Sachin Goyal

Visiting Assistant Professor

Professional Biography

Sachin Goyal received his education at Banaras Hindu University (BHU), India where he was awarded his B.Tech. (1997) degree in Mechanical Engineering. Following three years of industrial experience at Larson and Toubro Limited, India, he attended the University of Michigan in 2000 to pursue M.S. and Ph.D. degrees in Mechanical Engineering and Scientific Computing. Sachin's research expertise lies in the areas of continuum mechanics, dynamics and controls and their application to engineering and biological systems. During his PhD, he contributed a nonlinear computational rod model to simulate the mechanics of filament-like structures. His research was first inspired by looping and tangling of underwater cables, but also paved the way to understanding the long-length scale mechanics of DNA. He received the 2006 Ivor K. McIvor Award for his outstanding research and scholarship in Applied Mechanics and the Rackham Pre-doctoral Fellowship, a prestigious graduate student fellowship at the University of Michigan. After Ph.D. he pursued his research with a merit-based postdoctoral scholarship at Woods Hole Oceanographic Institution (WHOI) which is a research center affiliated to MIT. He joined the Cornell faculty in 2007.

Research Interests

A simulation video [reference: Goyal et al., 2008]

Structural interactions of DNA and proteins:

Vital functions of biomolecules like DNA and proteins are strongly governed by their structural deformability. Our goal is to model their structural behavior and biophysics with continuum mechanics and multi-body dynamics-based approaches. This is an ambitious and scientifically-rich area of research which is perhaps best introduced by posing three fundamental questions:

How does the structural deformations of DNA control gene expression?
How does the base (monomer) sequence of DNA influence its structural deformations?
How do gene-regulating proteins manipulate structural deformations in DNA?

DNA is a filament-like structure that can be modeled as an elastic rod. We are developing a continuum mechanics-based rod model that efficiently captures multi-physical phenomenon including arbitrarily large (nonlinear) dynamic behavior, interactions with an aqueous environment, non-homogenous and non-isotropic behavior, electro-static fields, and mechanical self-contact or excluded volume effects. Some of the specific challenges in doing so and ideas of immediate interest are described in the next paragraph.

First, a significant challenge is to design experiments that test the validity of mechanical models of DNA. Ideally, one seeks to design experiments guided by model simulations that dovetail well with current experimental capabilities. Second, the material properties of DNA are not yet well-characterized by single-molecule experiments. In fact, the experimental evidence is sometimes contradictory and certainly subject to debate. Third, very little is understood about the mechanical deformability of the proteins that control structural changes in DNA. Interestingly, advanced computational tools spawned from mechanical engineering within the special field of flexible multi-body dynamics (FMBD) could be adapted to model proteins. This idea is yet to be exploited perhaps due to the apparent barrier to engineering experts faced with learning the field of molecular biology/ biochemistry. The current Molecular Dynamics (MD) software remains computationally inefficient (if not unusable) when applied to long time and length scale simulations of DNA-protein interactions. However, MD tools could be interfaced with FMBD tools using multi-scale modeling techniques to yield dramatic improvements in both computational speed and accuracy.

As we delve deeper into our research mission laid out by the above three fundamental questions, we are also inspired to pursue many follow-on studies. Examples of follow-on studies are:

1. Inverse modeling applied to the rod model for DNA: The inverse modeling approaches have emerged as promising ways of leveraging theoretical models with experimental measurements for understanding physical phenomenon. Inverse problems are often perceived as the estimation of model parameters (e.g. sequence-dependent stiffness for rod model of DNA) from experimentally observed data (e.g. the looping behavior of DNA). However inverse modeling approaches also allow us to perform sensitivity analyses and to reverse engineer the physical system.

2. Nucleosome modeling: The nucleosome is DNA wrapped around spool-shaped proteins called histones. This wrapping provides an organized means to compact these very long molecules (by five orders of magnitude) enabling them to fit within the small confines of the cell nucleus in human beings. The accurate modeling of this wrapping is of crucial medical interest and it is an active area of research. One challenge in nucleosome modeling lies in simulating DNA as a charged stiff rod in the electrostatic environment of the histone protein along with contact and disruptive thermal fluctuations. Our rod model, recently reformulated for studying DNA/histone interactions, has yielded some promising early results.