MRC Institute of Genetics & Molecular Medicine
MRC Institute of Genetics & Molecular Medicine


HiP-HoP: A new inter-disciplinary approach to predict complex 3D genome folding

Gilbert and colleagues describe a new method for prediction of the 3 dimensional folding of chromatin in a Molecular Cell paper: October 2018

Inside every cell DNA is wrapped up with proteins to form a structure called chromatin.  How chromatin is folded is important for gene regulation and controls how proteins are made in different cell types. Although chromatin folding in cells is complex, chromatin fibres behave much like simple polymers. To investigate the properties important for chromatin folding we setup a collaboration with polymer physicists, to model how chromatin folds at specific genes including those important for human disease.  We first collected information about how proteins bound to the DNA at these genes of interest and painted this information onto our computer based 3D polymer simulations. Then, using current knowledge of genome organisation, we added different physical properties to the polymer, such as regions with a more crumpled structure and allowed the computer simulated chromatin polymers to fold-up 100s of different times. We compared the outcome of these simulations to the physical 3D folding of chromatin inside real cells to test how well the model predicted the real 3D structure. The simulations showed us there was striking variability in the shape and folding of chromatin in individual cells, especially at active gene regions and this maybe important for regulating gene expression.  We named this method the “highly predictive heteromorphic polymer model” or HiP-HoP model. HiP-HoP now allows us to accurately predict 3D folding using commonly available 2D data and understand some of the fundamental principles that organise DNA inside cells. We are now applying this method to understand how chromatin folding changes in disease, and how this in turn affects the expression of individual genes.


Nick Gilbert research group