With the large-scale metagenome sequencing data produced currently, alignment-free metagenomic profiling approaches have demonstrated the effectiveness of Markov models in addressing the limitations of alignment-based techniques, particularly in handling unclassified reads. The development of POSMM (Python Optimized Standard Markov Model), employing SMM (Standard Markov Model) algorithm, initially showcased competitive performance when compared to tools such as Kraken2. However, when subjected to simulated damages present in ancient metagenomics data, shortcomings emerged, leading to false positives or misclassified sequences that compromised overall classification accuracy. To address this problem, we developed a segmental genome model (SGM) algorithm based on the generation of the ensemble of models representing distinct classes of DNA segments in a genome. SGM incorporated a recursive segmentation and clustering approach to segregate regions of distinct composition in a microbial genome. An ensemble of higher-order Markov models is trained on DNA clusters generated for each genome. A database of models of genomes, with each genome represented by multiple Markov models are then queried to infer the origin of reads from a metagenome. SGM was benchmarked using diverse synthetic metagenome datasets of varying composition, read lengths, and error profiles. The comparative assessment showed that SGM consistently outperformed SMM. SGM brings in significant advances in alignment-free profiling, offering a new promising avenue for metagenomic exploration through its integration in the next version of POSMM. Furthermore, leveraging the power of integration of alignment-free and alignment-based approaches and highlighting the versatility and practicality of these methods in addressing critical public health challenges, we developed a statistical analysis and machine learning pipeline to identify candidate microbes associated with COVID-19. This involved a meta-analysis of the whole genome sequencing data of COVID-19 patients' samples and its predictive modeling to discern the distinctive microbial features. We improve and explore alignment-free metagenome profiling to raise the bar in metagenome profiling in complex real-world samples.

<p> This work describes the fundamental study of two bacterial toxins with computational methods, the rational design of a potent inhibitor using molecular dynamics, as well as the development of two bioinformatic methods for mining genomic data. Clostridium difficile is an opportunistic bacillus which produces two large glucosylating toxins. These toxins, TcdA and TcdB cause severe intestinal damage. As Clostridium difficile harbors considerable antibiotic resistance, one treatment strategy is to prevent the tissue damage that the toxins cause. The catalytic glucosyltransferase domain of TcdA and TcdB was studied using molecular dynamics in the presence of both a protein-protein binding partner and several substrates. These experiments were combined with lead optimization techniques to create a potent irreversible inhibitor which protects 95% of cells in vitro. Dynamics studies on a TcdB cysteine protease domain were performed to an allosteric communication pathway. Comparative analysis of the static and dynamic properties of the TcdA and TcdB glucosyltransferase domains were carried out to determine the basis for the differential lethality of these toxins. Large scale biological data is readily available in the post-genomic era, but it can be difficult to effectively use that data. Two bioinformatics methods were developed to process whole-genome data. Software was developed to return all genes containing a motif in single genome. This provides a list of genes which may be within the same regulatory network or targeted by a specific DNA binding factor. A second bioinformatic method was created to link the data from genome-wide association studies (GWAS) to specific genes. GWAS studies are frequently subjected to statistical analysis, but mutations are rarely investigated structurally. HyDn-SNP-S allows a researcher to find mutations in a gene that correlate to a GWAS studied phenotype. Across human DNA polymerases, this resulted in strongly predictive haplotypes for breast and prostate cancer. Molecular dynamics applied to DNA Polymerase Lambda suggested a structural explanation for the decrease in polymerase fidelity with that mutant. When applied to Histone Deacetylases, mutations were found that alter substrate binding, and post-translational modification.</p>

Transmembrane beta-barrels (TMBB) are a special structural class of proteins predominately found in the outer membranes of Gram-negative bacteria, mitochondria, and chloroplasts. TMBBs are surface-accessible proteins that perform a variety of functions ranging from iron acquisition to osmotic regulation. These properties make TMBBs tempting targets for vaccine or drug therapy development A prediction method based on the physicochemical properties of experimentally characterized TMBB structures was developed to predict TMBB-encoding genes from genomic databases. The algorithm's prediction efficiency was tested using a non-redundant set of sequences from proteins of known structure. The algorithm was based on the work of Wimley (2002), but was improved because of its disappointingly high false-positive prediction rate. The improved prediction algorithm developed in this study was shown to be more accurate than previously published prediction methods. Its accuracy is near 99% when using the most efficient prediction criteria, i.e. where the most known TMBBs are correctly predicted and the most non-TMBBs are correctly excluded. The improved algorithm was used to predict the abundance of TMBBs in 611 chromosomes from Gram-negative and acid-fast bacteria. The average predicted abundance of genomic TMBBs was 3%, which is consistent with previous estimates Predicted outer membrane protein L (OmpL) from Salmonella typhimurium LT2, was tested as a model for validating the prediction method. All of the physicochemical and spectroscopic properties exhibited by OmpL are consistent with other known TMBBs. Recombinant OmpL localizes to the outer membrane when expressed in Escherichia coli; has a beta-sheet-rich secondary structure with stable tertiary contacts in the presence of either detergent micelles or a lipid bilayer; OmpL also forms a pore through which small hydrophilic solutes can diffuse. Together, this data proves that OmpL is a true TMBB, which supports the computational prediction. This work significantly contributes to the advancement of TMBB research / acase@tulane.edu

Algorithms for DNA Sequence Assembly and Motif Search

Thesis (Ph. D.)--University of California, San Francisco, 2009. / Source: Dissertation Abstracts International, Volume: 70-06, Section: B, page: 3519. Adviser: Patricia C. Babbitt. Includes supplementary digital materials.

Thesis (Ph. D.)--University of Illinois at Urbana-Champaign, 2009. / Source: Dissertation Abstracts International, Volume: 70-06, Section: B, page: 3735. Adviser: Gerard Wong. Includes bibliographical references (leaves 114-119). Available on microfilm from Pro Quest Information and Learning.

<p> Base stacking provides stability to nucleic acid duplexes, and base unstacking is involved in numerous biological functions related to nucleic acids, including replication, repair, transcription, and translation. The patterns of base stacking and unstacking in available nucleic acid crystal structures were classified after separation into their individual single strand dinucleotide components and clustering using a <i>k</i>-means-based ensemble clustering method. The A- and B-form proximity of these dinucleotide structures were assessed to discover that RNA dinucleotides can approach B-form-like structures. Umbrella sampling molecular dynamics simulations were used to obtain the potential of mean force profiles for base unstacking at 5'-termini for all 16 dinucleotides. A rate calculation method was investigated and implemented using small test compounds and applied to a base unstacking transition to predict a rate for 5'-terminal base fraying. The findings can be applied for localized nucleic acid structure prediction, and for comparison of molecular dynamics simulation-based investigations of nucleic acid distortions to experimental structural data.</p>

<p> In both mice and humans, males and females exhibit differences in behavior and response to neurological and psychological diseases that are linked to the cortex and hippocampus. The perinatal exposure of males to testosterone secreted by the testes creates alterations in neural structures and behaviors that can persist throughout their lives; however, the molecular mechanisms that underlie the actions of sex steroids to produce these lasting changes are still unclear. Given that regulation of gene expression is a primary mechanism whereby sex steroids exert changes to an organism, I sought to identify genes expressed at different levels between the sexes in the cortex and hippocampus and to determine the effect of testosterone on expression of these genes. Using gene expression microarrays and RT-qPCR, I identified genes that are differentially expressed between the sexes in the neonatal mouse cortex and hippocampus; however, whether perinatal testosterone is regulating these differences remains unclear.</p>

<p> Cystic fibrosis (CF) is the most common lethal autosomal recessive genetic disorder that affects the Caucasian population. CF is caused by mutations in the CF transmembrane conductance regulator (CFTR), and is characterized by a viscous airway surface liquid (ASL) that impairs mucociliary function and facilitates bacterial infection. The molecular mechanisms by which these symptoms result from CFTR malfunction are unclear. We hypothesized that expression and secretion of innate immune proteins is altered in CF ASL. </p><p> We sought to use cell culture models in which the only source of secreted proteins was differentiated airway epithelium. Since CFTR localizes to the apical surface of airway submucosal glands (SMG) and ciliated epithelium, cell culture models that recapitulate two parts of respiratory tract epithelium were studied: 1) SMG acini and 2) mucociliary epithelium. </p><p> We developed a three-dimensional system wherein CF (&Delta;F508/&Delta;F508) and non-CF human bronchial epithelial (HBE) cells differentiated on Matrigel into polarized glandular acini with mature lumens by two weeks with no significant variability in size. Bronchial acini expressed and secreted SMG proteins, MUC5B and lysozyme, at day 22, and exhibited vectorial secretions that were collected along with acinar cell lysates. Proteome profiling demonstrated unique protein signatures for each cellular space. However, abundant contaminating proteins from Matrigel and growth media were identified. Therefore, the ALI cell culture model of airway epithelium was chosen for quantitative proteomic comparison of CF and non-CF HBE apical secretions because the protein-rich media does not contact the apical surface. </p><p> CF and non-CF HBE cells were labeled by stable isotope labeling with amino acids in cell culture and differentiated at ALI. LC-MS/MS and bioinformatic analysis identified seventy-one proteins with altered levels in CF secretions (+/&minus;1.5 fold-change; p-value&lt;0.05). Validation with antibody based biochemical assays demonstrated increased levels of MUC5AC, MUC5B, fibronectin and MMP9, and increased proteolysis/activation of complement C3, in CF secretions. Overall, the function of altered proteins in the CF secretome is indicative of an airway epithelium in a state of repair and altered immunity in the absence of infection, suggesting the downstream consequences of mutated CFTR in CF airways set the stage for chronic inflammation and infection.</p>

<p> Human genes are split into regions that code for protein, exons, and regions that don't, introns. Upon transcription, the removal of these intervening introns is necessary if a usable mRNA molecule is to be translated. The process of intron removal and subsequent ligation of exons is called splicing and is carried out by a large complex called the spliceosome. This process is driven by sequence elements within the pre-mRNA itself and is the major contributor of diversity to the human transcriptome. Due to the ubiquitous nature of alternative splicing in almost every multi-exon gene, the regulation pathways of exon inclusion are a subject of wide study. </p><p> The different lengths of introns and exons as well as location of splice sites in a pre-mRNA molecule have been shown to have differing affects on the spliceosomes ability to recognize them. Using <i>in vitro</i> splicing and complex formation assays in parallel with cell transfection experiments, we determined that the distance between two splice sites across the intron or across the exon are strong predictors of splice site usage. Additionally, we found that two splice sites interact differently when placed at different lengths apart. Our findings suggest a mechanism for observed selection of specific intron/exon architectures. </p><p> Splice site recognition is also influenced by the presence of protein binding sequence elements in the pre-mRNA that alter spliceosomal recruitment. Previously, these proteins and sequence elements had been rigidly classified into splice enhancing or inhibiting categories. We show that this rigid classification is incorrect. We found that the location of these elements relative to the splice site determines their enhancing or silencing effect. That is, an enhancing element found upstream of a splice site imposes a silencing effect when relocated downstream of the splice site (and vice versa). </p><p> Spliceosomal proteins are conserved from yeast to humans. The sequence elements used in pre-mRNA sequences have been evolving over time but under pressure from multiple cellular processes, including splicing. To observe the effect of splicing on evolution, we took advantage of the synonymous mutation positions that are under the least amount of selective pressure from the genetic code. We mutated these positions and found that some caused a large decrease in exon inclusion. When we analyzed the comparative alignment data, we found that these specific nucleotide mutations were selected against across species in order to maintain exon inclusion. SNP analysis showed that this pattern of selection was broadly observable at synonymous positions throughout the human genome.</p>