Thesis (Ph. D.)--Oregon State University, 2007. / Printout. Includes bibliographical references (leaves 160-180). Also available on the World Wide Web.

Thesis (Ph. D.)--University of Washington, 2006. / Vita. Includes bibliographical references (leaves 151-169).

Thesis (Ph.D.)--York University, 2005. Graduate Programme in Chemistry. / Typescript. Includes bibliographical references. Also available on the Internet. MODE OF ACCESS via web browser by entering the following URL: http://wwwlib.umi.com/cr/yorku/fullcit?pNR11631

Thesis (Ph. D.)--Ohio State University, 2008. / Title from first page of PDF file. Includes vita. Includes bibliographical references (p. 118-127).

Studies undertaken to investigate the binding of low molecular mass analogues of polyphenolic vegetable tannins to collagen have prompted the development of a new method to investigate protein-ligand interactions. This method, the continuous-flow dynamic dialysis method (CFDD), differs from conventional dialysis procedures used for protein-ligand binding studies. In this method, the ligand concentration in the diffusate is monitored automatically at successive closely spaced time intervals while being continuously eluted from the dialysis cell. The primary data obtained by this method consists of a series of spectrophotometric absorbance measurements representing the ligand concentration in the sink compartment of a dialysis cell. This primary data is recorded by means of a data logging device onto a punched paper tape for subsequent computer processing. Two original methods are presented for analysing the primary data to extract the protein-ligand binding isotherm. The first of these is a direct analysis which relies on Fick's first law of diffusion. In this method it is necessary to establish, by means of a control experiment, a value for the ligand permeation constant. This is used in a subsequent analysis to establish a relationship between the measured rate of diffusion of the ligand from a protein-ligand mixture and the concentration of unbound ligand which is in equilibrium with the protein-ligand complex. The protein-ligand binding isotherm is obtained from parametric equations which give the quantity of ligand bound to the protein and the concentration of unbound ligand in the sample compartment as functions of time. The second method, which is more general, utilizes the same primary data but is based on establishing a system transfer function to characterise the dialysis and eluting processes. This analysis depends on the linearity of the system and utilizes numerical laplace transforms of the primary data sets obtained from control and protein-ligand dialyses. Laplace transforms are used to effect a deconvolution of the transfer function from the primary data and yield the concentration of ligand in equilibrium with the protein-ligand complex. This procedure yields, simultaneously, both the total ligand concentration and the concentration of unbound ligand in the protein compartment of the dialysis cell. These quantities are used to establish the binding isotherm for the protein ligand system. Numerical inversion of the laplace transforms in this analysis is effected by their reduction to Fourier series. The experimental reliability of the continuous-flow dynamic dialysis method, and validity of the two analytical methods used to derive a binding isotherm from dialysis data are evaluated from studies of the binding of phenol red to bovine serum albumin (BSA) at 15⁰, 20⁰ and 25⁰ C, as well as from simulated binding curves generated by the numerical solution of the differential equations used to describe the dialysis and elution process in terms of a two-site Scatchard binding model. The method is used to investigate the binding to collagen of a series of low molecular mass phenolic compounds which can be isolated from Wattle and Quebracho vegetable tannin extracts. These compounds can be considered as monomeric precursor analogues of the polymeric vegetable tannins. The binding of these ligands to collagen is shown to be characterised by high capacity, low affinity binding in which the uptake of ligand by the protein increases linearly with increasing ligand concentration. Collagen exhibits no indication of site saturation for these ligands over the experimentally accessible concentration ranges investigated.

Infectious diseases from novel viruses have become a major public health concern. Rapid identification of virus–host interactions can reveal mechanistic insights into infectious diseases and shed light on potential treatments. Current computational prediction methods for novel viruses are based mainly on protein sequences. However, it is not clear to what extent other important features, such as the symptoms caused by the viruses, could contribute to a predictor. Disease phenotypes (i.e., signs and symptoms) are readily accessible from clinical diagnosis and we hypothesize that they may act as a potential proxy and an additional source of information for the underlying molecular interactions between the pathogens and hosts. We developed DeepViral, a deep learning based method that predicts protein– protein interactions (PPI) between humans and viruses. Motivated by the potential utility of infectious disease phenotypes, we first embedded human proteins and viruses in a shared space using their associated phenotypes and functions, supported by formalized background knowledge from biomedical ontologies. By jointly learning from protein sequences and phenotype features, DeepViral significantly improves over existing sequence-based methods for intra- and inter-species PPI prediction. Lastly, we propose a novel experimental setup to realistically evaluate prediction methods for novel viruses.

This thesis focuses on the two research projects which have applied machine learning techniques to the protein-related topics. The first project is to use protein sequences and the interaction graph to address the protein-protein interaction prediction problem. The second project is to leverage the sequences of protein loops within and beyond homologs to predict the protein loop structures. In the protein-protein interaction prediction project, we applied the pretrained language models, which were trained on large sets of protein sequences, as one of the protein feature extraction methods. Another feature extraction method is the graph learning on the protein interaction graph. The graph learning embeddings and the language model embeddings were fed into classification models to predict if two proteins are interacting or not. We trained and tested our methods on the S. cerevisiae dataset and the human dataset. Our results are comparable to or better than other state-of-art methods, with the advantages that our method is faster at the sample preparation step and has a larger application scope for requiring only protein sequences. We also did experiments with datasets from different similarity cutoffs between the train and test set of the human dataset, and our method has shown an effective prediction ability even with a strict similarity cutoff. In the protein loop prediction project, we utilized the attention-based encoder-decoder language models to predict the protein loop inter-residue distances from the protein loop sequences. We fed the model with the loop sequences and received arrays of numbers representing the distances between each C_α pair in the loops. We utilized two different strategies to reconstruct the loops from the predicted distances. One was firstly to calculate the C_α coordinates from the predicted distances, and then apply a fast full-atom reconstruction method starting from C_α coordinates to build the local loop structures. Our local loop structure prediction results of this method are very competitive with low local RMSDs, especially with the lowest standard deviations. The second method was to integrate the predicted inter-residue distances as constraints to the de novo loop prediction method PLOP (Jacobson et al. 2004). We tested the loop reconstruction process on the 8-res and 12-res loop benchmark sets. This method has the best performance compared to other state-of-art methods, and the incorporation of such machine learning step decreased the computing time of the standalone PLOP program.

Deformed epidermal autoregulatory factor 1 (DEAF1) encodes a transcription factor essential in early embryonic and neuronal development. In humans, mutations in the DNA binding domain of DEAF1 cause intellectual disability together with clinical characteristics collectively termed DEAF1-associated neurodevelopmental disorders (DAND). The objective of this study is to 1) assess the pathogenicity of newly identified variants using established functional assays, and 2) confirm and map the interaction domain of DEAF1 with HDAC1 and evaluate the importance of DEAF1-HDAC1 interaction on DEAF1-mediated transcriptional repression. Exome sequencing analysis identified six de novo DEAF1 mutations (p.D200Y, p.S201R, p.K250E, p.D251N, p.K253E, and p.F297S). Promoter activity experiments indicate DEAF1 transcriptional repression activity was altered by p.K250E, p.K253E, and p.F297S. Transcriptional activation activity was altered by p.K250E, p.K253E, p.F297S, and p.D251N. Combined with clinical phenotype of the patients, this work establishes the pathogenicity of new DEAF1 variants. Previous studies identified a potential protein interaction between DEAF1 and several proteins of the nucleosome remodeling and deacetylating (NuRD) complex including Histone Deacetylase 1 (HDAC1), Retinoblastoma Binding Protein 4 (RBBP4), Methyl CpG Binding Domain Protein 3 (MBD3). GST pull-down and co-immunoprecipitation (CoIP) assays confirmed and mapped the interaction with HDAC1 between amino acids 113 – 176 of DEAF1. To determine whether DEAF1-mediated repression requires HDAC1 activity, HEK293t wild type or CRISPR/Cas9-mediated DEAF1 knockout cells were treated with the HDAC inhibitor Trichostatin A (TSA). Interestingly, this study demonstrates that the requirement of HDAC1 activity on DEAF1-mediated transcriptional repression activity is target gene specific and expands our understanding of DEAF1 mediated transcriptional repression.

The PDZ domain of proteins mediates a protein-protein interaction by recognizing the hydrophobic C-terminal tail of the target protein. One of the challenges put forth by the DREAM (Discussions on Reverse Engineering Assessment and Methods) 2009 Challenge consists of predicting a position weight matrix (PWM) that describes the specificity profile of five PDZ domains to their target peptides. We consider the primary structures of each of the five PDZ domains as a numerical sequence derived from graph-theoretic models of each of the individual amino acids in the protein sequence. Using available PDZ domain databases to obtain known targets, the graph-theoretic based numerical sequences are then used to train a neural network to recognize their targets. Given the challenge sequences, the target probabilities are computed and a corresponding position weight matrix is derived. In this work we present our method. The results of our method placed second in the DREAM 2009 challenge.

Functional immobilization and lateral organization of proteins into micro- and nanopatterns is an important prerequisite for miniaturizing analytical and biotechnological devices. In this thesis I present novel and versatile approaches for high contrast surface micropatterning of proteins, artificial membranes and live cells based on maleimide photochemistry. The patterning strategy is carried-out on glass substrates exploiting a poly(ethylene glycol) PEG polymer layer as a compatible scaffold. The flexible PEG cushion prevents unspecific proteins attachment and cell adhesion to surfaces. The versatility of this method is demonstrated by means of different orthogonal chemistries using covalent- and affinity- based interactions for protein immobilization. Furthermore, using maleimide based alkyl-thiol chemistry, I utilized the patterning approach for capturing liposomes and proteoliposomes onto surfaces. Formation of fluid patterned polymer-supported membranes demonstrating lateral diffusion of lipids and proteins was confirmed by biophysical assays. A similar approach was used for micropatterning of transmembrane proteins in surface adhered live cells.