Good and hard

Howdy! It's been a rather busy week, highlighted with the submission of an NSF proposal a mere 15 minutes before it was due. Through it all, I've come to feel very average. It's a feeling that I don't much enjoy, quite frankly. I've done the absolute best that I could with the hand that I have been dealt but at times it is not enough. This fact has taught me new lessons about the universality of the Atonement. Because of Christ's grace, even after I have done all that I could and come up short, I can still have hope. Hope is what gives me the reason to smile despite it all and find joy in the journey.

Speaking of smiling, check on this cartoon from PhD comics here. It made me smile because it is so true to life.

In case any of you are wondering what I am hoping to do with myself during the next few years, I have pasted the body of my research proposal that I recently submitted.

A Mesoscale Model of RNA for Characterization of RNA Pseudoknots

Key Words: molecular dynamics, RNA, pseudoknot, folding

Ribonucleic acid (RNA), like its complement, DNA, is a fundamental component of cellular processes. While the role of RNA in biological systems is well-understood, little is known of its molecular morphology and dynamic behavior. I propose to develop a mesoscale model of RNA that will provide insight into the sequence/structure/function relationship of RNA tertiary structures.

The De Pablo group at the University of Wisconsin-Madison has established itself at the forefront of DNA simulation. With the development of coarse grain mesoscale DNA models and advanced simulation techniques, they were able to efficiently and accurately predict the thermodynamic properties and mechanical behavior of DNA [1]. They also demonstrated the ability to model the hybridization of complementary strands of DNA [2]. I will build on these efforts using my previous experience with simulation in order to develop a mesoscale model that, when combined with transition path sampling, will predict the dynamic folding of RNA tertiary structures.

Intellectual Merit

One of the most novel characteristics of RNA is its ability to double back on itself, creating a secondary structure that folds into a tertiary structure called a pseudoknot. DNA is simpler in that it does not exhibit this characteristic. This final tertiary structure is of particular interest because it influences the function of RNA and depends on the ionic conditions of the solvent. A change in the concentration of Mg²⁺ leads to a change in the shape of the pseudoknot, exposing information previously hidden within the folds of the RNA.

There are a wide variety of pseudoknots observed in nature. I propose to begin my research by studying the so-called ABCABC pseudoknot, consisting of three helices. This pseudoknot appears to be somewhat rare, having only been recognized in the E. coli ribosomal protein alpha and rpsO operons. However, such scarcity may only be the result of being overlooked in current protein databases. Despite its scarcity, the ABCABC pseudoknot has been studied with a variety of techniques including mutagenesis, calorimetry, and UV hyperchromicity. Consequently, experimental data exist for validation of the modeling proposed in the following paragraphs.

To predict the folding of an ABCABC pseudoknot, it is necessary to have a model that accurately captures the role of ordered hydration sites and solute-cation interactions. As was done in my previous research involving reacting flows, I propose to use a hierarchal modeling approach to develop a mesoscale RNA model. This hierarchal approach will involve atomistic simulations and coarser grain mesoscale simulations.

Atomistic molecular dynamics (MD) simulations will provide parameters such as bond angles that are crucial to the construction of the mesoscale model. I will perform atomistic simulations using codes developed by the De Pablo Group. Such codes have the ability to treat accurately the long-range electrostatic forces that will have a decisive effect on the final folded configuration of the pseudoknot. I will use an AMBER-99f force field and the Particle Mesh Ewald method to capture the interactions between the nucleic acids, water, and solvent ions. This approach will yield nanosecond and nanometer scale results that can be validated again existing experimental data.

The complexity of the RNA molecule makes it prohibitively demanding to simulate large molecules over relatively long time scales using traditional molecular dynamics. For the atomistic calculations described above, the helices must be constrained in order to run the calculations in a reasonable time. However, the breaking and forming of RNA-helices must be modeled in order to capture the dynamic folding of the ABCABC pseudoknot and to elucidate the pathway for folding, which is one of the primary aims of my project. Therefore, the main component of the proposed research is the development of a coarse-grained mesoscale model of RNA for use in Monte Carlo and Brownian dynamics simulations.

I will begin with the 3SPN (Three Sites Per Nucleotide) representation developed by Knotts and refined by Sambriski. This 3SPN model can quantitatively predict the melting temperature, mechanical properties, and rate of hybridization of DNA. I will modify the 3SPN model to be consistent with the aforementioned atomistic simulations. The initial parameters of the electrostatic model for monovalent ions at low concentrations will be obtained from Debye-Huckel theory. I will then develop a more rigorous model that will explicitly account for the presence of divalent counter ions such as Mg²⁺ and excess salts. I will develop advanced Monte Carlo techniques to determine the density of states using the model. Validation of this model will be performed using RNA folded structures deduced from X-Ray Diffraction experiments, atomistic simulations and experimentally measured melting temperatures and heat capacities.

Once the mesoscale model is completed, I will calculate trajectories from the unfolded to folded states of RNA. I will then use transition path sampling (TPS) [3] to sample the ensemble of states corresponding to the transition from the unfolded to folded configurations. The configurations calculated by TPS will describe the dynamic folding of an RNA pseudoknot in different ionic environments. The resulting understanding of RNA folding could potentially lead to new treatments for cancer and the genomic origins of multiple diseases.

In order to perform the necessary calculations, I will draw on Condor, the high throughput computing (HTC) system at UW. Additionally, as an NSF fellow, I will utilize the Teragrid resources to expedite the simulation of the ABCABC pseudoknot and other RNA structures.

Broader Impact

Despite the computational aspect of my research, it is just as experimental as gel electrophoresis. As an NSF fellow, I will be a proponent of the inclusion of computational experiments which are currently neglected in the secondary curriculum. With the proliferation of computers in secondary schools, educators in the biological and physical sciences have the necessary infrastructure for fostering an environment of numerical inquiry in the classroom. I will develop a series of lectures and simple algorithms accessible via an easy-to-use graphical interface that will allow students to reproduce in silico common biological experiments performed in vitro in the classroom. These tools will be shared with K-12 educators as part of the Wisconsin Teacher Enhancement Program (WisTEP).

To sharpen my skills as an educator, I will obtain a Teaching and Learning Certificate from the Delta Program (UW-Madison's implementation of CIRTL). I will also participate in the Genomic Sciences Training Program (GSTP) to disseminate those lessons learned in my research amongst biologists, computer scientists, statisticians and others. I am particularly excited about the prospect of participating in existing initiatives within the De Pablo group and the University of Wisconsin that are aimed at improving the participation of under-represented groups in computational science and engineering. One such initiative teaches 3-week workshops to high school students from at-risk backgrounds every summer. I look forward to working as an instructor in such workshops. I strongly believe that my previous experiences abroad have prepared for the challenges associated with getting younger students motivated about science, and I will be very effective in exposing them to previously unknown fields within biology and computational science.