Research

Spatial Organisation of Chromatin and Dynamics of Transcription Factors

Transcription factors (TF) are proteins that bind to a target site on chromatin and orchestrate transcription. We try to model the interphase chromatin as a network where each bead of the simulated polymer acts as a node and the TF as an unbiased random walker, hopping to other nodes in the neighborhood. From existing literature on the network, it is known that the mean first passage time (MFPT) of a random walker between any two nodes on a general network follows a set of particular scaling laws.

  • Our main goal is to verify whether the MFPT scaling law (for a network) holds in the chromatin context, i.e., the possibility of characterizing the MFPT of a TF based on walk dimension and fractal dimension of chromatin mesh.
  • The biologically relevant question is, if a TF is stuck inside a wrong TAD (chromatin mesh with no target site), how long it would take to come out of it. In other words, we are trying to figure out the dependence of mean exit time from a TAD on the size of the domain (radius of gyration) and structure of this domain.
  • It is now well accepted that chromatin are dynamic polymers. This dynamic nature arises due to the tethering of chromatin to the nuclear lamina, the interplay of interactions with various proteins, and the dynamic loop extrusion process via cohesin.Previous works show that the performance of a random walker (walking on polymer) is increased if the polymer has its dynamics. The mean square displacement of a random walker along a static polymer shows only diffusive behavior, whereas polymer dynamics can make the walk superdiffusive. It will be interesting to see whether chromatin dynamics can reduce the TF's mean target search time (or mean exit time from a TAD).
  • exit times

  • Effect of collapsed structure: Chromatin is a collapsed polymer and it makes frequent contact with itself. TF can jump to distant loci if there are contacts between far-away loci. We constructed an artificial domain where each locus is making contacts with other loci with a constant probability. Mean exit time from the domain for a random walker shows a non-monotonic behavior. If the contact probability is small, the walk is only along the 1D chain and can not hop to distant loci. Again, if contact probability is very high, there are too many contacts for each locus and the walker takes more time to reach the domain boundary.
  • Exit time analysis: We constructed an ensemble of connectivity matrices (Cij) (with entries 1 and 0) for a region on human chromosome 7 according to the available Hi-C contact matrix. A random walker was allowed to walk on the network following the connectivity matrix (i.e. only if Cij=1, jump is allowed between i and j ). The chosen region consists a TAD and we calculated exit time distribution. Simulation have shown that total mean exit time from a bigger domain is less than sum of mean exit times from the sub-domains.
  • Exit time scaling: We collected more TAD structures from the Hi-C map in order to find out a proper scaling of mean exit times with TAD lengths.
  • Finite-size of the TAD: As discussed earlier, Our goal is to calculate mean exit time scaling with the size of the domain. However, the actual TAD size varies in the range 0.5Mbp to 2Mbp. Hence, getting a scaling law remains a challenge with the available data.

  • Investigating the role of Lamin interactions in regulating chromatin organisation

    Lamins are principal nucleoskeletal proteins of all mammalian cells, which impart rigidity to nuclear morphology and ensure the resilience of the nucleus by forming filamentous meshwork under the inner nuclear membrane, called nuclear lamina. Over the recent years, mutations in lamins have been found to cause a cluster of human diseases, collectively termed laminopathies (e.g., dilated cardiomyopathy (DCM), Hutchinson–Gilford progeria syndrome etc.). One of the significant features of such mutations is the abrupt disappearance of the nuclear lamina and the formation of multiple lamin aggregates in the nucleoplasm. In this study, we perform coarse-grained simulations (using LAMMPS) of a small region near the membrane of the nucleus to investigate the role of interactions between different nuclear components (chromatin, lamin and nuclear membrane) in WT cells and possible changes in those interactions which can lead to a mutant configurations.

    recombination

    Improved chromatin models using systematic coarse-graining data

    Coarse-grained chromatin models are essential to interpret various experimental observations and predicting physical principles behind chromatin organization. Most existing polymer models consider chromatin as an array of beads connected by springs. Each bead represents chromatin segments of genomic length 1 kbp to 1 Mbp (depending on the context). Models also assume various polymer properties as input, such as the spring constant of the connecting spring between beads, bending rigidity, coarse-grained bead sizes, equilibrium bond lengths between neighboring beads, equilibrium angles between three consecutive beads etc. In a recent study, using extensive simulations, we have predicted these quantities essential for the polymer representation of chromatin.

    recombination
    Modeling chromatin as a FRC polymer Using the distribution of bond angles and bond lengths from systematic coarse-graining, we ask whether chromatin-like polymer config- urations can be generated by incorporating these coarse-grained parameters into simple polymer models. We construct a freely rotating chain (FRC) polymer using these parameters. We show that while it is possible to generate TAD-like domains as seen from the FRC model, and also the average 3D distance between the same genomic loci matches with experimentally observed values, there is a significant deviation of the scaling of the radius of gyration with the genomic length for this FRC polymer from experimentally observed data. This disagreement indicates that “only” heterogeneous bond lengths and angles are not sufficient to model chromatin segments, and we must construct polymer models incorporating the effects of soft interaction potential and sequence heterogeneity


    Past Research Experience


    Summer Internship
    SINP Kolkata (May 2018-July 2018)

    Course Projects

    PH 549:Physics of Biological Systems
    IIT Bombay (2020)


    PH 543:Advanced Statistical Mechanics
    IIT Bombay (2020)