Available Rotations

Our primary research projects fall into two camps: algorithm development and biological investigation. These are detailed below. Specific rotation openings and skills are listed first.

Spring 2005


Algorithm development
Required skills: programming, preferably in C, C++ or Java and scripting languages (e.g., Perl or Python).

  • Improving the accuracy of phylogenetic tree construction, multiple sequence alignment and comparative (homology) models:
    These tasks are all related, and we have several development projects in place that rotation students can join. This project involves method development and evaluation for three of the primary tasks in constructing a comparative model: (1) Template selection, (2) Constructing a pairwise alignment between the target (of unknown structure) and the template protein (whose structure is known), (3) Determining which region(s) of the target-template alignment are usable for model construction and which region(s) of the target must be flagged for either ab initio loop modeling or matching against a library of fragments of solved structures. We have preliminary results on two novel approaches (Simultaneous Alignment and Tree Construction using Hidden Markov Models and TreeHMMs) from our lab with superior results to all other alignment methods, based on evaluation against structurally aligned pairs of proteins. One student can work on improvements to the SATCHMO algorithm, and the second student can work on improvements to TreeHMMs.

  • Predicting key functional residues in proteins.
    This project involves integration of protein structure information, multiple sequence alignment and phylogenetic tree topology to predict amino acids playing catalytic roles in proteins, participating in binding pockets, or showing evidence of diversifying selection.

    Biological investigation

  • These projects are performed in collaboration with biologists here at Berkeley, as well as several international collaborations (UK, Switzerland, and Asia) and involve protein structure-function prediction. Students interested in this type of rotation should have taken Bioe144 (Intro. to Protein Informatics) and received a grade of A- or better, or have other similar experience preparing them for this work.

    Other projects in our lab include the following:

    Computational Method Development

  • Clustering and alignment
  • Phylogenetic tree inference
  • Identifying key functional positions
  • Domain Identification
  • Constructing homology models
  • Inferring networks from noisy experimental data
  • Profile-profile scoring and alignment
  • Protein structure prediction and HMM methods in general

    Biological Investigation

  • Animal proteome analysis and functional annotation of animal genomes
  • Plant proteome analysis and functional annotation of plant genomes
  • Innate Immunity in Animals
  • Protein Superfamily Evolution

    User Interface and Webserver/Database Development

  • HMM library development
  • Graphical user interfaces for phylogenomic inference

    Student Research Opportunities

    These are available for qualified students. These are of two basic types. The first type of rotation involves algorithm (computational method) development for problems in molecular biology. Students interested in this type of rotation should have solid programming skills (i.e., C/C++/Java and Perl or Python). The second type of rotation involves using informatics methods to predict protein function or structure, typically in collaboration with biologists here on campus. Students interested in this latter type of rotation need some solid coursework in molecular biology, and should preferably have taken Bioengineering 144, Introduction to Protein Informatics.

    To apply, submit

  • a transcript of your grades at Berkeley (unofficial is acceptable),
  • the names and contact information for two references (faculty who know your work is best, TAs are acceptable), and
  • a brief description of your personal research interests, what kind of work you like to do, and your background in programming, mathematics, and biology.

    Compute Infrastructure

    Our compute infrastructure is being designed by Systems Architect John Lee.