News and Announcements

Do you speak MATLAB? Over one million people around the world do and now you can too. Free! The University of Tennessee is accelerating its pace in science and engineering by providing MATLAB, Simulink and 54 companion products free to all faculty, staff and students. In July 2008, a UT system-wide license agreement with The MathWorks was acquired and the MATLAB family of products is ready for distribution. Our MATLAB rollout event is August 27-28 in the University Center Shiloh room and everyone is invited. There is no pre-registration. The keynote speaker is Cleve Moler, the author of the original MATLAB code and co-founder of The MathWorks. After the keynote session, MathWorks’ engineers will provide training on MATLAB, Simulink and various companion products. [More...]

WHEN: Wednesday, August 13th at 12:30 p.m. in CL 211

TOPIC: Addressing Challenges in a Graph-Based Analysis of High-Throughput Biological Data

ADVISOR: Michael Langston

Abstract
Graph-based methods used in the analysis of DNA microarray technology can be powerful tools in the elucidation of biological relationships. As these methods are developed and applied to various types of data, challenges arise that test the limits of current algorithms. These challenges arise in all phases of data analysis: data normalization, modeling biological networks, and interpreting results.

Spectral graph theory methods are investigated as means of threshold selection, a key step in constructing graphical models of biological data. Also important in constructing graphs is the selection of an appropriate gene-gene similarity metric, and a sort-of compendium of similarity profiles for some biological data sets is present, along with a similarity thresholding method based upon structural properties of random graphs.

The identification of altered relationships between two or more conditions is a goal of many microarray gene expression studies. Clique-based methods can identify sets of coexpressed genes within each group, but additional computational methods are required to uncover the differential relationships and sets of genes changing together between groups. Differential filters are reviewed to highlight those changing interactions and sets of changing genes. The effect of various normalization methods on these differential results is also studied.

Finally, how methods commonly used in the analysis of gene expression data can be used to investigate relationships in noisy and incomplete historical ecosystem data is explored.

414 Ferris Hall
Wednesday, August 6, 2008 @ 9:00 A.M.
Design and Development of an APD Algorithm Development Board for Positron Emission Tomography
Rakesh Mallem, Master’s Candidate
Dr. S. K. Islam, Major Professor

Abstract
Using the PET scanner, three dimensional images of the human body with sufficient detail can be viewed which help physicians to visualize both normal metabolic functions and discover the chemical processes underlying physical abnormalities. Commercial PET scanners employ Photo Multiplier Tubes to detect the anti-matter annihilation photons and amplify the signals to a suitable level for digital sampling. Photomultiplier tubes provide extremely high sensitivity and exceptionally low noise compared to other photosensitive devices currently used to detect radiant energy in the ultraviolet, visible and near infrared regions. A combined magnetic resonance positron emission tomography (MR-PET) modality would require a solid-state photo detector due to the known gain/timing variation of PMTs with variable magnetic field. PET detector block designs have been described and implemented in the literature using APD photo detectors at moderate values of gain.

The APD Algorithm Development Board is basically a signal processing board which receives the integrated APD analog signals and outputs a digital event packet composing of position and timing data for each detected photon. These digital event packets are digitally transmitted to a downstream module for comparison with opposing detectors to detect the coincidence photons fundamental to PET. The main functions are to process analog signals from the APDs to determine if an energy qualified gamma ray event has been detected, localize the crystal position and time of the event, and transmit the event information to the control interface, en route to a coincidence processor.

DATE: Friday, July 25, 2008

TIME: 10:00am

PLACE: Claxton 202

TITLE: Hardware Sensors Monitoring with PAPI

ABSTRACT
Modern computer hardware have sensors to monitor the health and status of different computer components. Most of these sensors can be read and monitored, if you know how to do it. But, even if you can read them, you still need to know how to properly interpret (massage) the information just read. In this presentation, I will talk about the details of the work I did to enable the PAPI framework to conveniently access these sensors. I will also show results of some thermal profiling with the goal of understanding better how the computer resources are being used, and particularly, how are we using power.

COMMITTEE

• Dr. Jack Dongarra (Represented by Dr. Shirley Moore)
• Dr. Xiaorui Wang
• Dr. Yuanshun Dai (Represented by Dr. Stan Tomov)