(2ed) Computational Methods to Engineer Proteins for Health and Environmental Applications | AIChE

(2ed) Computational Methods to Engineer Proteins for Health and Environmental Applications


Samanta, R. - Presenter, The University of Texas at Austin
Research Interests

Proteins are the building blocks of all cells in our body and living creatures. They are nanomolecular machines that perform many functions ranging from digestion of food, ultrasensitive detection of small molecules (olfactory receptors), and acting as a checkpoint that regulates the passage of ingredients and information through cells (membrane proteins) and catalyzing chemical reactions (enzymes), etc. Natural proteins consist of a sequence of amino acids that determine their function. A change in the protein sequence can affect its structure and, thereby, functions. Therefore, if we can understand how the functions are encoded well, proteins can be engineered to diagnose genetic disorders, arrest carcinogenic proliferations, develop stimulus-responsive drug systems, and develop new proteins for desirable functions. Furthermore, synthetic proteins can be designed through computer-aided designs and produced through directed evolution in laboratories. However, it is challenging to develop accurate physics-based models since proteins involve thousands of coupled interactions, complex environment, and dynamics across multiple lengths and time scales. Even though Machine learning techniques such as AlphaFold, and RosettaFold methods are highly accurate and help get around the complex physics, they face a roadblock where there is limited reliable data. Towards this, the broader goal of my research group will be to develop physics-based computational models over multiple time and length scales using all-atom, coarse-grained, and combining them with machine learning approaches to understand the mechanism and design of new proteins for therapeutic and environmental applications.

The wide range of themes that will form the basis of my research group are as follows:

  1. Understanding the fundamentals of dimerization of Receptor Tyrosine Kinase (dysregulation of which is a marker of dwarfism and cancer progression) and using the insights to develop therapeutics.
  2. Develop methods to understand the effect of polymer architecture on protein and polymer phase behavior. Further, develop an inverse design platform to predict the properties of proteins and polymers for the desired phase behavior.
  3. Develop methods to design lipid membrane captured self-assembled protein nano-assemblies.
  4. Rational design of enzymes for sustainable plastic degradation.

Based on the experiences I have gained as a graduate student and a postdoctoral fellow, I am prepared to tackle these challenges as an independent researcher. My group will develop physics-based and machine learning-based robust in-silico methods to study protein design and soft matter problems, using the principles of statistical thermodynamics and polymer physics. We will closely work with experimental collaborators to test and validate our prediction.

In my graduate research with Prof.Venkat Ganesan at University of Texas (UT) Austin, my primary interest was studying the phase behavior of self-assembled globular proteins and polymers and the effect of their physical properties on the complexes. This system is ubiquitously found in improving biosensors; mainly, the oppositely charged polymers are used to stabilize enzymes in biosensors which enhances their shelf life and accuracy; polycations tend to condensate DNA by forming aggregates with negatively charged serum proteins and tissues to engineer synthetic cells and artificial organelles and use of oppositely charged peptides to purify proteins. I developed a mean-field-based coarse-grained model in canonical and semi-grand canonical ensembles to understand the physics of proteins with the polymer solution. We quantified the influence of different physical properties like dielectric contrast between the proteins and polyelectrolytes,1 charge heterogeneity of the protein surfaces,2 charge modulation due to pH of the solution,3 and various other physical properties of proteins on the self-assembled structure of multiple proteins and polymer complex. As a postdoctoral fellow (with Prof. Jeffrey J Gray, Johns Hopkins University), my main research focus has been to develop a physics-based all-atom hybrid molecular dynamics and Monte Carlo based implicit model for membrane proteins. Although membrane proteins are essential for protecting the cells and keeping them healthy, drug targets and ion channels in membranes are critical to the nervous system. They constitute 30% of the total proteins and are targeted by 50% of pharmaceuticals in the market. However, they are challenging to analyze due to the lipid membranes, the diversity in their composition, and their complex features resulting in the scarcity of experimental data. Existing methods are time-consuming, built for individual applications, or rely on manual design strategies. Motivated by these issues, I am developing a biologically realistic in-silico implicit solvent-based approach to capture the electrostatic and hydrophobic interactions of membrane proteins within the lipid bilayer. Our model can predict the structural properties of single and multi-pass membrane proteins, determine stability, and determine the native structure and native sequence recovery to analyze its ability to design and docking applications.4-5 I am also working on applying machine learning techniques to predict protein lipidation sites.

Teaching Interest

I believe that as a teacher; my responsibility will be to challenge and support my students in pursuit of three learning objectives: (1) develop engineering problem-solving skills; (2) understand foundational science and engineering concepts; (3) cultivate an interest in taking classroom problem-solving skillset to real-world scenarios. These objectives are best achieved by connecting existing problems with ongoing research and reasoning them with a fundamental theory. I will promote an active learning environment to engage students in the classroom to foster better understanding. I have supplemented my research experience with a foundational study of pedagogy through a Teaching certificate program at the Johns Hopkins Teaching Academy and through teaching workshops at UT Austin. The teaching certificate was an extensive program that included participating in lectures on different learning techniques and a 3-day workshop on instructional material development and independent teaching. I was selected as a Hopkins Engineering and Research Tutorial (HEART) fellow to teach first-year undergraduate students an introductory course on advanced research topics. The fellowship allowed me to apply my teaching skills as an independent course instructor for the course "Polymer nanotechnology: emerging applications for composite materials." I used multiple active learning tools such as polls for multiple-choice questions, think-pair-share fostering discussion among peers, and exit ticket questions to get feedback about the class. A Teaching Institute observer for my class particularly appreciated me for integrating supporting video, polls and facilitating discussion and summarizing after each section. As a graduate student, I also served as a teaching assistant (TA) for undergraduate Transport Phenomenon and Thermodynamics. As a TA, I strove to develop an inclusive, safe, and interactive learning environment by engaging all students in understanding the subject's fundamental concepts rather than just solving the questions. In the future, I would prefer to teach courses such as Transport phenomena, Thermodynamics, Numerical Methods, and Fluid mechanics. However, I am flexible in teaching any courses in the Chemical Engineering curriculum. Additionally, I would be interested in offering more research-based courses such as molecular simulations, polymeric materials, and computational protein design, which can be tailored to both graduate and undergraduate classes.

In the classroom and my research group, I will try to foster a diverse, transparent, inclusive, accessible, and welcoming work environment where peers can learn from each other. I firmly believe that a diverse community increases the likelihood of novel approaches and applications of science and enriches the overall experience of all members. Working with a diverse group of colleagues and mentoring undergraduate and graduate students with diverse backgrounds has been very rewarding. Thus, beside graduate students, I would continue to involve undergraduate students in research programs through course projects, semester-long projects, or hosting summer research programs. Furthermore, I believe early intervention works best in encouraging high school students in STEM education. Towards such an interest, as a graduate student and postdoc, I am and have been involved in various activities such as organizing a computational biology workshop, introducing games such as FoldIT, and showcasing fun science experiments for high school students. The collective objective of such programs is to connect with underprivileged students early to level the playing field and foster a sense of belonging to STEM to improve equity and inclusion. I will continue to develop and participate in such activities in the future.

  1. R Samanta, Venkat Ganesan, Influence of dielectric inhomogeneities on the structure of charged nanoparticles in neutral polymer solutions; Soft Matter 14, 3748 (2018)
  2. R Samanta, Venkat Ganesan, Influence of protein charge patches on the structure of protein-polyelectrolyte complexes; Soft Matter, 9475-9488 (2018)
  3. R Samanta, Avani Halabe, Venkat Ganesan, Influence of Charge Regulation and Charge Heterogeneity on Complexation between Polyelectrolytes and Proteins; The Journal of Physical Chemistry B 124, 22, 4421– 4435 (2020)
  4. RF Alford, R Samanta, JJ Gray, Diverse Scientific Benchmarks for Implicit Membrane Energy Functions, Journal of Chemical Theory and Computation 17, 5248-5261 (2021)
  5. Koehler Leman, J., Lyskov, S., Lewis, S.M. et al. Ensuring scientific reproducibility in bio-macromolecular modeling via extensive, automated benchmarks. Nat Commun 12, 6947 (2021)