It can be natural to believe that many of the traditional issues of scale have been eliminated or at least greatly reduced via cloud computing. That is, if one can create a seemingly wellfunctioning cloud application that operates correctly on small or moderate-sized problems, then the very nature of cloud programming abstractions means that the same application will run as well on potentially significantly larger problems. In this paper, we present our experiences taking MODISAzure, our satellite data processing system built on the Windows Azure cloud computing platform, from the proof-of-concept stage to a point of being able to run on significantly larger problem sizes (e.g., from national-scale data sizes to global-scale data sizes). To our knowledge, this is the longest-running eScience application on the nascent Windows Azure platform. We found that while many infrastructure-level issues were thankfully masked from us by the cloud infrastructure, it was valuable to design additional redundancy and fault-tolerance capabilities such as transparent idempotent task retry and logging to support debugging of user code encountering unanticipated data issues. Further, we found that using a commercial cloud means anticipating inconsistent performance and black-box behavior of virtualized compute instances, as well as leveraging changing platform capabilities over time. We believe that the experiences presented in this paper can help future eScience cloud application developers on Windows Azure and other commercial cloud providers.

This dissertation consists of five chapters addressing analytically and empirically U.S. Postwar business cycle fluctuations. Markov Switching models and Bayesian estimation methods are used to investigate United States macroeconomic dynamics in the last 60 years. Chapter 1 introduces the structure of this dissertation. Chapter 2 proposes a dynamic stochastic general equilibrium (DSGE) model with Markov Switching and heteroskedastic shocks to examine the role of agents' beliefs separately from changes in monetary policy in explaining inflation fluctuations. Bayesian analysis is conducted with Markov Switching to support regime switches in the private sector, in the implementation of monetary policy and in the volatility of shocks in the U.S. Postwar economy, which are related to the "Great Inflation", the "Great Moderation" and the 2008 financial crisis. A counterfacutal analysis found that if agents maintained a weak response to macroeconomic dynamics over time, there would be lower inflation during the "Great Inflation". In addition, irrespectively to monetary policy regimes, supply shocks are the main driver of inflation fluctuations, while demand shocks are the main source of changes in the output gap. However, when agents maintain a higher risk aversion towards consumption with a higher slope in the Phillips curve, demand shocks also play a role in driving inflation, even though supply shocks are still the main driver of inflation. Chapter 3 emphasizes on the monetary policy with an investigation on the assumption that policymakers commit to a Taylor rule, using a time-varying inflation-unemployment dynamic model on U.S. economy. This chapter is based on the conjecture that potential policymakers' misperception may be originated from unobserved deviations of unemployment from its natural rate. Five processes are proposed for policymakers' belief under commitment to inflation and unemployment and compare them with a baseline autoregressive process without commitment. The models are estimated using Bayesian techniques. Empirical results are as follows: First, policymakers' belief is very persistent even when it commits to a Taylor-type policy rule. Second, the run-up of U.S. inflation around 1980 can be mostly attributed to policymakers' misperception while the peak surge of inflation in 1974 is possibly a result of non-policy shocks. Third, models with commitment dominate models without commitment, especially in periods of large oscillations in inflation. In particular, when policymakers are committed to respond to a Taylor-type policy rule, the average loss function is considerably reduced over time, thus effectively lessening potential misperceptions. Chapter 4 introduces a simple version of adaptive expectation to a dynamic stochastic general equilibrium (DSGE) model to evaluate goodness of fitness and forecasting performance on U.S. macroeconomic indicators. Analytical maximum likelihood estimation results represent a DSGE model with adaptive expectation outperforms a DSGE model with rational expectation. In addition to providing a better fit of inflation and output gap in the U.S. Postwar macro economy, a DSGE model with adaptive expectation also leads to redundant lagged inflation in fitting inflation dynamics. Chapter 5 concludes and proposes future extension.

Proteins play the most dynamic and diverse roles among all the biomacromolecules in living organisms. With the fast development in biotechnology, protein gains more and more interests for a wide range of applications, such as biochemical synthesis, sensing, environmental protection and medical applications. However, the vulnerable nature of proteins has hindered their boarder applications. Developing vectors and stabilizers that help to more effectively deliver protein therapeutics or stabilize proteins has been an essential theme of the field.

In this dissertation, novel strategies have been developed for protein delivery by a method of in-situ free-radical polymerization or by a self-assembling/ crosslinking approach. Protein nanocapsules made from the in-situ polymerization method were also used building blocks to synthesize protein composites with dramatically enhanced stability. This dissertation research consist with four topics outlined below:

1. Organophosphorus hydrolase (OPH) nanocapsules were synthesized by in-situ free-radical polymerization with enhanced activity and stability. These nanocapsules are highly potent for decontamination, as well as in vivo detoxification of organophosphorus as prophylactics or antidotes.

2. Protein nanocapsules made by the in-situ polymerization technique were used as building blocks to synthesize protein-silica composites through a sol-gel process. The microenvironment around the proteins was engineered through judicious choices of the polymer monomers and the silica precursors, enabling the synthesis robust enzyme composites with high activity for various industry applications.

3. Protein nanocapsules were synthesized by assembling the proteins with self-crosslinking cell penetrating polymer (SCP). Self-crosslinking by disulfide-bond formation within the SCP leads to the formation of robust protein nanocapsules with highly retained activity. These nanocapsules were able to be effectively internalized by cells without significant cytotoxicity, and the protein cargo could be effectively released upon exposure to glutathione that break down the disulfide bonds.

4. Protein nanocapsules with zwitterionic shells were synthesized by a self-assembling approach. The zwitterionic shells protected the nanocapsules from being uptake by macrophages, while such shells could be detached once exposing to a low pH environment. This approach provides a suitable platform towards protein therapeutics with prolonged circulation time with ability of being delivered into the cells.

Overall, this dissertation work provides novel strategies toward better protein-delivery vectors, as well as composites with better protein stability and activity, for a broad range of applications.

Theoretical studies of molecules have historically relied on deterministic algorithms, stochastic simulations, and physical models. Recently modern data-driven methods are starting to infiltrate into various fields of molecular science, opening new possibilities for solving problems that are difficult to tackle through traditional approaches. The accumulation of data, advancement of machine learning algorithms and improvement in hardware enables a plethora of data-driven approaches to surpass traditional methods in terms of accuracy and efficiency, but questions remain about how well these data-driven methods can generalize to unseen data to do true prediction. In this dissertation, I will show that when data-driven models are combined with physics-based approaches, through either feature design, or exerting constrains on the machine learning models, new standards can be established in the fields of molecule characterization and generation.

Nuclear magnetic resonance (NMR) chemical shifts (CS) are extremely sensitive to the local atomic environments for different nuclei in a molecule, and therefore is a common technique in molecule characterization. In chapter 2, I focus on the design of the UCBShift predictor for CSs for proteins in aqueous solution. The UCBShift method uniquely fuses a transfer prediction module, which employs sequence and structure alignments to select reference chemical shifts from a database, with a machine learning model that uses carefully curated and physics-inspired features, to predict CSs for proteins with higher accuracy and better reliability compared to all popular methods such as SHIFTX2 and SPARTA+. This chapter further delineates how UCBShift benefits from realistic data that has not been heavily curated, and surpasses existing CS calculators in terms of real-world performance without eliminating test predictions ad hoc.

However, in order to achieve rigorous and consistent improvement for an arbitrary molecular system, carefully curated feature sets specifically for proteins can be limiting, and we seek features from theoretical calculations. In Chapter 3 I describe the development of a novel neural network model which employs quantum mechanical (QM) features from affordable Density Functional Theory (DFT) calculations, along with geometric features of the molecular systems, to predict NMR chemical shieldings. The resulting iShiftML model predicts chemical shieldings approaching the highest level of accuracy under the modern theoretical framework of CCSD(T) in the complete basis set limit, but without the computational burden that limits its applicability to large systems. Not only does the iShiftML model demonstrate excellent predictive performance when compared with small molecule gas phase experimental CSs, but it also offers a capability to predict chemical shifts for much more complex natural products, and can be used for differentiating diasteromers based on chemical shift assignments. This chapter unveils new possibilities for integrating machine learning and QM calculations for accurate and transferable molecular characterization.

In Chapters 4 an 5, my research addresses fundamental issues for large and small molecule generation relevant to proteins and drug molecules. Chapter 4 describes the Int2Cart method that uses a recurrent neural network to predict the correlations between bond lengths, bond angles and backbone torsion angles and amino acid sequence of a protein. By incorporating these correlations, proteins reconstructed from just torsion angles display not only physically more accurate bond lengths and bond angles, but the reconstructed proteins are closer to their crystal structures than under the common assumption that bond lengths and bond angles are fixed, or that coming from a static library that only relies on local residue geometries. I have also shown potential applications of this method in estimating model quality for AlphaFold2 predicted structures, and reconstructing intrinsically disordered protein (IDP) ensembles with decreased steric overlap. Chapter 5 describes the combination of deep generative networks trained by reinforcement learning and physical docking study. I developed the iMiner method, which generates de novo drug-like molecules with an augmented binding potency towards specific protein targets, facilitating the discovery of potential new therapeutic targets. SARS-COV-2 Main Protease was used as an example to show that our generated molecules cover a broader chemical space than crowdsourcing efforts, and the newly generated molecules exert optimized interactions and correct shape for the compatibility with the binding pocket.

To summarize, this dissertation contains multiple methods that harmonize data-driven models and physics-based approaches in the area of NMR spectroscopy, protein structure modelling and de novo drug discovery, which provides a new perspective for researchers striving to leverage computational methods in molecular science and chemical biology.