Jonathan Carlson

I'm a fourth-year graduate student in the Computational Molecular Biology program and the CS Department at the University of Washington in Seattle, WA. I spend much of my time at Microsoft Research, working with my advisor David Heckerman on modeling HIV adaptation. Although I'm learning to be a Husky, my heart will always be with the Big Green of Dartmouth, where I met my beautiful wife.

Index

• Curriculum Vitae
• Research
• Publications
• Course Projects and Reports
• Pictures
• Useful tricks

Research

I am interested in the interface between computer science and biology. While I'm in the computer science department, my goal is to further my undergrad training in biology as much as possible, for computational biology without the biology is little more than a theoretical exercise. That said, my core interests really lie in the computational aspects, hence my decision to pursue a CS PhD.

HIV Adaptation

HIV's tremendous capacity for rapid adaptation is the root of drug resistance, vaccine failure, and immune failure (AIDS). Our work is on computational models for identifying sources of selection pressure and the specific adapations that arise as a result of that selection pressure. For example, when a patient is given an anti-viral drug, a new selection pressure is introduced to the virus. Eventually, resistance mutations are selected for that protect the virus from the drug.

Our primary focus is on how HIV adapts to the cellular immune response. We are working with a number of collaborators on this project, with the goal of informing new approaches to vaccine design. Our models allow us to identify the host genetic variations (HLA alleles) that correlate with specific HIV mutations. To do this in a statistically sound way requires that we account for the evolutionary history of the the virus, linkage disequilibrium among HLAs, and compensatory mutations that lead to a dense network of dependencies among the HIV codons.

This work is based at Microsoft Research under the direction of my advisor David Heckerman. The source code is available via an open source license (note that the current version of the open sourced code supports only pairwise correlations). The press announcement was received well and lead to some nice articles on David and MSR in Business Week and on NPR.

Robust Motif Finding

From the fall of 2003 to the spring of 2007, I worked with Prof. Bob Gross and Dr. Arijit Chakravarty on a motif finding program. Our first papers focused on identifying key bounds that yielded efficient computation of DNA motifs of arbitrary length and degeneracy under any objective function. BEAM, PRISM and SPACER (see pubs) were each focused on the identificiation of non-degenerate, degenerated, and biparatite motifs, respectively. We found that by using beam-search algorithms we were able to heuristically limit the search space in such a way that expressive consensus motifs can be learned. Although the consensus motif representation is not as expressive as the position weight matrix, we found that a good search over the full space of consensus motifs yields better results than heuristic searches over PWM space that are quite prone to local optima.

Most recently, we combined these three focused motif finders into an ensemble method. SCOPE runs BEAM, PRISM and SPACER, then uses a unified scoring metric to combine the results. Really all it's doing is running the three focused motif finders, each of which uses the same scoring metric, which accounts for the size of the search space. We found that searching for three classes of motifs with motif finders that are good at those classes results in an ensemble that significantly outperforms those methods, as well as a number of other approaches.

Virtify.ImagingLib

Over the summer of 2005 I worked at Virtify. I built a computer vision library for feature extraction of high throughput immunoflourescent images in the context of lead compound discovery.

Genomic Motif Distributions

My undergrad thesis was on genome-wide distributions of consensus motifs. I built a relational database of all 9mers in the upstream regions of A. thaliana (a plant) and used statistical measures of the motifs distributions with respect to the translation start sites to find transcription factor binding sites. The project was promising but never carried through to publication. We presented our results at TIGR's annual conference.

Top

Publications

Journal Publications

Phylogenies and computational immunology

• Jonathan M. Carlson and Zabrina L. Brumme. "HIV evolution in response to HLA-restricted CTL selection pressures: a population-based perspective." Microbes and Infection, (February 2008) doi:10.1016/j.micinf.2008.01.013.    [pdf]
• Christine M. Rousseau, Marcus G. Daniels, Jonathan M. Carlson, Carl Kadie, Haley Crawford, Andrew Prendergast, Philippa Matthews, Dana N. Raugi, Brandon S. Maust, Gerald H. Learn, David C. Nickle, Nicole Frahm, Christian Brander, Bruce D. Walker, Philip J. R. Goulder, Tanmoy Bhattacharya,, David E. Heckerman, Bette T. Korber, James I. Mullins (2008) "HLA Class-I Driven Evolution of Human Immunodeficiency Virus Type 1 Subtype C Proteome: Immune Escape and Viral Load." Journal of Virology, 82(13):6434-6446.
• Zabrina Brumme, Iris Tao, Sharon Szeto, Chanson Brumme, Jonathan M. Carlson, Dennison Chan, Carl Kadie, Nicole Frahm, Christian Brander, Bruce D. Walker, David Heckerman, P. Richard Harrigan (2008) "HLA-specific polymorphisms in HIV-1 Gag and their association with viral load in chronic untreated infection." AIDS, 22(11):1277-1286.
• Paul Goepfert, Wendy Lumm, Paul Farmer, Philippa Matthews, Andrew Prendergast, Jonathan M. Carlson, Cynthia A. Derdeyn, Jianming Tang, Richard A. Kaslow, Anju Bansal, Karina Yusim, David Heckerman, Joseph Mulenga, Susan Allen, Philip Goulder, Eric Hunter (May 2008) "Transmission of HIV-1 Gag Immune Escape Mutations is Associated with Reduced Viral Load in Linked Recipients." Journal of Experimental Medicine , 205(5):1009-1017.
• Jonathan M. Carlson, Carl Kadie, Simon Mallal, David Heckerman. "Leveraging hierarchical population structure in discrete association studies." PLoS One, 2(7):e591 (July 2007).
• Z.L. Brumme, C.J. Brumme, D. Heckerman, B.T. Korber, M. Daniels, J.M. Carlson, C. Kadie, T. Bhattacharya, C. Chui, J. Szinger, T. Mo, R.S. Hogg, J.S.G. Montaner, N. Frahm, C. Brander, B.D. Walker, P.R. Harrigan. "Evidence of differential HLA class I-mediated viral evolution in functional and accessory/regulatory genes of Human Immunodeficiency Virus (HIV) type-1." PLoS Pathogens, 3(7):e94 (July 2007). [In the news: NewScientist EurekAlert]
• T. Bhattacharya, M. Daniels, D. Heckerman, B. Foley, N. Frahm, C. Kadie, J.M. Carlson, K. Yusim, B. McMahon, B. Gashen, S. Mallal, J. Mullins, D. Nickle, J. Herbeck, C. Rousseau, G. Learn, T. Miura, C. Brander, B. Walker, B. Korber. "Founder effects in the assessment of HIV polymorphisms and HLA allele associations." Science, 315(5818):1583-1586 (March 2007). [commentary] [In the news: MS release CNet ]

Transcription factor binding site prediction

• A. Chakravarty, J.M. Carlson, R.S. Khetani and R.H. Gross "A novel ensemble learning method for de novo computational identification DNA binding sites." BMC Bioinformatics, 8:249 (July 2007).
• J.M. Carlson, A. Chakravarty, R.S. Khetani, C.E. DeZiel and R.H. Gross. "SCOPE: A web server for practical de novo motif discovery." Nucleic Acids Research, 35(Web Server issue):W259-264 (July 2007). [pdf]
• A. Chakravarty, J.M. Carlson, R.S. Khetani, C.E. DeZiel, R.H. Gross. "SPACER: Identification of cis-regulatory elements with non-contiguous critical residues." Bioinformatics, 23(8):1029-1031 (May 2007).
• J.M. Carlson*, A. Chakravarty*, R.S. Khetani and R.H. Gross "Bounded search for de novo identification of degenerate cis-regulatory elements." BMC Bioinformatics, 7(1):254, (May 2006).     [pdf]
• J.M. Carlson*, A. Chakravarty* and R.H. Gross. "BEAM: A beam search algorithm for the identification of cis-regulatory elements in groups of genes." Journal of Computational Biology, 13(3):686-701, (April 2006).     [pdf]

Theses

• J.M. Carlson. "Creation of a Relational Database for Identifying Functional DNA Sequence Motifs in A. thaliana and Other Genomes." Undergraduate Honors thesis, Dept. Biology, Dartmouth College (2003).     [pdf]

Top

Course Projects and Reports

CSE 557: Graphics (Winter 2006)

• Impressionist Project. Built an app that lets a user modify a digital photo to look like an impressionist painting. Worked with Rob Nash.
• Ray Tracer. Built a Whitted ray tracer that implements BSP tree acceleration, adaptive sampling and the Phong illumination model.

CSE 561: Networks (Fall 2005)

We built from the ground up a network that uses sound as its medium. By careful use of encoding schemes and reliable transport, we were able to achieve reliable communication at distances of up to 15 feet in noisy environments at 160 bits per second. We were able to maintain 50% medium usage under high load using our novel congestion control mechanism. We also learned a great deal about C# and DirectX sound.

• J Carlson, J Froehlich, L Ritter. " LoudNetworks." Fall, 2005.

CSE 527: Computational Molecular Biology (Fall 2004)

• J Carlson. " Critical comparison of motif models." Fall, 2004.

Top

Pictures

	National Parks Tour 2006	1 pictures	September 17, 2006
	Oregon Coast	19 pictures	August 06, 2006
	Mt. Rainier	72 pictures	July 22, 2006
	Seattle	10 pictures	July 17, 2006
	Rattlesnake Ridge	30 pictures	July 15, 2006
	Lake George 2006	9 pictures	June 09, 2006
	Ray Tracer	13 pictures	February 23, 2006
	Impressionist Project	7 pictures	January 25, 2006
	Houston Christmas 2005	28 pictures	December 25, 2005
	Mt. St. Helens	24 pictures	September 28, 2005
	Wedding	84 pictures	July 17, 2004

Top