abstract: While techniques for reading DNA in some capacity has been possible for decades, the ability to accurately edit genomes at scale has remained elusive. Novel techniques have been introduced recently to aid in the writing of DNA sequences. While writing DNA is more accessible, it still remains expensive, justifying the increased interest in in silico predictions of cell behavior. In order to accurately predict the behavior of cells it is necessary to extensively model the cell environment, including gene-to-gene interactions as completely as possible. Significant algorithmic advances have been made for identifying these interactions, but despite these improvements current techniques fail to infer some edges, and fail to capture some complexities in the network. Much of this limitation is due to heavily underdetermined problems, whereby tens of thousands of variables are to be inferred using datasets with the power to resolve only a small fraction of the variables. Additionally, failure to correctly resolve gene isoforms using short reads contributes significantly to noise in gene quantification measures. This dissertation introduces novel mathematical models, machine learning techniques, and biological techniques to solve the problems described above. Mathematical models are proposed for simulation of gene network motifs, and raw read simulation. Machine learning techniques are shown for DNA sequence matching, and DNA sequence correction. Results provide novel insights into the low level functionality of gene networks. Also shown is the ability to use normalization techniques to aggregate data for gene network inference leading to larger data sets while minimizing increases in inter-experimental noise. Results also demonstrate that high error rates experienced by third generation sequencing are significantly different than previous error profiles, and that these errors can be modeled, simulated, and rectified. Finally, techniques are provided for amending this DNA error that preserve the benefits of third generation sequencing. / Dissertation/Thesis / Doctoral Dissertation Computer Science 2017

Pipes in water distribution systems may change as they age. The accumulation of corrosion byproducts and suspended particles on the inside wall of aged pipes can increase pipe roughness and reduce pipe diameter. To quantify the hydraulic effects of irregular accumulation on the pipe walls, eleven aged pipes ranging in diameter from 0.020-m (0.75-in) to 0.100-m (4-in) and with varying degrees of turberculation were located and subjected to laboratory testing. The laboratory test results were used to determine a relationship between pipe diameter reduction and Hazen-Williams C. This relationship, combined with a manipulation of the Hazen-Williams equation, provided a simple and direct method for correcting the diameters of aged pipes in distribution models. Using EPANET 2, the importance of correcting pipe diameters when modeling water distribution systems containing aged pipes was investigated. Correcting the pipe diameters in the sample network reduced the modeled water age by up to 10% and changed the pattern of fluctuating water age that occurred as waters with different sources moved through the pipe network. In addition, two of the aforementioned aged pipes with diameters of 0.025-m (1-in) and 0.050-m (2-in) were modeled using Reynolds-Averaged Navier-Stokes (RANS) turbulence modeling. Flow was computed at Reynolds numbers ranging from 6700 to 31,000 using three turbulence models including a 4-equation v2-f model, and 2-equation realizable k-e; and k-ω models. In comparing the RANS results to the laboratory testing, the v2-f model was found to be most accurate, producing Darcy-Weisbach friction factors from 5% higher to 15% lower than laboratory-obtained values. The capability of RANS modeling to provide a detailed characterization of the flow in aged pipes was demonstrated. Large eddy simulation (LES) was also performed on a single 0.050-m (2-in) pipe at a Reynolds number of 6800. The Darcy-Weisbach friction factor calculated using LES was 20% less than obtained from experimental tests. Roughness elements smaller than the grid scale and deficiencies in the subgrid-scale model at modeling the complex three-dimensional flow structures due to the irregular pipe boundary were identified as likely sources of error. Even so, the utility of LES for describing complex flows was established.

Indiana University-Purdue University Indianapolis (IUPUI) / Heating, ventilation and air conditioning systems (HVAC) are the single largest consumer of energy in commercial and residential sectors. Minimizing its energy consumption without compromising indoor air quality (IAQ) and thermal comfort would result in environmental and financial benefits. Currently, most buildings still utilize constant air volume (CAV) systems with on/off control to meet the thermal loads. Such systems, without any consideration of occupancy, may ventilate a zone excessively and result in energy waste. Previous studies showed that CO2-based demand-controlled ventilation (DCV) methods are the most widely used strategies to determine the optimal level of supply air volume. However, conventional CO2 mass balanced models do not yield an optimal estimation accuracy. In this study, feed-forward neural network algorithm (FFNN) was proposed to estimate the zone occupancy using CO2 concentrations, observed occupancy data and the zone schedule. The occupancy prediction result was then utilized to optimize supply fan operation of the air handling unit (AHU) associated with the zone. IAQ and thermal comfort standards were also taken into consideration as the active constraints of this optimization. As for the validation, the experiment was carried out in an auditorium located on a university campus. The results revealed that utilizing neural network occupancy estimation model can reduce the daily ventilation energy by 74.2% when compared to the current on/off control.

No description available.

Firebrand pile ignition of adjacent materials requires an in-depth understanding of heat transfer and flow profile within the firebrand pile. Modeling the firebrand pile as a fibrous porous medium, this study identified a porosity-permeability correlation that accurately describes the transport properties of a firebrand pile. The conduction-based model and Kozeny-Carman model were identified and examined by experiment, where firebrand porosity and permeability were collected with a wind tunnel. The conduction-based model was more stable and more accurate in the porosity range of interest. Pore network models were developed for the simulation of flow profiles utilizing the permeability data collected. The non-uniform network, which better represents a randomly stack firebrand pile, resulted in a more complex multidimensional flow within the pile. / Master of Science / Firebrands are known to be one of the primary ways wildfires can spread. They are mostly small pieces of flammable materials originating from vegetation or wooden structures that can be carried by wind ahead of the fire. The accumulation of firebrands on flammable materials tends to create ignitions, which calls for an in-depth understanding of temperature and airflow within the firebrand pile. Simplifying the firebrand pile as a porous medium, this study identified a relationship between how much void is present in the pile and the resistance of airflow of a firebrand pile. The conduction-based model and Kozeny-Carman model were identified and examined by experiment with a wind tunnel. The conduction-based model was determined to better describe the relationship. Pore network models were developed for the simulation of flow through the firebrand pile utilizing the data collected in the experiment, which provided an understanding of how airflow behaves inside the pile. A non-uniform flow network inside the pile led to a more complex, multidimensional flow through the firebrand pile.

Intrusive thoughts are aversive, private thoughts that are unwanted but intrude into consciousness, and are a ubiquitous phenomenon that approximately 93% of the population experiences (Radomsky et. al., 2014). Obsessional thoughts are a key etiological component of obsessive-compulsive disorder (OCD). Cognitive behavioral models of OCD conceptualize intrusive thoughts and obsessive thoughts as the same phenomenon occurring on a spectrum, with obsessional thoughts being a particular type of intrusion that is integral to the development and maintenance of OCD (Moulding, 2014). However, there is little evidence to demonstrate this relationship. This study examined intrusive thoughts across stratified groups based on intrusion frequency using ecological momentary assessment. This exploratory study examined potential idiographic differences in reported experiences of people ranging from low to high levels of intrusive thought frequency. Personalized contemporaneous networks were constructed from participant data and examined for differences in topography, measures of centrality, and magnitude of relationships between nodes. These networks are visually distinct, providing a glimpse into a wide variety of ways in which participants experience and relate to their intrusive thoughts.

No description available.

Carefully engineered mine ventilation is critical to the safe operation of underground longwall mines. Currently, there are several options for simulation of mine ventilation. This research was conducted to rapidly simulate an underground longwall mine, especially for the use of tracer gas in an emergency situation. In an emergency situation, limited information about the state of mine ventilation system is known, and it is difficult to make informed decisions about safety of the mine for rescue personnel. With careful planning, tracer gases can be used to remotely ascertain changes in the ventilation system. In the meantime, simulation of the tracer gas can be conducted to understand the airflow behavior for improvements during normal operation. Better informed decisions can be made with the help of both tracer gas technique and different modeling approaches. This research was made up of two main parts. One was a field study conducted in an underground longwall mine in the western U.S. The other one was a simulation of the underground longwall mine with different approaches, such as network modeling and Computational Fluid Dynamics (CFD) models. Networking modeling is the most prevalent modeling technique in the mining industry. However, a gob area, which is a void zone filled with broken rocks after the longwall mining, cannot be simulated in an accurate way with networking modeling. CFD is a powerful tool for modeling different kinds of flows under various situations. However, it requires a significant time investment for the expert user as well as considerable computing power. To take advantage of both network modeling and CFD, the hybrid approach, which is a combination of network modeling and CFD was established. Since tracer gas was released and collected in the field study, the tracer gas concentration profile was separately simulated in network modeling, CFD model, and hybrid model in this study. The simulated results of airflow and tracer gas flow were analyzed and compared with the experimental results from the field study. Two commercial network modeling software packages were analyzed in this study. One of the network modeling software also has the capability to couple with CFD. A two-dimensional (2D) CFD model without gob was built to first analyze the accuracy of CFD. More 2D CFD models with gob were generated to determine how much detail was necessary for the gob model. Several three-dimensional (3D) CFD models with gob were then created. A mesh independence study and a sensitivity study for the porosity and permeability values were created to determine the optimal mesh size, porosity and permeability values for the 3D CFD model, and steady-state simulation and transient simulations were conducted in the 3D CFD models. In the steady-state simulation, a comparison was made between the 3D CFD models with and without taking the diffusivity of SF6 in air into account. Finally, the different simulation techniques were compared to measured field data, and assessed to determine if the hybrid approach was considerably simpler, while also providing results superior to a simple network model. / Master of Science

Master of Science / Department of Biology / Samantha Wisely / As human populations increase world-wide, land use and land cover are altered to support the rapid anthropogenic expansion. These landscape alterations influence patterns of zoonotic infectious disease emergence and propagation. It is therefore becoming increasingly important to study emerging and re-emerging diseases to predict and manage for future epidemics. Studies of directly-transmitted infectious diseases should consider three components of disease epidemiology: characteristics of the pathogen, ecology of the host, and habitat configuration of the underlying landscape. I studied the influence of both the host ecology of the striped skunk (Mephitis mephitis) and the alteration of the underlying landscape on the epidemiology of rabies virus in the Flint Hills of Kansas. This tall-grass prairie is experiencing woody expansion due to anthropogenic disturbance, altering the landscape on which the rabies virus emerges and spreads. We first studied the behavioral and social ecology of the striped skunk using field and genetic methods. We concluded that 1) striped skunks reached high population densities in anthropogenically disturbed habitats, 2) these individuals were not closely related, and 3) contact rates could be influenced by temperature. Using habitat-specific skunk densities from this initial study, we created spatially-explicit contact networks of skunk populations across the Upper Kansas River Watershed and simulated the emergence and spread of rabies through the system. This modeling approach revealed a threshold of forest habitat beyond which striped skunks became increasingly connected and the rabies virus reached greater extents across the landscape. Based on these findings we recommend fire regimes and land cover alterations to reduce woody encroachment across the Flint Hills and to avoid future disease epidemics in the region.

Travel time reliability is deemed as one of the most important factors affecting travelers’ route choice decisions. However, existing practices mostly consider average travel time only. This dissertation establishes a methodology framework to overcome such limitation. Semi-standard deviation is first proposed as the measure of reliability to quantify the risk under uncertain conditions on the network. This measure only accounts for travel times that exceed certain pre-specified benchmark, which offers a better behavioral interpretation and theoretical foundation than some currently used measures such as standard deviation and the probability of on-time arrival. Two path finding models are then developed by integrating both average travel time and semi-standard deviation. The single objective model tries to minimize the weighted sum of average travel time and semi-standard deviation, while the multi-objective model treats them as separate objectives and seeks to minimize them simultaneously. The multi-objective formulation is preferred to the single objective model, because it eliminates the need for prior knowledge of reliability ratios. It offers an additional benefit of providing multiple attractive paths for traveler’s further decision making. The sampling based approach using archived travel time data is applied to derive the path semi-standard deviation. The approach provides a nice workaround to the problem that there is no exact solution to analytically derive the measure. Through this process, the correlation structure can be implicitly accounted for while simultaneously avoiding the complicated link travel time distribution fitting and convolution process. Furthermore, the metaheuristic algorithm and stochastic dominance based approach are adapted to solve the proposed models. Both approaches address the issue where classical shortest path algorithms are not applicable due to non-additive semi-standard deviation. However, the stochastic dominance based approach is preferred because it is more computationally efficient and can always find the true optimal paths. In addition to semi-standard deviation, on-time arrival probability and scheduling delay measures are also investigated. Although these three measures share similar mathematical structures, they exhibit different behaviors in response to large deviations from the pre-specified travel time benchmark. Theoretical connections between these measures and the first three stochastic dominance rules are also established. This enables us to incorporate on-time arrival probability and scheduling delay measures into the methodology framework as well.