Project Summary
A grant is awarded to Stony Brook University to generalize and expand computational approaches for designing gene sequences into a coherent set of tools for three areas of molecular biology: 1) sequence design for optimized gene expression; 2) signal identification and location; and 3) self-assembly. This project will improve our understanding of several sequence phenomena that occur in genomes, such as codon-pair bias, how RNA secondary structure and tRNA usage modulates gene expression, and how to use these results to develop gene design software to optimize gene expression.
A new approach to identify the locations of critical DNA or RNA sequence signals (e.g. binding domains, secondary structures, and mRNAi targets) couples large-scale synthesis with sophisticated designs based on ideas from combinatorial group testing. This project will improve the efficiency and generality of this procedure, and will develop new homology-based software tools to identify candidate signals prior to employing combinatorial search approaches. New array-based oligo synthesis technologies provide access to tens or even hundreds of thousands of short custom sequences, but even greater power will become enabled if oligonucleotides can be designed to self-assemble into chromosome-sized sequences. Recombination systems in yeast provide the platform to conduct such experiments. This project will develop algorithms for the design of large-scale, self-assembling sequences, to best exploit new array-based synthesis technologies, and test how they work in practice.
This collaboration between computational and life sciences researchers will advance both disciplines, through new results in combinatorial algorithm design and discrete optimization as well as fundamental discoveries regarding gene expression, signal detection, and self-assembly. Broader impacts of this project will result in software and experimental tools to advance broad areas of molecular biology, including the design of vaccines. Software and results of this project will be available from the website http://www.cs.stonybrook.edu/~skiena/dna.
Project Summary
Vaccines and antibiotics have proven to be tremendously important agents against viral and bacterial infections, respectively. However, current vaccines are ineffective against certain viruses (e.g. rhinovirus) and potentially dangerous against others (e.g. smallpox). Further, the evolution of antibiotic-resistant strains of bacteria has become a critical problem in treating diseases such as tuberculosis.
We propose to exploit the redundancy inherent in the genetic code to create both safer and more effective vaccines and an improved class of anti-bacteriological agents. We apply design and synthesis techniques to replace viral genes so they code for identical proteins as in the wildtype virus, but in different ways. For the first application, we seek to weaken viral strains by introducing mutations that alter translational efficiency and RNA secondary structure without affecting protein coding so as to create better vaccine candidates. For the second application, we seek to strengthen bacteriophages (viruses which attack bacteria) by eliminating important restriction sites so as to improve their ability to combat pathogenic infections.
The PI (Skiena) is a computer scientist who has worked extensively on theoretical sequence design problems which seek optimized DNA sequences which code for a particular protein while simultaneously accomplishing some other objective. The co-PI (Wimmer) is a microbiologist who attracted worldwide attention in July 2002 by synthesizing poliovirus completely from off-to-shelf components. But once you can synthesize an existing genome from scratch, you can do the same for new and better designs as well. We will collaborate on a series of computational and wet lab experiments using poliovirus to test and develop this vision.
Broader impacts of the proposed research include benefits to society
through developing new types of medical therapeutics, and its
impact on interdisciplinary bioinformatics education.
See our image of the RNA secondary structure of synthetic gene coding for a protein (accession number AB010831) designed to maximize RNA secondary structure. It is 2.25 times as stable as the wildtype encoding of the protein.