Conidiation is a common and critical asexual reproductive mode in fungi. The ascomycetes, the largest group in the kingdom Fungi undergo conidiation. The wide array of morphological difference in a conidiophore and conidial size, shape, and cellular organization demonstrates the importance of evolution in driving the morphological and functional diversity. An important unanswered question is how these conidiation processes evolve. We hypothesized that a conidiation regulatory pathway was present in the ancestral species, and became specialized in the extant species to lead to morphological and functional diversity. To address this hypothesis we assessed the conserved function of conidiation regulators in two distantly related ascomycetes, Aspergillus nidulans and Neurospora crassa. Using sequence similarity analysis, N. crassa orthologs were characterized to seven main conidiation regulatory genes in A. nidulans (fluG, flbC, flbD, abaA, wetA, medA, and stuA). Expression of the N. crassa orthologs complemented defective conidiation in the A. nidulans fluG, flbD, wetA, medA, and stuA mutants. In contrast, abaA and flbC and the N. crassa orthologs did not share conserved biochemical function. Taken in context of other recent studies of conidiation regulators, there are four distinct evolutionary patterns: (i) Non-homologous genes with analogous roles in conidiation (‘brlA’ and ‘fl’), (ii) Orthologs with retained biochemical function that lack analogous role in conidiation (‘fluG’, ‘flbD’, and ‘wetA’), (iii) Orthologs with retained biochemical function and analogous roles in conidiation (‘medA’ and ‘stuA’), and (iv) Orthologs with biochemical function not conserved but with analogous roles in conidiation (‘abaA’ and ‘flbC’). These studies set the stage for long-term studies of how evolution proceeded during the evolution of conidiation at different levels of phylogenetic diversity. An understanding of how evolutionary mechanisms shape the dynamics of developmental pathways will be significant for our understanding of fungal evolution of other novel adaptations such as pathogenesis.

<p>Orthology identification is central to comparative and evolutionary genomics and is an active area of research. Despite a recent shift towards tree reconciliation and other phylogenetic methods, previous comparisons between different algorithms relied on real datasets where true orthology relationships are unknown and did not conclusively show whether phylogenetic methods truly outperform sequence similarity-based methods. Using simulated datasets generated from programs we developed, we show that tree reconciliation does perform better than similarity-based methods when the true species phylogeny is known. Even slight deviations in the species phylogeny can have adverse effects on the performance of reconciliation algorithms and in those cases similarity-based methods may perform better. Fusion and fission complicate orthology identification and are not explicitly considered in most existing algorithms. Programs designed specifically to investigate fusion and fission events are either unavailable or are not specific enough to identify events affecting orthologous genes. We developed a pipeline of programs called FusionFinder that perform this task, gaining new insights to the contributions of fusion and fission to bacterial protein evolution and uncover an unexpected abundance of fissions in <em>Bacillus anthracis</em> that to our knowledge yet to be reported.</p> / Master of Science (MS)

Transcription factor Wrinkled (WRI) 1 is associated with triacylglycerol (TAG) biosynthesis and accumulation in plant tissues. In avocado (Persea americana), a basal angiosperm, four WRI orthologs (1-4) were identified by transcriptome studies and the gene expression of WRI1, 2 and 3 was associated with TAG accumulation in mesocarp tissue. Therefore, it is hypothesized that putative PaWRI1, 2 and 3 but not PaWRI4 are responsible for TAG synthesis in non-seed tissues. To this extent, various in silico analyses were performed to identify similarities and distinct features of putative WRI genes in basal angiosperm relative to maize and Arabidopsis, a monocot and dicot respectively. Predicted structural comparison of these orthologs is expected to reveal the distinct features of avocado WRI paralogs that are associated with the regulation of oil biosynthesis in non-seed tissues.

The chb operon constitutes the genes essential for utilization of chitooligosaccharides in Escherichia coli and related species. The six genes of the operon code for a transcriptional regulator (ChbR) of the operon, a permease (ChbBCA), a monodeacetylase (ChbG), and a phospho-beta-glucosidase (ChbF). In the absence of the substrate, the operon is maintained in a transcriptionally repressed state, while presence of the substrate leads to transcriptional activation. Regulation of the chb operon is brought about by the concerted action of three proteins, the negative regulator NagC coded by the nag operon, the dual function regulator ChbR coded by the chb operon and the universal regulatory protein CRP. Mutations that lead to alterations in the regulation of the operon can facilitate utilization of cellobiose, in addition to chitooligosaccharides by E. coli. The studies presented in Chapter II were aimed at understanding the evolution of cellobiose utilization in Shigella sonnei, which is phylogenetically very close to E. coli. Cel+ mutants were isolated from a Cel- wild type S. sonnei strain. Interestingly, Cel+ mutants arose relatively faster on MacConkey cellobiose agar from the S. sonnei wild type strain compared to E. coli. Similar to E. coli, the Cel+ phenotype in S. sonnei mutants was linked to the chb operon. Deletion of the phospho-β-glucosidase gene, chbF also resulted in loss of the Cel+ phenotype, indicating that ChbF is responsible for hydrolysis of cellobiose in these mutants. Previous work from the lab has shown that acquisition of two classes of mutations is necessary and sufficient to give rise to Cel+ mutants in E. coli. The first class of mutations either within the nagC locus or at the NagC binding site within the chb promoter, lead to NagC derepression. The second class consisting of gain-of-function mutations in chbR enable the recognition of cellobiose as an inducer by ChbR and subsequent activation of the operon. However, in S. sonnei a single mutational event of an IS element insertion resulted in acquisition of this phenotype. Depending on the type and location of the insertion, the mutants were grouped as Type I, and Type II. In Type I mutants an 1S600 insertion between the inherent -10 and -35 elements within the chb promoter leads to ChbR-independent constitutive activation of the operon, while in Type II mutants, an IS2/600 insertion at -113/-114, leads to ChbR-dependent, cellobiose-inducible expression of the operon. The results presented also indicate that in addition to relieving NagC mediated repression, the insertion in Type II mutants also leads to increase in basal transcription from the chb promoter. Constitutive expression of the chb operon also results in utilization of the aromatic β-glucosides salicin and arbutin, in addition to cellobiose in Type I mutants, which indicates the promiscuous nature of permease and hydrolysis enzyme of the chb operon. This part of the thesis essentially demonstrates the different trajectories taken for the evolution of new metabolic function under conditions of nutrient stress by two closely related species. It emphasizes the significance of the strain background, namely the diversity of transposable elements in the acquisition of the novel function. The second part of this research investigation, detailed in Chapter III deals with experiments to characterize the eukaryotic orthologs of the last gene of the chb operon. The chbG gene of E. coli codes for a monodeacetylase of chitooligosaccharides like chitobiose and chitotriose. The protein belongs to a highly conserved, but less explored family of proteins called YdjC, whose orthologs are present in many prokaryotes and eukaryotes including mammals. The human YDJC locus located on chromosome 22 is linked to a variety of inflammatory diseases and the transcript levels are relatively high in stem cells and a few cancer cells. In silico analysis suggested that the mammalian YdjC orthologs possess sequence and structural similarity with the prokaryotic counterpart. The full length mouse YdjC ortholog, which is 85% identical to the human ortholog was cloned into a bacterial vector and expressed in a chbG deletion strain of E. coli. The mouse YdjC ortholog could neither promote growth of the strain on chitobiose nor induce transcription from the chb promoter. The purified mouse YdjC ortholog could not deacetylate chitobiose in vitro as well, suggesting that the mouse ortholog failed to complement the function of the E. coli counterpart, ChbG under the conditions tested in this study. In order to characterize the mammalian YdjC orthologs more elaborately, further experimentation was performed in mammalian cell lines. The results indicate that YdjC is expressed in mammalian cell lines of different tissue origin and the expression was seen throughout the cell. Overexpression of mouse Ydjc in a few mammalian cells also resulted in increased proliferation and migration, indicating a direct or indirect role of this protein in cell growth/proliferation. The mammalian orthologs of ChbG therefore appear to have related but distinct activities and substrates compared to the bacterial counterpart that need to be elucidated further.

Recent progress in fundamental biological sciences and medicine has considerably increased the quantity ofdata that can be studied and processed. The main limitation now is not retrieving data, but rather extractinguseful biological insights from the large datasets accumulated. More recent advances have provided detailedhigh-density data regarding metabolism (metabolomics) and protein expression (proteomics). Clearly, no single analytic methods, can provide a comprehensive understanding. Rather, the ability to link available datatogether in a coherent manner is required to obtain a complete view. The improving application of MachineLearning (ML) techniques provides the means to make continuous progress in processing complex data sets.A brief discussion is offered on the advantages of ML, the state-of-the-art in Deep Learning (DL) for proteinpredictions and the importance of ML in biological data processing. Noise stemming from incorrect classification or arbitrary/ambiguous labelling of data may arise when ML techniques are applied to large datasets. Furthermore, the stochasticity of biological systems needs to be considered for correctly evaluating theoutputs. Here we show the potential of a workflow to respond biological questions taking into consideration aperturbation of the biological data.  For controlling the applicability of models and maximizing the predictivity, in silico filtering schemescan usefully be applied as an “Ockham’s razor” before using any ML technique. After reviewing differentDL approaches for protein prediction purposes, this work shows that a computational approach in filteringsteps is a valuable tool for proteins classification when biological features are not fully annotated or reviewed.The in silico approach has identified putative proline transporters in fungi and plants as well as carotenoidbiosynthetic gene products in the plant family Brassicaceae. The proposed method is suitable for extractingfeatures of classification and then maximizing the use of a DL approach.

Advances in genomics allow researchers to measure the complete set of transcripts in cells. These transcripts include messenger RNAs (which encode for proteins) and microRNAs, short RNAs that play an important regulatory role in cellular networks. While this data is a great resource for reconstructing the activity of networks in cells, it also presents several computational challenges. These challenges include the data collection stage which often results in incomplete and noisy measurement, developing methods to integrate several experiments within and across species, and designing methods that can use this data to map the interactions and networks that are activated in specific conditions. Novel and efficient algorithms are required to successfully address these challenges. In this thesis, we present probabilistic models to address the set of challenges associated with expression data. First, we present a novel probabilistic error correction method for RNA-Seq reads. RNA-Seq generates large and comprehensive datasets that have revolutionized our ability to accurately recover the set of transcripts in cells. However, sequencing reads inevitably contain errors, which affect all downstream analyses. To address these problems, we develop an efficient hidden Markov modelbased error correction method for RNA-Seq data . Second, for the analysis of expression data across species, we develop clustering and distance function learning methods for querying large expression databases. The methods use a Dirichlet Process Mixture Model with latent matchings and infer soft assignments between genes in two species to allow comparison and clustering across species. Third, we introduce new probabilistic models to integrate expression and interaction data in order to predict targets and networks regulated by microRNAs. Combined, the methods developed in this thesis provide a solution to the pipeline of expression analysis used by experimentalists when performing expression experiments.

Cotton has been a world-wide economic staple in textiles and oil production. There has been a concerted effort for cotton improvement to increase yield and quality to compete with non-natural man-made fibers. Unfortunately, cultivated cotton has limited genetic diversity; therefore finding new marketable traits within cultivated cotton has reached a plateau. To alleviate this problem, traditional breeding programs have been attempting to incorporate practical traits from wild relatives into cultivated lines. This incorporation has presented a new problem: uncultivated cotton hampered by photoperiodism. Traditionally, due to differing floral times, wild and cultivated cotton species were unable to be bred together in many commercial production areas world-wide. This worldwide breeding problem has inhibited new trait incorporation. Before favorable traits from undomesticated cotton could be integrated into cultivated elite lines using marker-assisted selection breeding, the markers associated with photoperiod independence needed to be discovered. In order to increase information about this debilitating trait, we set out to identify informative markers associated with photoperiodism. This study was segmented into four areas. First, we reviewed the history of cotton to highlight current problems in production. Next, we explored cotton’s floral development through a study of floral transition candidate genes. The third area was an in-depth analysis of Phytochrome C (previously linked to photoperiod independence in other crops). In the final area of study, we used Genotype-By-Sequencing (GBS), in a segregating population, was used to determine photoperiod independence associated with single nucleotide polymorphisms (SNPs). In short, this research reported SNP differences in thirty-eight candidate gene homologs within the flowering time network, including photoreceptors, light dependent transcripts, circadian clock regulators, and floral integrators. Also, our research linked other discrete SNP differences, in addition to those contained within candidate genes, to photoperiodicity within cotton. In conclusion, the SNP markers that our study found may be used in future marker assisted selection (MAS) breeding schemas to incorporate desirable traits into elite lines without the introgression of photoperiod sensitivity.