<p>In this work, I develop computational and visual tools for analyzing and manipulating the backbone of macromolecules, and I demonstrate that these tools support building better structures than currently done. These visualization and analysis tools belong to an "Intelligence Amplification" (IA) tradition (rather than complete Artificial Intelligence (AI) automation), empowering users to improve structures.</p><p>Proteins and nucleic acids are among the most important molecules in biology, mediating the majority of biochemical processes that comprise a living organism. Therefore, these macromolecules are important targets, both for basic research to improve understanding of how life works, and for medical research as possible drug targets. </p><p>The function of these macromolecules is largely determined by their 3D structure. Although these macromolecules are chemically fairly simple, made up of linear sequences of a few possible subunits, they physically fold into complex, compact structures. Overall, structural biology aims to determine the general relationship between sequence and structure of these macromolecules, through determination of the positions of the atoms within individual macromolecules. </p><p>Because it is currently impossible to directly see the position of atoms in a molecule, all structural determination techniques, including X-ray crystallography, NMR, and homology modeling, result in an interpreted <italic>model</italic> of a structure. Nearly all of these models contain mistakes, in which atoms are fit in incorrect or impossible positions. These mistakes, especially at a functionally-important location in a structure, can mislead both basic and medical research, making it critical for structural biologists to build the highest quality models possible. </p><p>This document details how my dissertation work enables the building of better macromolecular structure models. This work follows an iterative development cycle, where visual analysis of models spurs development of better tools, which in turn improves the analysis. First, I describe how my analysis of protein loops from X-ray crystal structures reveals that the traditional definition of loop endpoints is too restrictive. Second, I create a protein backbone analysis and modeling tool, using a new peptide-centric division system. I show how this tool makes it easier to study protein loops, and also how it improves an algorithm for calculating core protein models from NMR residual dipolar coupling (RDC) data. Third, I describe how 3D visualization of RDCs in their structural context improves understanding of RDCs and validates NMR models in a novel way. Fourth, I describe how local quality analysis can diagnose problems in homology models. Fifth, I demonstrate that local quality analysis can be successfully used in conjunction with model rebuilding software to correct errors in low resolution structures. The various tools and software packages I created during the course of my work are freely available and have already made a positive impact on structures being generated by the community.</p><p>Archive versions of several of these software packages (JiffiLoop, RDCvis, and KiNG) should be included with this document; current versions can be downloaded from http://kinemage.biochem.duke.edu.</p> / Dissertation

Dissertation (Ph.D.)--Auburn University,2006. / Abstract. Vita. Includes bibliographic references (p.110-113).

Linearity enhancement, especially within the front end of a wireless receiver IC design, is highly desirable since it allows the front-end to withstand strong interferers from co-existing communication standards or other wireless radiators. We propose an interferer suppression method based on feed-forward cancellation that uses an injectionlocked oscillator (ILO) to extract the interferer from the incident spectrum. The technique is expected to be useful in environments where a strong narrowband interferer appears along with a wideband desired signal, such as ultra-wideband (UWB) and emerging cognitive-radio applications. The ILO is further embedded within a phase-locked loop which provides several advantages including ILO center frequency self tuning and automatic phase alignment between the main signal path and the auxiliary path. An IC that uses this approach is implemented in a UMC 0.18[mu]m RFCMOS process. In measurement, the chip demonstrates 20dB suppression for phase and frequency modulated interferers while maintaining around 18dB power gain and noise figure below 5dB, measured with an off-chip balun for the desired signal. Techniques for canceling amplitude modulated interferers, though not included in the integrated circuit, were also demonstrated with an off chip amplitude control loop. Over 20dB rejection was obtained with AM interferers with properly scaled envelop signal applied to the ILO bias port. A second LNA was connected in cascade with the system to emulate the input stage of a down-conversion mixer and the cascaded P1dB was improved over 16dB with cancellation on. Gain compression above 13dB was also observed when auxiliary path was disabled, at the same input level as the P1dB with cancellation applied. / text

This project focused on two aspects of magnetite thin films. The first was to find optimum parameters and conditions for deposition of stoichiometeric Magnetite films using pulsed laser deposition (PLD). The second aspect was the characterization of the magnetic and electrical properties in order to broaden the spectrum of understanding of PLD Magnetite films. These properties were also investigated in terms of the substrates on which the films were deposited. Discussed in this thesis are deposition parameters, structural characteristics, magnetic and electrical characteristics of the films in terms of different substrates and film thicknesses. The discussion consists of structural parameters obtained using X- ray diffraction (XRD), Scanning electron microscopy (SEM), Energy dispersive spectroscopy (EDS), and electric properties such as resistance as a function of temperature and voltage dependence on the applied current. The magnetic properties measured were the magneto-resistance, M-H hysteresis loop, and magnetization as a function of temperature. The results obtained are then compared with pre-existing literature data. It will be shown that there is an impurity phase that may be seen when magnetite films are deposited on Sillicon dioxide substrates.

The Phase Locked Loops (PLLs) are widely used in contemporary electronic systems for frequency synthesis, clock recovery, frequency multiplication and other purposes. Because of continuous increase in operating frequency of clocking systems the requirements on the clock spectral purity and low jitter became very demanding and are one of major designers' concerns.Frequency synthesizers used in microprocessors are integrated on the same substrate as the rest of the circuit and thus suffer from a substantial switching noise injected into global supply and ground busses. Usually when the reference signal comes from a crystal oscillator, VCO becomes a main source of phase noise. A designer of VCO needs to determine the best circuit structure by considering different prototypical perturbations scenarios and predicting the worst case and jitter response when the perturbation signals are switched on and off. Therefore the time efficient estimation of the jitter effects resulting from many shapes, frequencies and phases of perturbation is critical.The presented dissertation demonstrates simulation methodology for rapid estimation of jitter in oscillators, particularly in VCOs, caused by perturbation sources such as power supply and substrate couplings. The methodology is also extended to these types of PLLs in which the VCO instability is a main contributor to the output timing jitter.Simulation of oscillatory circuits is strongly effected by the round-off errors. Special technique was developed to eliminate these effects.The technique is applicable for predicting timing non-idealities for arbitrary perturbation shapes, frequencies and phases. Different jitter metrics can be easily extracted for all important perturbation scenarios.The methodology utilizes the new concept of the transient multi-cycle Voltage Impulse Sensitivity Function (VISF), which has been developed in the dissertation. It contains information about sensitivity of oscillator to noise injection and also allows for efficient prediction of the transient effects caused by switching on and off the perturbation sources. The methodology offers efficiency and great simplicity of use, which frees designers from complicated, time consuming analysis of data generated by a simulator. The very involved postprocessing of simulation data can be fully automated.

An understanding of the blockage potential for an oil dominated system is an important step in moving from hydrate prevention to hydrate management. To better understand this problem a series of experiments were performed by varying the water cut, fluid velocity, and gas-liquid volume fraction using the ExxonMobil (XoM) flow loop in Houston, Texas, USA. The XoM large loop is a three pass, four inch internal diameter flow loop with a sliding vane pump capable of generating liquid velocities of up to 4 m/s. The systems that were studied include a range of water cuts from 5%-50% in a light crude oil (Conroe crude) and a gas phase of either pure methane for sI or 75% methane and 25% ethane which has sII as the thermodynamically stable phase. The results are compared with the hydrate plug prediction tool, CSMHyK, integrated into the multiphase flow simulator OLGA5®. The comparison between the model and the flow loop results serve as a basis for improving hydrate formation and plug prediction. In addition, the experimental variables that promote plug formation in the flow loop and how these may translate into the field are discussed.

Frequency synthesizers are essential components for modern wireless and wireline communication systems as they provide the local oscillator signal required to transmit and receive data at very high rates. They are also vital for computing devices and microcontrollers as they generate the clocks required to run all the digital circuitry responsible for the high speed computations. Data rates and clocking speeds are continuously increasing to accommodate for the ever growing demand on data and computational power. This places stringent requirements on the performance metrics of frequency synthesizers. They are required to run at higher speeds, cover a wide range of frequencies, provide a low jitter/phase noise output and consume minimum power and area. In this work, we present new techniques and architectures for implementing high speed frequency synthesizers which fulfill the aforementioned requirements. We propose a new architecture and design approach for the realization of wideband millimeter-wave frequency synthesizers. This architecture uses two-step multi-order harmonic generation of a low frequency phase-locked signal to generate wideband mm-wave frequencies. A prototype of the proposed system is designed and fabricated in 90nm Complementary Metal Oxide Semiconductor (CMOS) technology. Measurement results demonstrated that a very wide tuning range of 5 to 32 GHz can be achieved, which is costly to implement using conventional techniques. Moreover the power consumption per octave resembles that of state-of-the art reports. Next, we propose the N-Push cyclic coupled ring oscillator (CCRO) architecture to implement two high performance oscillators: (1) a wideband N-Push/M-Push CCRO operating from 3.16-12.8GHz implemented by two harmonic generation operations using the availability of different phases from the CCRO, and (2) a 13-25GHz millimeter-wave N-Push CCRO with a low phase noise performance of -118dBc/Hz at 10MHz. The proposed oscillators achieve low phase noise with higher FOM than state of the art work. Finally, we present some improvement techniques applied to the performance of phase locked loops (PLLs). We present an adaptive low pass filtering technique which can reduce the reference spur of integer-N charge-pump based PLLs by around 20dB while maintaining the settling time of the original PLL. Another PLL is presented, which features very low power consumption targeting the Medical Implantable Communication Standard. It operates at 402-405 MHz while consuming 600microW from a 1V supply.

The performance of the High voltage dc systems is dependent on the stiffness of the ac bus, it is connected to. With the traditional synchronous reference frame-phase locked loops (SRF-PLL), voltage source converters (VSC) systems with large PLL gains, connected to weak ac networks are shown to be prone to instabilities, when subject to disturbances. In this thesis a new Adaptive PLL is designed with a pre-filter topology which extracts the fundamental positive sequence component of the input voltage, to be fed into the SRF-PLL for tracking of its phase angle. Compared with other traditional PLL topologies, this Adaptive PLL shows superior immunity to voltage distortions, and also has a faster dynamic performance. The thesis presents a comparative analysis of the performance of the traditional SRF-PLL with the Adaptive PLL in a VSC control system, and its impact on stability for VSCs connected to weak ac systems (up to SCR=1.3).

Research suggests that the cerebellum is involved in cognition, but its exact role is unclear. The efficiency theory posits that the cerebellum supports processing speed. Other researchers argue that the cerebellum is functionally heterogeneous, and damage to lobes of the cerebellum causes selective loss of cognitive functions. This study sought to determine whether selective impairment in motor, verbal fluency, or processing speed occurred depending on the lobe of the cerebellum that was lesioned. Lesion mapping was used to measure lesion size and volumetric methods were used to measure atrophy in 25 adult survivors of cerebellar tumors. Participants had too a high degree of heterogeneous cerebellar lesions and accompanying atrophy to explore specialization. However, total cerebellar atrophy negatively impacted written and oral processing speed to a greater degree than total cerebellar lesion size. Younger ages at diagnosis and radiation therapy were associated with greater cerebellar atrophy.

SMA belongs to a class of so-called “smart materials” which possess properties that can be controlled by application of various types of stimuli – stress, temperature, electric field or magnetic field. In particular, SMA is a smart material which undergoes a temperature- or stress-dependent phase transformation giving it the property of remembering its original shape. Once deformed (up to a certain recoverable strain), SMA returns to its original shape upon heating. In this thesis, a study of SMA models and techniques to improve the performance of SMA actuators was carried out. In general, an SMA model is required for 3 main purposes: simulation, analysis and for model-based hysteresis compensation. In this work, the reduced-order form of John Shaw’s partial-differential equation model is chosen for implementation and simulation. The reduced-order form is used because its simpler structure makes it more useful for real-time control applications. The parameters were estimated for the John Shaw model followed by its implementation in MATLAB. From the view of control applications, a limitation of the John Shaw model is the inability to reproduce the so-called ‘minor loop behavior’ which is observed when the material is subject to cycling resulting in incomplete phase transformations. Modeling minor loop behavior is particularly important in closed-loop strain (or position) control applications since achieving a specific target strain between the two (load-dependent) extremes requires partial phase transformation. Herein, the governing equations are modified to include minor loop behavior. This behavior was tested using damped signals which would be expected to trigger minor loops in the actual SMA and reasonable match is observed from the simulations. The use of SMA actuators is limited by the relatively slow response time compared to other smart materials. The conventional current saturation (CS) scheme limits the maximum current into the wire at the manufacturer-specified safe current values in order to protect the wire from damage due to overheating. However, this is a conservative limit on the maximum current and hence, the response is artificially slowed. In order to improve the response time, a model-based temperature saturation (MBTS) scheme was developed, in which current is saturated based on model-predicted temperature. The MBTS scheme allows much higher currents to be applied to the wire, while ensuring that the wire is not damaged. Based on simulations using the reduced-order John Shaw model, it is observed that better tracking occurs using the MBTS scheme in the actuation scheme as compared to the CS scheme.