Glioblastoma (GB) is one of the deadliest cancers and the most common form of adult primary brain tumors. SGEF (ARHGEF26) has been previously shown to be overexpressed in GB tumors, play a role in cell invasion/migration, and increase temozolomide (TMZ) resistance.[3] It was hypothesized parental LN229 cell lines with SGEF knockdown (LN229-SGEFi) will show decreased metabolism in the MTS assay and decreased colony formation in a colony formation assay compared to parental LN229 cells after challenging the two cell lines with TMZ. For WB and co-immunoprecipitation (co-IP), parental LN229 cells with endogenous SGEF and BRCA were expected to interact and stain in the BRCA1:IP WB. LN229-SGEFi cells were expected to show very little SGEF precipitated due to shRNA targeted knockdown of SGEF. In conditions with mutations in the BRCA1 binding site (LN229-SGEFi + AdBRCAm/AdDM), SGEF expression was expected to decrease compared to parental LN229 or LN229-SGEFi cells reconstituted with WT SGEF (LN229-SGEFi + AdWT). LN229 infected with AdSGEF with a mutated nuclear localization signal (LN229-SGEFi + AdNLS12m) were expected to show BRCA and SGEF interaction since whole cell lysates were used for the co-IP. MTS data showed no significant differences in metabolism between the two cell lines at all three time points (3, 5, and 7 days). Western blot analysis was successful at imaging both SGEF and BRCA1 protein bands from whole cell lysate. The CFA showed no significant difference between cell lines after being challenged with 500uM TMZ. The co-IP immunoblot showed staining for BRCA1 and SGEF for all lysate samples, including unexpected lysates such as LN229-SGEFi, LN229-SGEFi + AdBRCAm, and LN229-SGEFi + AdDM. These results suggested either an indirect protein interaction between BRCA1 and SGEF, an additional BRCA binding site not included in the consensus, or possible detection of the translocated SGEF in knockdown cells lines since shRNA cannot enter the nucleus. Further optimization of CO-IP protocol, MTS assay, and CFA will be needed to characterize the SGEF/BRCA1 interaction and its role in cell survival.

Cancer rates vary between people, between cultures, and between tissue types, driven by clinically relevant distinctions in the risk factors that lead to different cancer types. Despite the importance of cancer location in human health, little is known about tissue-specific cancers in non-human animals. We can gain significant insight into how evolutionary history has shaped mechanisms of cancer suppression by examining how life history traits impact cancer susceptibility across species. Here, we perform multi-level analysis to test how species-level life history strategies are associated with differences in neoplasia prevalence, and apply this to mammary neoplasia within mammals. We propose that the same patterns of cancer prevalence that have been reported across species will be maintained at the tissue-specific level. We used a combination of factor analysis and phylogenetic regression on 13 life history traits across 90 mammalian species to determine the correlation between a life history trait and how it relates to mammary neoplasia prevalence. The factor analysis presented ways to calculate quantifiable underlying factors that contribute to covariance of entangled life history variables. A greater risk of mammary neoplasia was found to be correlated most significantly with shorter gestation length. With this analysis, a framework is provided for how different life history modalities can influence cancer vulnerability. Additionally, statistical methods developed for this project present a framework for future comparative oncology studies and have the potential for many diverse applications.

The research presented in this Honors Thesis provides development in machine learning models which predict future states of a system with unknown dynamics, based on observations of the system. Two case studies are presented for (1) a non-conservative pendulum and (2) a differential game dictating a two-car uncontrolled intersection scenario. In the paper we investigate how learning architectures can be manipulated for problem specific geometry. The result of this research provides that these problem specific models are valuable for accurate learning and predicting the dynamics of physics systems.<br/><br/>In order to properly model the physics of a real pendulum, modifications were made to a prior architecture which was sufficient in modeling an ideal pendulum. The necessary modifications to the previous network [13] were problem specific and not transferrable to all other non-conservative physics scenarios. The modified architecture successfully models real pendulum dynamics. This case study provides a basis for future research in augmenting the symplectic gradient of a Hamiltonian energy function to provide a generalized, non-conservative physics model.<br/><br/>A problem specific architecture was also utilized to create an accurate model for the two-car intersection case. The Costate Network proved to be an improvement from the previously used Value Network [17]. Note that this comparison is applied lightly due to slight implementation differences. The development of the Costate Network provides a basis for using characteristics to decompose functions and create a simplified learning problem.<br/><br/>This paper is successful in creating new opportunities to develop physics models, in which the sample cases should be used as a guide for modeling other real and pseudo physics. Although the focused models in this paper are not generalizable, it is important to note that these cases provide direction for future research.

High-entropy alloys possessing mechanical, chemical, and electrical properties that far exceed those of conventional alloys have the potential to make a significant impact on many areas of engineering. Identifying element combinations and configurations to form these alloys, however, is a difficult, time-consuming, computationally intensive task. Machine learning has revolutionized many different fields due to its ability to generalize well to different problems and produce computationally efficient, accurate predictions regarding the system of interest. In this thesis, we demonstrate the effectiveness of machine learning models applied to toy cases representative of simplified physics that are relevant to high-entropy alloy simulation. We show these models are effective at learning nonlinear dynamics for single and multi-particle cases and that more work is needed to accurately represent complex cases in which the system dynamics are chaotic. This thesis serves as a demonstration of the potential benefits of machine learning applied to high-entropy alloy simulations to generate fast, accurate predictions of nonlinear dynamics.

Robots are often used in long-duration scenarios, such as on the surface of Mars,where they may need to adapt to environmental changes. Typically, robots have been built specifically for single tasks, such as moving boxes in a warehouse or surveying construction sites. However, there is a modern trend away from human hand-engineering and toward robot learning. To this end, the ideal robot is not engineered,but automatically designed for a specific task. This thesis focuses on robots which learn path-planning algorithms for specific environments. Learning is accomplished via genetic programming. Path-planners are represented as Python code, which is optimized via Pareto evolution. These planners are encouraged to explore curiously and efficiently. This research asks the questions: “How can robots exhibit life-long learning where they adapt to changing environments in a robust way?”, and “How can robots learn to be curious?”.

Colorimetric assays are an important tool in point-of-care testing that offers several advantages to traditional testing methods such as rapid response times and inexpensive costs. A factor that currently limits the portability and accessibility of these assays are methods that can objectively determine the results of these assays. Current solutions consist of creating a test reader that standardizes the conditions the strip is under before being measured in some way. However, this increases the cost and decreases the portability of these assays. The focus of this study is to create a machine learning algorithm that can objectively determine results of colorimetric assays under varying conditions. To ensure the flexibility of a model to several types of colorimetric assays, three models were trained on the same convolutional neural network with different datasets. The images these models are trained on consist of positive and negative images of ETG, fentanyl, and HPV Antibodies test strips taken under different lighting and background conditions. A fourth model is trained on an image set composed of all three strip types. The results from these models show it is able to predict positive and negative results to a high level of accuracy.

The goal of this project was to design and create a genetic construct that would allow for <br/>tumor growth to be induced in the center of the wing imaginal disc of Drosophila larvae, the <br/>R85E08 domain, using a heat shock. The resulting transgene would be combined with other <br/>transgenes in a single fly that would allow for simultaneous expression of the oncogene and, in <br/>the surrounding cells, other genes of interest. This system would help establish Drosophila as a <br/>more versatile and reliable model organism for cancer research. Furthermore, pilot studies were <br/>performed, using elements of the final proposed system, to determine if tumor growth is possible <br/>in the center of the disc, which oncogene produces the best results, and if oncogene expression <br/>induced later in development causes tumor growth. Three different candidate genes were <br/>investigated: RasV12, PvrACT, and Avli.

The field of biomedical research relies on the knowledge of binding interactions between various proteins of interest to create novel molecular targets for therapeutic purposes. While many of these interactions remain a mystery, knowledge of these properties and interactions could have significant medical applications in terms of understanding cell signaling and immunological defenses. Furthermore, there is evidence that machine learning and peptide microarrays can be used to make reliable predictions of where proteins could interact with each other without the definitive knowledge of the interactions. In this case, a neural network was used to predict the unknown binding interactions of TNFR2 onto LT-ɑ and TRAF2, and PD-L1 onto CD80, based off of the binding data from a sampling of protein-peptide interactions on a microarray. The accuracy and reliability of these predictions would rely on future research to confirm the interactions of these proteins, but the knowledge from these methods and predictions could have a future impact with regards to rational and structure-based drug design.