Assessing the limits of restraint-based 3D modeling of genomes and genomic domains

Abstract

Restraint-based modeling of genomes has been recently explored with the advent of Chromosome Conformation Capture (3C-based) experiments. We previously developed a reconstruction method to resolve the 3D architecture of both prokaryotic and eukaryotic genomes using 3C-based data. These models were congruent with fluorescent imaging validation. However, the limits of such methods have not systematically been assessed. Here we propose the first evaluation of a mean-field restraint-based reconstruction of genomes by considering diverse chromosome architectures and different levels of data noise and structural variability. The results show that: first, current scoring functions for 3D reconstruction correlate with the accuracy of the models; second, reconstructed models are robust to noise but sensitive to structural variability; third, the local structure organization of genomes, such as Topologically Associating Domains, results in more accurate models; fourth, to a certain extent, the models capture the intrinsic structural variability in the input matrices and fifth, the accuracy of the models can be a priori predicted by analyzing the properties of the interaction matrices. In summary, our work provides a systematic analysis of the limitations of a mean-field restrain-based method, which could be taken into consideration in further development of methods as well as their applications.

Defined chromosome structure in the genome-reduced bacterium Mycoplasma pneumoniae

“…Second, TADbit used a Monte Carlo simulated annealing sampling procedure to identify 3D models that best satisfy all of the imposed restraints. The contact map obtained from the final models resulted in a Pearson correlation of 0.83 with the input Hi-C interaction matrix, which is indicative of good model accuracy47.…”

“…To build the 3D models, we applied a restraint-based modelling approach using the TADbit python library164748. The genome was defined by 82 particles, determined by the resolution of the contact map at 10 kb.…”

An integrated 3-Dimensional Genome Modeling Engine for data-driven simulation of spatial genome organization

“…This can be inferred from the contact data itself or determined using a separate data set. In particular, there are now several software packages designed to identify topologically associating domains and loop regions from Hi-C data sets (for review, see Trussart et al 2015), and these structures can be used in 3D-GNOME in the same manner as CCDs.…”

“…However, applying well-established ISD techniques to 3D genome mapping data is not straightforward, as the precise relationship between IFs and physical distances is unknown and may differ according to specific chromosomal segments and locations in the nuclear space. Nevertheless, recent results suggest that chromosome structure prediction is robust with respect to experimental noise but more sensitive to different structural-functional relationships (Trussart et al 2015).…”

On the demultiplexing of chromosome capture conformation data

“…They are thus well poised to handle the heterogeneity properties of chromatin in vivo. The possibility of properly quantifying the structural variability of chromosomes nevertheless remains an important challenge [21].…”