Biology

Permanent URI for this communityhttp://hdl.handle.net/1903/11810

Browse

Subject: search.filters.subject.Random Forest

Search Results

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    Simple statistical models predict C-to-U edited sites in plant mitochondrial RNA
    (Springer Nature, 2004-09-16) Cummings, Michael P; Myers, Daniel S
    RNA editing is the process whereby an RNA sequence is modified from the sequence of the corresponding DNA template. In the mitochondria of land plants, some cytidines are converted to uridines before translation. Despite substantial study, the molecular biological mechanism by which C-to-U RNA editing proceeds remains relatively obscure, although several experimental studies have implicated a role for cis-recognition. A highly non-random distribution of nucleotides is observed in the immediate vicinity of edited sites (within 20 nucleotides 5' and 3'), but no precise consensus motif has been identified. Data for analysis were derived from the the complete mitochondrial genomes of Arabidopsis thaliana, Brassica napus, and Oryza sativa; additionally, a combined data set of observations across all three genomes was generated. We selected datasets based on the 20 nucleotides 5' and the 20 nucleotides 3' of edited sites and an equivalently sized and appropriately constructed null-set of non-edited sites. We used tree-based statistical methods and random forests to generate models of C-to-U RNA editing based on the nucleotides surrounding the edited/non-edited sites and on the estimated folding energies of those regions. Tree-based statistical methods based on primary sequence data surrounding edited/non-edited sites and estimates of free energy of folding yield models with optimistic re-substitution-based estimates of ~0.71 accuracy, ~0.64 sensitivity, and ~0.88 specificity. Random forest analysis yielded better models and more exact performance estimates with ~0.74 accuracy, ~0.72 sensitivity, and ~0.81 specificity for the combined observations. Simple models do moderately well in predicting which cytidines will be edited to uridines, and provide the first quantitative predictive models for RNA edited sites in plant mitochondria. Our analysis shows that the identity of the nucleotide -1 to the edited C and the estimated free energy of folding for a 41 nt region surrounding the edited C are the most important variables that distinguish most edited from non-edited sites. However, the results suggest that primary sequence data and simple free energy of folding calculations alone are insufficient to make highly accurate predictions.
  • Thumbnail Image
    Item
    Few amino acid positions in rpoB are associated with most of the rifampin resistance in Mycobacterium tuberculosis
    (Springer Nature, 2004-09-28) Cummings, Michael P; Segal, Mark R
    Mutations in rpoB, the gene encoding the β subunit of DNA-dependent RNA polymerase, are associated with rifampin resistance in Mycobacterium tuberculosis. Several studies have been conducted where minimum inhibitory concentration (MIC, which is defined as the minimum concentration of the antibiotic in a given culture medium below which bacterial growth is not inhibited) of rifampin has been measured and partial DNA sequences have been determined for rpoB in different isolates of M. tuberculosis. However, no model has been constructed to predict rifampin resistance based on sequence information alone. Such a model might provide the basis for quantifying rifampin resistance status based exclusively on DNA sequence data and thus eliminate the requirements for time consuming culturing and antibiotic testing of clinical isolates. Sequence data for amino acid positions 511–533 of rpoB and associated MIC of rifampin for different isolates of M. tuberculosis were taken from studies examining rifampin resistance in clinical samples from New York City and throughout Japan. We used tree-based statistical methods and random forests to generate models of the relationships between rpoB amino acid sequence and rifampin resistance. The proportion of variance explained by a relatively simple tree-based cross-validated regression model involving two amino acid positions (526 and 531) is 0.679. The first partition in the data, based on position 531, results in groups that differ one hundredfold in mean MIC (1.596 μg/ml and 159.676 μg/ml). The subsequent partition based on position 526, the most variable in this region, results in a > 354-fold difference in MIC. When considered as a classification problem (susceptible or resistant), a cross-validated tree-based model correctly classified most (0.884) of the observations and was very similar to the regression model. Random forest analysis of the MIC data as a continuous variable, a regression problem, produced a model that explained 0.861 of the variance. The random forest analysis of the MIC data as discrete classes produced a model that correctly classified 0.942 of the observations with sensitivity of 0.958 and specificity of 0.885. Highly accurate regression and classification models of rifampin resistance can be made based on this short sequence region. Models may be better with improved (and consistent) measurements of MIC and more sequence data.
  • Thumbnail Image
    Item
    IDENTIFICATION AND CHARACTERIZATION OF REGULATORY MIRNAS AND MRNAS IN THE LONGITUDINAL HUMAN HOST RESPONSE TO VAGINAL MICROBIOTA
    (2017) Smith, Steven Bradley; Ravel, Jacques; El-Sayed, Najib; Biology; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
    The human vagina and the bacterial communities that reside therein exist in a finely balanced mutualistic association. Dysbiotic states of the vaginal microbiota, including bacterial vaginosis (BV), are characterized by a paucity of Lactobacillus spp., the presence of a wide array of strict and facultative anaerobes, and a pH >4.5. Symptoms such as odor and discharge can accompany these microbial dysbiotic states, however, epidemiologically, vaginal dysbioses have been associated with increased susceptibility to STIs, including chlamydia. The mechanisms by which vaginal microbiota protect or increase the risk to infections remain unknown. This thesis aimed to identify the molecular factors that control host cellular responses to Lactobacillus spp.-dominated and dysbiotic microbiota. Chapter 2 characterized the in vivo host microRNA (miRNA) response to different types of vaginal microbiota to gain insight into host functions that play a role in vaginal homeostasis. Leveraging daily collected vaginal samples in conjunction with a machine learning approach, eight miRNAs were discovered to be differently controlled by vaginal microbiota. Of these, expression of miR-193b, known to regulate host cell proliferation, was increased by Lactobacillus spp.-dominated microbiota. In vitro, vaginal epithelial cells exposed to Lactobacillus spp. culture supernatants exhibited reduced epithelial cell proliferation, high miRNA-193b expression and decreased abundance of cyclin D1. More importantly, epithelial cell proliferation was identified as a requirement for efficient Chlamydia trachomatis infection. Chapter 3 characterized the in vitro transcriptome of epithelial cells exposed to Lactobacillus spp. relative to Gardnerella vaginalis, a surrogate for dysbiotic vaginal microbiota. Immune response and cell cycle pathways were found to be among the most modulated by Lactobacillus spp. Longitudinal gene expression suggested a role of histone deacetylases (HDAC) as an intermediary between immune stimulation and cell proliferation. Additionally, the epidermal growth factor receptor (EGFR), required for C. trachomatis infection, was decreased when epithelial cells were exposed to Lactobacillus spp. These findings contribute to the fundamental understanding of the vaginal microbiota’s role in cellular homeostasis as a requirement for resistance to STI agents such as C. trachomatis, and ultimately will lead to improved preventive strategies against STIs through the modulation of vaginal microbiota composition and function.