Urmi Majumder


I received my PhD on Biomolecular Computing and Self Assembly of DNA nanostructures from the Department of Computer Science, Duke University in Spring 2009. My thesis advisors were Professor John Reif and Professor Thom LaBean   Before Duke I was in India where I received my Bachelors of Technology degree from the Department of Computer Science and Engineering, Indian Institute of Technology, Kharagpur, in Spring 2004. Even though I am an American citizen now, I'm proud of my Bengali heritage and cherish the fond memories of growing up in Kolkata, India. 


I am currently a team lead with experience in both product development and project management in the E-Business Suite Supply Chain Management Division at Oracle Inc in Reston, Virginia (right outside the nation's capital, Washington DC). Recently (FY2012), I was awarded the 'Extra Mile Award' for my exceptional performance in the Supply Chain Management Group at Oracle.


It has never been easier to reach people. You can reach me via LinkedIn or email at:  urmim at urmim dot org.



Research @ Duke:


My graduate school research involved understanding biomolecular computation using a synthetic approach. Specifically, I spent an exciting 4.5 years at Duke investigating how to engineer a system of biomolecules to process information and carry out specific tasks. The notion of algorithms is central to such a question. But my research was not limited to simply applying algorithmic concepts to synthetic biology but also implementing them using real biomolecules. My work was well received by the biomolecular computation research community and I was awarded scholarships to present my work at GHC 2008, UC 2008, NSTI 2008, and DNA13, DNA15, FNANO conferences. I was also awarded the Outstanding Research Initiation Project Award in our department in 2006.  Additionally, apart from being awarded the departmental graduate fellowship (2004-2005) and research and teaching assistantships (2005-2009) for completing my doctoral studies in Computer Science, I was awarded the Graduate Nanoscience Program (GPNANO) Fellowship in Spring 2006 to complete a certificate program in Nanoscience (awarded May 2009 along with my PhD in Computer Science). 


Publication Summary:

U Majumder & A Rangnekar, K V. Gothelf,  J H Reif and T H LaBean, Design and Construction of Double-Decker Tile as a Route to  Three-Dimensional Periodic   Assembly of DNA, 

            Journal American Chemical Society (JACS), Vol. 133, no. 11, pp. 3843—3845 (Feb. 2011).

           U. Majumder, S. Sahu, T. LaBean, J. H. Reif, Design and Simulation of Self-Repairing DNA Lattices, DNA 12, LNCS 4287, pp: 195-214, Springer-Verlag, NY 2007.


Presentation Summary:

Oral Talks:


Towards Compact Robust DNA Self-Assembly based Computation: Modeling, Simulation and Experiments, PhD Forum, GHC, October 2008.


A Framework for Designing Novel Magnetic Tiles Capable of Complex Computation, Contributed Presentation, UC, August 2008.


Towards Compact Robust DNA Self-Assembly based Computation: Modeling, Simulation and Experiments, Masters Defense Presentation, Duke, October 2007.


Activatable Tiles for Compact, Robust Programmable Assembly and other Applications, Contributed Presentation, DNA 13, June 2007.


Poster Presentations:

Stochastic Analysis of Reversible Assembly, GHC Technical Poster Session, GHC 2008. 


Probabilistic Models for Damage and Self-Repair in DNA Self-Assembly, NSTI Nanotech 2008. 


Autonomous Reactivating Whiplash PCR for Programmable Molecular Computation, FNANO 08. 


Isothermal and autocatalytic Whiplash PCR machines for autonomous computing, 7th Annual Graduate Research Day, Duke, April 2008.


Activatable DNA Tiles for Compact, Error-Resilient Algorithmic Assembly, inDUKE Frontiers 2007. 


Activatable Tiles for Compact Error-Resilient Directional Assembly, FNANO 07. 


Reversible Self-Assembly of Squares as a Rapidly Mixing Markov Chain, FNANO 07. 


Design and Simulation of Self-Repairing DNA Lattices, FNANO 06. 


Design and Simulation of Self-Repairing DNA Lattices, 6th Annual Grad. Res. Day. Duke, 2006.


Self-assembly across scales, Research Initiation Project Presentation, Duke, May 2005.



Service:

Life at Duke went beyond just working towards my dissertation on biomolecular computation. Among others,


I used to coordinate Biomolecular Computation Journal Club, Duke CS (Fall 2006-Spring 2009). 

I was the President & co-founder of Duke Chapter of ACM-W, January 2008 - May 2009. 

I managed registration, finances, exhibits and other related responsibilities at FNANO Conference, Snowbird, Utah (2005 - 2008). •

I was also a Referee to the Reviewing Committee of ACM-JETC 2011,  DNA12,  Natural Computing,  BioSystems,  DISC08,  SODA09. 

I was a co-author in research grant: NSF EMT Grant CCF-0829798: EMT/NANO, 2008 - 2010. 

I also a member of  ACM (2006 - 2010), ACM-W (2008 - 2010), ISNSCE (2005 - 2010).



Dissertations:

Doctor of Philosophy Thesis, Fall 2007 - Fall 2009

(Committee: John Reif, Thom LaBean, Bruce Donald, Chris Dwyer) 

Molecular Computing with DNA self-assembly


Synthetic biology is an emerging technology field whose goal is to use biology as a substrate for engineering. Self-assembly is one of the several methods for fabricating such synthetic bio-systems. Self-assembly is a process where components spontaneously arrange themselves into organized aggregates by the selective affinity of substructures. DNA is an excellent candidate for making synthetic biological systems using self-assembly because of its modular structure and simple chemistry for assembly. My thesis work uses DNA as a nano-construction material as well. As a computer scientist, however, I have explored the computational power of DNA self-assembly too. In fact, computational DNA self-assembly has been shown to be theoretically capable of forming complex patterns. However, experimental demonstrations of these patterns have been plagued by significant assembly errors.  Personally I verified this fact with my first wet laboratory demonstration of a DNA self-assembled binary counter. Although several redundancy based error-correction techniques exist to address this problem, in practice they are quite difficult and expensive to implement. We addressed the problem of fault-tolerant computation with DNA self-assembly by realizing that theoretical models of self-assembly make an unrealistic assumption of directional growth and this assumption leads to a gap in theory and reality. Hence, we designed a new kind of self-assembling unit that we will hence forth refer to as a tile (composed of multiple DNA single strands and considered to be the basic unit of self-assembly) called an activatable tile that enforced the direction of assembly growth using two very common biochemistry techniques (branch migration and polymerization).  While implementing my DNA based binary counter without error-correction, I observed that the nano-structures are very fragile and prone to mechanical and thermal damage. In order to study damage and self-repair of fragile DNA nano-structures, I proposed probabilistic models for damage and self-repair and though not validated with experiments, the simulation results can be used to understand the behavior of the system under external impact. We also proposed a compact tile set that can self-heal given that the computation encoded in the tiles is reversible. Beyond fabrication, self-assembly can be used for actual computation as well. However, for the autonomous operation of such a machine we need to devise an autocatalytic system. In this context, simple biochemistry techniques like branch migration and polymerization allowed us to design an isothermal and autocatalytic machines that can run distinct programs in parallel in the same test-tube. Given that this machine can operate without outside intervention, it can have a variety of applications in a variety of environments from simple bio-sensing to complex medical diagnostic tools. As synthetic biology matures into a robust discipline, it is going to revolutionize both the biotechnology as well as the pharmaceutical industries. At that point, devices have to be mass produced and hence we need to understand the yields of our devices and their rate of production. We addressed the question of yield and time for production of large batches of self-assembled nano-devices by building a stochastic model for reversible assemblies that allowed us to estimate several thermodynamic properties of the assemblies. Till date, the DNA self-assembled nano-structures community have had tremendous success in building structures and devices in two dimensions. However, success in three dimensions has been very limited. Thus we designed a new DNA motif and experimentally demonstrated building of structures in two dimensions. If 3D assembly with this motif be successful, it would be an important milestone in synthetic biology. 


Masters of Science Thesis, Fall 2004 - Fall 2007

(Committee: John Reif, Thom LaBean, Bruce Donald, Chris Dwyer) 

Towards compact robust DNA self-assembly: modeling, simulation and experiments


Since our physical experiments showed that computational assemblies we developed error correction techniques for increasing robustness in computational assemblies. We also developed a stochastic model for computing equilibrium yields and convergence rates of erroneous assemblies and error rates at near equilibrium.


Bachelors of Technology Thesis, Spring 2003 - Spring 2004

(Supervisor, Sudeb Prasant Pal, IIT Kharagpur) 

Quality of Service for internet, multimedia and other real time systems


We developed an analytical model for the top (application) layer of the internet in order to deliver better latency and QoS for various internet services and evaluated its performance using simulation models. This also includes study of dynamic schemes for resource management of memory and bandwidth in the internet so that latency can be improved for the average case as well as the worst case. The issue of real time and multimedia traffic is one step ahead of that for networks. We incorporated parametric decompression into the existing model to improve the document retrieval latency for dynamic requests from huge databases.


Teaching:

Graduate Coursework:

Undergraduate Courses


Miscellaneous:


Despite crazy traffic in the beltway, DC metro area is a wonderful place to live in terms of weather, cultural activities, entertainment, shopping, festivals, outdoor recreation, and great restaurants. Weekend gateways in the mountains are just as easily accessible as the ones in the beach. Additionally, because of the amazing diversity of the people living in the area, good ethnic restaurants are not too hard to find. Above all, as the nation's capital, the impact of the government is very high in the region as is evident from the frequent political demonstrations and daily proceedings on Capitol Hill. Some favorite links from the area:


DCist              DC Foodies                We the people                Sustainable DC                WOD Trail                Wolf Trap Performances                Screen on the Green                Taste of DC