Lab web servers
  • CASTp
  • CPalign
  • TheraBill
  • Facebook page
  • Myspace
  • Surrender Dorothy
  • Click Tribe
  • The Lesson Studio
  • The Long family
  • Joseph M. Dundas Ph.D.
    PostDoctoral Research Associate
    University of Illinois at Chicago
    Bioengineering/Bioinformatics M/C 563
    835 South Wolcott
    Chicago, IL 60612-7340
    Click picture to view personal photo page.
    Research projects

  • Topology independent protein structural alignment: Protein structure alignment is an indespensable tool used for many different studies in bioinformatics. Most structural alignment algorithms assume that the structural units of two similar proteins will align sequintially. This assumption is flawed, and as a result proteins with similar structure but with permutated sequence arrangement are often missed. We present a solution to the problem based on an approximation algorithm that finds a sequence-order indpendent structural alignment that is close to optimal.
  • Sequence order independent local surface alignment: The algorithm is based on a multi-objective optimization method described by Xian-Sun Zhang's group. Their method decomposes the structural alignment problem into two subproblems. First a linear programming subproblem to obtain the equivalence relation between residues and second a least squre subproblem for coordinate transformation. A bipartite matching algorithm is implemented in the linear programming subproblem while the singular value decomposition method is used for coordinate transormation.
  • Trajectories in Darwinian evolution: Beta-lactamase is an enzyme that confers resistance to beta-lactam antibiotics. Certain point mutations on the TEM-1 beta-lacatamase produce extended spectrum beta-lactamases (ESBLs). Unlike TEM-1, ESBLs are able to hydrolyze third generation cephalosporins, deactivating the molecule's antibacterial properties. The increasing frequency of ESBL-producing organisms may be a result of extensive beta-lactam antibiotics use, implying that developing new drugs capable of inhibiting ESBL-producing microbials may not be the best approach to deal with this problem. A better approach might be stopping the emergence of the ESBL. To do this, it's important to understand how TEM-1 evolves to ESBL. A specific ESBL (TEM*) has five point mutations that increase its resistance to cefotaxime approximately 100,000 fold. By assuming that evolution only acts to increase drug resistance, Weinreich et. al. showed that only 10 of the 120 possible mutational evolutionary trajectories from TEM-1 to TEM* are realizable. In their study, Minimum Inhibitory Concentration (MIC) was used as a measure of resistance, which assumes a changing spatiotemporal drug concentration for each trajectory. To test the validity of the increasing resistance assumption, we constructed the phylogenetic tree from a set of non-redundant beta-lactamases and predicted the ancestral sequences. The phylogenetic analysis showed instances of decreasing resistance within trajectories whose end results was higher resistant ESBL. We then developed a model that allows consideration of trajectories other than those that increase drug resistance at each evolutionary step. We defined the allele resistance as the response at a certain drug concentration on a dose-response curve derived from experimental MIC values treated as the inhibitory concentration. The resulting model allowed us to study the effect of drug concentration on the evolutionary trajectories from TEM-1 to TEM*. Our results indicate all trajectories are evolutionarily feasible at low drug concentrations. However, 30 trajectories are more probable at all drug concentrations.

  • Joe Dundas, Zheng Ouyang, Jeffery Tseng, Andrew Binkowski, Yaron Turpaz and Jie Liang. (2006) CASTp: comuter atlas of surface topography of proteins with structural and topographical mapping of functionally annotated residues. Nucleic Acids Research, 34:W116-118.
  • Joe Dundas, T.A. Binkowski, Bhaskar DasGupta and Jie Liang, Topology Independent Protein Structural Alignment. 7th Workshop on Algorithms in Bioinformatics, R. Giancarlo and S. Hannenhalli (Eds.), LNBI 4645, Springer-Verlag Berlin Heidelberg, pp. 171-182, September 2007.
  • Research interests
  • Structural computational biology
  • Protein structure alignment
  • Internet based bioinformatics applications
  • Data mining