No / A hybrid molecular mechanics and quantum mechanics solid state DFT method is used to re-rank the stability of racemic and enantiopure crystal structures of four molecules; 4-hydroxymethyl-2-oxazolidinone, 5-hydroxymethyl-2-oxazolidinone, 2-(4-hydroxyphenyl)-2,5,5-trimethylpyrrolidine-1-oxy and 2-(3-hydroxyphenyl)-2,5,5-trimethylpyrrolidine-1-oxy. Previous work using a force field based method to predict these crystal structures indicated that the lattice energy may be a suitable criterion for predicting whether a chiral molecule will resolve spontaneously on crystallisation. However, in some cases, the method had predicted an unrealistically high lattice energy for the structure corresponding to the experimentally observed one. The Hybrid DFT method successfully predicts those molecules which resolve spontaneously and furthermore predicts satisfactory lattice energies for all experimentally observed structures. Based on a comparison of the predicted lattice energies from the two methods it is concluded that the force fields used were not sufficiently accurate to predict spontaneous resolution with any confidence. However, the Hybrid DFT method is shown to be sufficiently accurate for making such predictions.

Low-dimensional systems are an important and intensely studied area of condensed matter physics. When a material is forced to adopt a low-dimensional structure, its behaviour is often dramatically different to that of the bulk phase. It is vital to predict the structures of low-dimensional systems in order to reliably predict their properties. To this end, the ab initio random structure searching (AIRSS) method, which has previously been used to identify the structures of bulk materials, has been extended to deal with the case of nanowires encapsulated inside carbon nanotubes. Such systems are a rapidly developing area of research with important nanotechnological applications, including information storage, energy storage and chemical sensing. The extended AIRSS method for encapsulated nanowires (ENWs) was implemented and used to identify the structures formed by germanium telluride, silver chloride, and molybdenum diselenide ENWs. In each of these cases, a number of novel nanowire structures were identified, and a phase diagram predicting the ground state nanowire structure as a function of the radius of the encapsulating nanotube was calculated. In the case of germanium telluride, which is a technologically important phase-change material, the potential use of GeTe ENWs as switchable nanoscale memory devices was investigated. The vibrational properties of silver chloride ENWs were also considered, and a novel scheme was developed to predict the Raman spectra of systems which can be decomposed into multiple weakly interacting subsystems. This scheme was used to obtain a close approximation to the Raman spectra of AgCl ENWs at a fraction of the computational cost that would otherwise be necessary. The encapsulation of AgCl was shown to produce substantial shifts in the Raman spectra of nanotubes, providing an important link with experiment. A method was developed to predict the stress-strain response of an ENW based on a polygonal representation of its surface, and was used to investigate the elastic response of molybdenum diselenide ENWs. This was used to predict stress-radius phase diagrams for MoSe_2 ENWs, and hence to investigate stress-induced phase change within such systems. The X-ray diffraction of ENWs was also considered. A program was written to simulate X-ray diffraction in low-dimensional systems, and was used to predict the diffraction patterns of some of the encapsulated GeTe nanowire structures predicted by AIRSS. By modelling the interactions within a bundle of nanotubes, diffraction patterns for bundles of ENWs were obtained.

<p>Proteins play a crucial roll in all biological processes. The wide range of protein functions is made possible through the many different conformations that the protein chain can adopt. The structure of a protein is extremely important for its function, but to determine the structure of protein experimentally is both difficult and time consuming. In fact with the current methods it is not possible to study all the billions of proteins in the world by experiments. Hence, for the vast majority of proteins the only way to get structural information is through the use of a method that predicts the structure of a protein based on the amino acid sequence.</p><p>This thesis focuses on improving the current protein structure prediction methods by combining different prediction approaches together with machine-learning techniques. This work has resulted in some of the best automatic servers in world – Pcons and Pmodeller. As a part of the improvement of our automatic servers, I have also developed one of the best methods for predicting the quality of a protein model – ProQ. In addition, I have also developed methods to predict the local quality of a protein, based on the structure – ProQres and based on evolutionary information – ProQprof. Finally, I have also performed the first large-scale benchmark of publicly available homology modeling programs.</p>

The goals of gene prediction using computational approaches are to determine gene location and the corresponding functionality of the coding region. A subset of gene prediction is the gene structure prediction problem, which is to define the exon-intron boundaries of a gene. Gene prediction follows two general approaches: statistical patterns identification and sequence similarity comparison. Similarity based approaches have gained increasing popularity with the recent vast increase in genomic data in GenBank. The proposed gene prediction algorithm is a similarity based algorithm which capitalizes on the fact that similar sequences bear similar functions. The proposed algorithm, like most other similarity based algorithms, is based on dynamic programming. Given a genomic DNA, X = x1 xn and a closely related cDNA, Y = y1 yn, these sequences are aligned with matching pairs stored in a data set. These indexes of matching sets contain a large jumble of all matching pairs, with a lot of cross over indexes. Dynamic programming alignment is again used to retrieve the longest common non-crossing subsequence from the collection of matching fragments in the data set. This algorithm was implemented in Java on the Unix platform. Statistical comparisons were made against other software programs in the field. Statistical evaluation at both the DNA and exonic level were made against Est2genome, Sim4, Spidey, and Fgenesh-C. The proposed gene structure prediction algorithm, by far, has the best performance in the specificity category. The resulting specificity was greater than 98%. The proposed algorithm also has on par results in terms of sensitivity and correlation coeffcient. The goal of developing an algorithm to predict exonic regions with a very high level of correctness was achieved.

Proteins play a crucial roll in all biological processes. The wide range of protein functions is made possible through the many different conformations that the protein chain can adopt. The structure of a protein is extremely important for its function, but to determine the structure of protein experimentally is both difficult and time consuming. In fact with the current methods it is not possible to study all the billions of proteins in the world by experiments. Hence, for the vast majority of proteins the only way to get structural information is through the use of a method that predicts the structure of a protein based on the amino acid sequence. This thesis focuses on improving the current protein structure prediction methods by combining different prediction approaches together with machine-learning techniques. This work has resulted in some of the best automatic servers in world – Pcons and Pmodeller. As a part of the improvement of our automatic servers, I have also developed one of the best methods for predicting the quality of a protein model – ProQ. In addition, I have also developed methods to predict the local quality of a protein, based on the structure – ProQres and based on evolutionary information – ProQprof. Finally, I have also performed the first large-scale benchmark of publicly available homology modeling programs.

RNA molecules play important roles, including catalysis of chemical reactions and control of gene expression, and their functions largely depend on their folded structures. Since determining these structures by biochemical means is expensive, there is increased demand for computational predictions of RNA structures. One computational approach is to find the secondary structure (a set of base pairs) that minimizes a free energy function for a given RNA conformation. The forces driving RNA folding can be approximated by means of a free energy model, which associates a free energy parameter to a distinct considered feature. The main goal of this thesis is to develop state-of-the-art computational approaches that can significantly increase the accuracy (i.e., maximize the number of correctly predicted base pairs) of RNA secondary structure prediction methods, by improving and refining the parameters of the underlying RNA free energy model. We propose two general approaches to estimate RNA free energy parameters. The Constraint Generation (CG) approach is based on iteratively generating constraints that enforce known structures to have energies lower than other structures for the same molecule. The Boltzmann Likelihood (BL) approach infers a set of RNA free energy parameters which maximize the conditional likelihood of a set of known RNA structures. We discuss several variants and extensions of these two approaches, including a linear Gaussian Bayesian network that defines relationships between features. Overall, BL gives slightly better results than CG, but it is over ten times more expensive to run. In addition, CG requires software that is much simpler to implement. We obtain significant improvements in the accuracy of RNA minimum free energy secondary structure prediction with and without pseudoknots (regions of non-nested base pairs), when measured on large sets of RNA molecules with known structures. For the Turner model, which has been the gold-standard model without pseudoknots for more than a decade, the average prediction accuracy of our new parameters increases from 60% to 71%. For two models with pseudoknots, we obtain an increase of 9% and 6%, respectively. To the best of our knowledge, our parameters are currently state-of-the-art for the three considered models.

Protein structure prediction is concerned with the prediction of a protein's three dimensional structure from its amino acid sequence. Such predictions are commonly performed by searching the possible structures and evaluating each structure by using some scoring function. If it is assumed that the target protein structure resembles the structure of a known protein, the search space can be significantly reduced. Such an approach is referred to as comparative structure prediction. When such an assumption is not made, the approach is known as ab initio structure prediction. There are several difficulties in devising efficient searches or in computing the scoring function. Many of these problems have ready solutions from known mathematical methods. However, the problems that are yet unsolved have hindered structure prediction methods from more ideal predictions. The objective of this study is to present a complete framework for ab initio protein structure prediction. To achieve this, a new search strategy is proposed, and better techniques are devised for computing the known scoring functions. Some of the remaining problems in protein structure prediction are revisited. Several of them are shown to be intractable. In many of these cases, approximation methods are suggested as alternative solutions. The primary issues addressed in this thesis are concerned with local structures prediction, structure assembly or sampling, side chain packing, model comparison, and structural alignment. For brevity, we do not elaborate on these problems here; a concise introduction is given in the first section of this thesis. Results from these studies prompted the development of several programs, forming a utility suite for ab initio protein structure prediction. Due to the general usefulness of these programs, some of them are released with open source licenses to benefit the community.

Protein structure prediction is concerned with the prediction of a protein's three dimensional structure from its amino acid sequence. Such predictions are commonly performed by searching the possible structures and evaluating each structure by using some scoring function. If it is assumed that the target protein structure resembles the structure of a known protein, the search space can be significantly reduced. Such an approach is referred to as comparative structure prediction. When such an assumption is not made, the approach is known as ab initio structure prediction. There are several difficulties in devising efficient searches or in computing the scoring function. Many of these problems have ready solutions from known mathematical methods. However, the problems that are yet unsolved have hindered structure prediction methods from more ideal predictions. The objective of this study is to present a complete framework for ab initio protein structure prediction. To achieve this, a new search strategy is proposed, and better techniques are devised for computing the known scoring functions. Some of the remaining problems in protein structure prediction are revisited. Several of them are shown to be intractable. In many of these cases, approximation methods are suggested as alternative solutions. The primary issues addressed in this thesis are concerned with local structures prediction, structure assembly or sampling, side chain packing, model comparison, and structural alignment. For brevity, we do not elaborate on these problems here; a concise introduction is given in the first section of this thesis. Results from these studies prompted the development of several programs, forming a utility suite for ab initio protein structure prediction. Due to the general usefulness of these programs, some of them are released with open source licenses to benefit the community.

The protein side chain prediction is an essential issue, in protein structure prediction, protein design, and protein docking problems. The protein side chain packing problem has been proved to be NP-hard. Our method for solving this problem is first to reduce it to the clique finding problem, and then we can apply the Ant Colony Optimization (ACO) algorithm to solve it. In knowledge-based methods, the rotamers are chosen from the rotamer library, which are based on the pair of dihedral angles, £r and £p, of backbones. We take the coordinate rotamer library as the template, so we do not need the complicated energy function to calculate the bond length and bond angle. We use a simple score function to evaluate the goodness of a solution of the ACO algorithm. The score function combines some factors, such as charge-charge interaction, intermolecular hydrogen bonds, disulfide bonds and van der Waals interactions. The experimental results show that our score function is biologically sensible. We compare our computational results with the results of SCWRL 3.0 and the residue-rotamer-reduction (R3) algorithm. The accuracy of our method outperforms both SCWRL 3.0 and R3 methods.

The objective of this thesis is to study the crystal structures and electronic properties of solids at high pressure using state-of-the-art electronic structure computational methods. The thesis is divided into two main sections. The first part is to examine the performance and reliability of several current density functionals in the description of the electronic structures of small band gap materials and strongly correlated systems. The second part is to compare and evaluate two recently proposed first-principles methods for the prediction of stable structures of solids at high pressure. To accomplish the first goal, first-principle electronic structure calculations employing density functional theory (DFT) and several “correlation corrected” functionals calculations were used to investigate the properties of solid AlH3 and EuO at high pressure. The primary reason to study AlH3 is to resolve a discrepancy between previously predicted superconductivity behavior at 110 GPa but was not observed in experimental resistance measurements. The key to resolve the discrepancy is an accurate calculation of the valence and conduction band energies. The results shows that the Fermi surface is modified by the “improved” functionals over the previous calculations using “standard” gradient corrected functional. These changes in the Fermi surface topology removed the possibility of nesting of the electronic bands, therefore, solid AlH3 above 100 GPa is a poor metal instead of a superconductor. In the second system, we have studied EuO with highly localized electrons in the Eu 4f orbitals. A particular interest in this compound is the report of an anomalous isostructural phase transition with a significant volume reduction at 35-40 GPa and the relationship with the electronic state of Eu at high pressure. Using the Hubbard on-site repulsion model (LDA+U), we successfully predicted the insulator  metal transition of EuO at 12 GPa and the trend in the Mössbauer isomer shifts. However, the isostructural transition was not reproduced. The U on-site repulsion to localized Eu 4f orbtials helped to ameliorate some deficiencies of the PBE functional and improved the agreement with experimental observations but not all the properties were correctly reproduced. The second objective of this investigation is to predict energetically stable crystalline structures at high pressure. The reliability and relative efficiency of two recently proposed structure prediction methods, viz, Particle Swarm Optimization (PSO) and the Genetic Algorithm (GA) were critically examined. We applied the techniques to two separate systems. The first system is solid CS2. The motivation is that this compound was recently found to be a superconductor with a critical temperature of 6 K from 60 – 120 GPa. However, no crystalline structure was found by experiment in this pressure range. Our calculations suggest the energetic favorable structures contain segregated regions of carbon and sulfur atoms. The sulfur atoms adopt a planar closed pack arrangement forming 2D square or hexagonal networks and the carbon atoms tend to form hexagonal rings. A global minimum crystalline structure with structural features observed in the amorphous structure was found and shown to be superconductive. In the second case, we studied the possibility on the existence of Xe-halides (XeHn (H=Cl, Br and I, n = 1, 2 and 4)) compounds below 60 GPa. We reported the stability, crystal and electronic structures, vibrational and optical spectra of a number of stoichiometric crystalline polymorphs. We found that only XeCl and XeCl2 form thermodynamically stable compounds at pressure exceeding 60 GPa. A stable cubic fcc structure of XeBr2 was found to be a superconductor with critical temperature of 1.4 K. From these studies, we found both merits and shortcomings with the two structural prediction approaches. In the end, we proposed a hybrid approach to assure the same stable structure is predicted from both computational strategies.