Eric Williams

edwst7@cs.pitt.edu

RESEARCH INTERESTS

Machine learning, data mining, computational scientific discovery, computational biology, biomedical informatics, protein x-ray crystallography

EDUCATION

University of Pittsburgh, Pittsburgh, PA

Ph.D. in Intelligent Systems, 2012 (estimated)

Advisor: Dr. John Rosenberg

Comprehensive Exam passed December 16, 2008

 

University of Pittsburgh, Pittsburgh, PA

M.S. in Intelligent Systems, 2007

Project: “Aggregating Validation Measures to Remove Potentially Problematic Protein Structure Models from Large Data Sets”

Advisor: Dr. Vanathi Gopalakrishnan

Project Defense passed November 16, 2007

 

University of Pittsburgh, Pittsburgh, PA

M.S. in Electrical Engineering, 2001

 

University of Pittsburgh, Pittsburgh, PA

B.S. in Computer Engineering, 1999

GPA: 3.23

AWARDS

University of Pittsburgh Innovator Award, 2006

For work on improving RL, an inductive rule learning tool.

PUBLICATIONS AND PAPERS

Williams, E., Rosenberg, J.

“Sweeping Sins Under the Rug (and Other Statistical Trends in B-factors Reported in the PDB)”

To be submitted to Proteins: Structure, Function, and Bioinformatics. In Preparation

 

Williams, E., Rosenberg, J.

“Sweeping Sins Under the Rug (and Other Statistical Trends in B-factors Reported in the PDB)”

Poster presented at Science2010: Transformations, Pittsburgh, PA. 2010

 

Hennessey, D., Williams, E., Rosenberg, J.

Relationships between pH, estimated net charge and parameters of protein crystallization and outcomes”

To be submitted to Protein Science. In Preparation

 

Fasnacht, M., Hodor, P., Williams, E., Buchanan, B., Caruana, R., Rosenberg, J.

A Method for Automatically Finding Structural Motifs in Proteins”

To be submitted to Proteins: Structure, Function, and Bioinformatics. In Preparation

 

Williams, E., Hodor, P., Caruana, R., Fasnacht, M., Hennessey, D., Buchanan, B. and Rosenberg, J.

Mass Assessment of Protein Structure Models”

To be submitted to Protein Science. In Preparation

 

Ranganathan, S., Williams, E., Ganchev, P., Gopalakrishnan, V., Lacomis, D., Urbinelli, L., et al.

Proteomic profiling of cerebrospinal fluid identifies biomarkers for amyotrophic lateral sclerosis”

Journal of Neurochemistry. 95(5): p. 1461-1471. 2005

 

Lustgarten, J., Williams, E.

Preliminary Studies Comparing MALDI- and SELDI-TOF Mass Spectral Profiling of Cerebrospinal Fluid from Patients With Amytrophic Lateral Sclerosis”

Poster presented at Science2004: Beyond Boundaries, Pittsburgh, PA. 2004

 

Gopalakrishnan, V., Williams, E., Ranganathan, S., Bowser, R., Cudkowic, M. E., Novelli, M., et al.

Proteomic data mining challenges in identification of disease-specific biomarkers from variable resolution mass spectra”

Paper presented at the Proceedings of SIAM Bioinformatics Workshop, SIAM Data Mining 2004, Nashville, TN. 2004

 

RESEARCH EXPERIENCE

Statistical Analysis of B-factors and Related Disorder Parameters

PI: Dr. John Rosenberg

2009 – Present

We studied B-factors deposited in the Protein Data Bank (PDB). We hypothesized that B-factor statistics would be correlated with indicators of model quality, providing evidence of neglected or highly dispersed errors. Mean model B-factors were found to be correlated with data resolution. Deviations from resolution-based mean value expectations and the degree of B-factor variation are associated with degradations in several model quality indicators.

 

Prediction of Degree of Protein Crystallization Success

PI: Dr. John Rosenberg

2004 – Present

We developed a set of rules for predicting and optimizing the likelihood that particular crystallization conditions will result in a crystal of a particular protein that is usable for x-ray diffraction and will yield a highly accurate structure model.

 

Automated Detection of Structural Motif Patterns

PI: Dr. John Rosenberg

2004 – Present

We developed a method to automatically extract structural motifs in proteins used to find structurally equivalent sets of motifs that can be used for study of the underlying amino acid sequence.

 

Automated Estimation of Expert Confidence in Protein Structure Models

PI: Dr. John Rosenberg

2004 – Present

We developed a model evaluation measure, called Omega, formed from an aggregation of established and broadly available measures of confidence in structure model accuracy.

 

Classification and Identification of Potential Biomarkers from Proteomics Mass Spectra

PI: Dr. Vanathi Gopalakrishnan

2003 – 2004

We applied machine learning and data mining methods to the problem of discovering discriminative biomarkers for and classifying proteomic mass spectra. Particular emphasis was placed on inductive rule learning.

 

Inductive Rule Learning

PI: Dr. Bruce Buchanan

2002 – 2003

We refined the RL4 rule learning program and applied it to problems in biomedical informatics.

Parkinson’s Disease

PI: Dr. J.T. Cain

2000 – 2001

We began development of a tool to facilitate evaluation of the severity of Parkinsonian tremors using a tablet PC.

 

TEACHING EXPERIENCE

University of Pittsburgh, Pittsburgh, PA

Teaching Assistant for Various Computer Engineering Undergraduate Courses

1999 - 2001

Administered all homework and lab report grades, guided recitation sessions, held office hours, met with students upon request.

 

REFERENCES

John Rosenberg, Ph.D.

(412) 624-4636

314 Clapp Hall

4249 Fifth Avenue

Pittsburgh, PA 15260

 

Gregory Cooper, M.D., Ph.D.

(412) 647-7113

Parkvale Building M-191

200 Meyran Avenue

Pittsburgh, PA 15213

 

Shyam Visweswaran, M.D., Ph.D.

(412) 648 – 6753

Parkvale Building 156
200 Meyran Avenue

Pittsburgh, PA 15213