods for measuring genome architecture providing practical guidance for selecting techniques) as well as demonstrations of computational tools – including predictive models – allowing the prediction of genome folding based on DNA sequence and assessing the potential impact of genetic variants on genome structure and function.

The results of the project will be particularly useful in biotechnology, biomedicine, pharmacy, the design of innovative drugs and therapies, as well as in the development of new diagnostic approaches. Above all, they will indirectly accelerate research on health and disease – the U.S. National Institutes of Health (NIH) emphasise that changes in DNA organisation within the nucleus may be linked to diseases (e.g. cancer) and responses to infectious agents, and that a better understanding of nuclear organisation is essential for health.

The most direct beneficiaries of the results are research teams in molecular biology and genomics, which require reference maps and comparative standards, as well as the bioinformatics and data science sectors in life sciences, due to the publicly available data, tools, and protocols that build an open science ecosystem. In the longer term, depending on further research and translation, the beneficiaries will also include biotechnology, pharmaceutical industries, and precision medicine, wherever the interpretation of genetic variants – especially non-coding ones – requires understanding which regulatory elements interact with which genes in 3D.

"Our publication also points to the direction of building models that allow inference of the potential impact of sequence variants on genome folding, which in turn translates into better interpretation of genomics in medicine. An additional ‘public benefit’ is open access to data and standardised resources, which accelerates scientific progress – there is no need to repeat the same costly experiments in multiple locations," explains Prof. Plewczyński.

Tasks of WUT researchers: analysis, integration, modelling

The contribution of the team from WUT and CeNT UW focused on computational analyses, data integration, and 3D modelling: benchmarking methods, integrative inference of genomic domain localisation relative to nuclear structures, and co-developing analyses and 3D genome models.

The work of Prof. Dariusz Plewczyński, Mateusz Chiliński, PhD, and Kaustav Sengupta, PhD, supported, among others, the benchmarking of 3D/4D genome mapping methods and the development of integrative approaches that enabled the transition from multiple separate measurements to a coherent, multi-scale picture of genome organisation. All data and resources were made publicly available through the 4DN portal, facilitating their reuse in both basic and applied research – including studies on gene regulation mechanisms and the interpretation of disease-related variants.

The most demanding aspect of the work (particularly for computational teams, including SFGLab) involved the integration of heterogeneous data (different technologies, resolutions, interaction ranges, artefacts, and normalisation), as well as ensuring comparability – hence the strong emphasis on benchmarking, integrative and 3D modelling, and maintaining quality and reproducibility (pipelines, metadata, data sharing).

This is the second phase of the 4D Nucleome project funded by the U.S. National Institutes of Health (NIH). The first phase began in 2015 and focused strongly on the development and benchmarking of methods, as well as computational and modelling approaches. The second phase of the NIH Common Fund 4DN, launched in 2020, places greater emphasis on health and disease, as well as on further development of data integration and visualisation tools.

The article can be accessed here.