Forensic analysis of DNA from sexual assault kits is a laborious process. These samples may be a mixture of sperm and male or female epithelial cells (E-cells). Generally, it is the sperm cells that are of greatest forensic value. Since its introduction in 1985 by Gill, Jefferys and Warrett, differential extraction has remained an essential pre-PCR extraction procedure adopted by most forensic laboratories for the preferential lysis of E-cells and isolation of sperm cells/male fraction prior to DNA profiling. The differential extraction procedure operates based on the packaging of DNA in these two types of cells. The E cells are first lysed by sodium dodecyl sulfate (SDS) and Proteinase K which leaves the sperm cells intact. The mixture is centrifuged leaving E-cell DNA in the supernatant and sperm cells in the pellet. After several wash steps to remove residual E cell DNA, the sperm fraction is then subjected to lysis using SDS, proteinase K, and dithiothreitol (DTT). DTT reduces the disulfide bonds present in the sperm nucleus, thereby releasing sperm cell DNA. The traditional Gill method of differential extraction, while proven to be highly effective in providing two separate fractions for a simplified interpretation of profiles, is a labor intensive and time-consuming process, requiring approximately six hours of an analyst’s concentration. In a casework scenario where an evidence sample is of a higher E cell concentration compared to sperm cells, it is inevitable to obtain mixture profiles that becomes more difficult to interpret. To mitigate carryover from the female fraction, the sperm cell fraction is usually subjected to multiple wash steps. Furthermore, the resulting fractions must be subjected to additional pre-PCR DNA purification procedures to remove PCR inhibitors such as SDS and Proteinase K which result in varying degrees on DNA loss. Progress has been made over the years to introduce methods that allow for PCR-ready lysates without additional purification steps, often referred to as direct lysis methods. However, none have been proven to be viable options for use in sexual assault samples. Our laboratory has developed a novel differential extraction procedure that is not only time-efficient and less laborious but also utilizes a direct-lysis procedure requiring no further pre-PCR purification for most samples. The novel procedure uses ZyGEM, which contains the thermophilic EA1 protease proven to effectively digest biological samples and produce PCR-ready lysates suitable for downstream nucleic acid amplification, thereby minimizing DNA loss. The procedure uses a multi-enzymatic approach and utilizes the different optimal activity temperatures of the enzymes to perform most of the process in a DNA extraction lab thermocycler, requiring only a single centrifugation for the usual separation of the E-cell fraction and no subsequent washing steps for the sperm cell fraction. It has the potential to be a rapid, robust procedure that can be easily implemented in any forensic laboratory. This thesis will describe the procedure and report progress in the procedure optimization. / 2019-10-24T00:00:00Z

No description available.

No description available.

M.S.

M.S.

Supervised approaches, especially those employing deep neural networks, have showcased impressive performance, relying on a significant volume of manual annotations. However, their effectiveness encounters challenges when attempting to generalize to new languages, domains, or types, particularly in the absence of sufficient annotations. Current methods fall short in effectively addressing information extraction (IE) under limited supervision. In this dissertation, we approach information extraction with limited supervision from three perspectives. Firstly, we refine the previous classification-based extraction paradigm by introducing a query-and-extract framework, which uses target information as natural language queries to extract candidate information from the input text. Additionally, we leverage the excellent generation capability of large language models (LLMs) to produce high-quality annotation data, enriching IE semantics within limited annotation data. We also utilize LLMs' instruction-following capability to iteratively refine and optimize solutions through a debating process. Beyond text-only IE, we define a new multimodal IE task that links an entity mention within heterogeneous information sources to a knowledge base with limited annotation data. We demonstrate that excellent multimodal IE performance can be achieved, even with limited annotation data, by leveraging monomodal external information. These combined efforts aim to make optimal use of limited knowledge, ensuring more robust and generalizable solutions. / Doctor of Philosophy / This dissertation explores the development of information extraction (IE) algorithms and systems that work effectively with limited supervision. Information extraction is a complex and challenging task that involves extracting structured data from plain text. Traditional IE systems are often tailored to specific tasks and domains where ample annotated data is available, limiting their ability to adapt to new domains. This research focuses on developing IE systems that can generalize to new domains with limited supervision, reducing the reliance on extensive annotations. The proposed solutions demonstrate the potential to transfer knowledge from existing annotations to new tasks and domains, emphasizing the importance of learning from limited data and improving knowledge transfer to previously unknown domains.

Any operation that will effect a saving in heat is vitally important to industry. A principle which has reseived increasing attention during the past few years from those who are concerned with the separation of close boiling mixtures is extractive distillation. The selective action of a solvent, which is relatively non-volatile compared to the components of a mixture to be separated, is utilized. The solvent and extract are then separated by distillation, and part of the extract is recycled as reflux. A spiral-packed column for use in extractive distillation was designed and constructed. The extraction assembly consists of a single helicoid flight with a pitch of two, wound around and welded to a standard two-inch pipe, and machined to a “light driving fit” within a five-foot length of standard five-inch heavy duty pipe which had been bored to a diameter of five inches. The five-foot length of annular space between the two pipes is thus converted into a 27.9 foot spiral column of 1.175 by 1.75 inches cross-section. The principle of design involved is an attempt to obtain the effect of a 28-foot tower in one compact five-foot column. The feed from overhead tanks enters at a point midway in the column through a dispersion device. The solvent enters at the top of the column, and is saturated with a recycle of the raffinate product. A calculated amount of extract product is recycled as reflux near the base of the column. The raffinate product leaves the column at the top, and the solvent saturated with extract flows from the bottom. The equipment was tested with the system methylcyclohexane-petroleum naphtha-aniline at 25°C. ± 1°C. Saturation of equilibrium data with the corresponding refractive indices were determined. Rates of flow and reflux ratios were calculated by the material balance of Varteressian and Fenake which is based on the properties of the triangular diagram. From a 50-50 per cent mixture by weight on a solvent-free basis of methylcyclohexane and petroleum naphtha, a product containing 63 per cent by weight on a solvent-free basis of petroleum naphtha at a reflux ratio of 2.69, and a product containing 74 per cent by weight on a solvent-free basis of petroleum naphtha using a reflux of 6.5 can be obtained. / M.S.

Call number: LD2668 .T4 1962 G47

ITC/USA 2005 Conference Proceedings / The Forty-First Annual International Telemetering Conference and Technical Exhibition / October 24-27, 2005 / Riviera Hotel & Convention Center, Las Vegas, Nevada / Often it is necessary to retrieve segments of video with certain characteristics, or features, from a large archive of footage. This paper discusses how image processing algorithms can be used to automatically create a relational database, which indexes the video archive. This feature extraction can be performed either upon acquisition or in post processing. The database can then be queried to quickly locate and recover video segments with certain specified key features

No description available.