Predicting nucleic acid hybridization and melting profiles.
Speaker: Michael Zuker , Dept of Mathematical Sciences, Rensselaer Polytechnic Institute
Date: April 3 2006
Time: 11:30AM to 1:00PM
Location: TOC Lab- 32-G575
Host: P Clote/ BC & B Berger/ MIT
Contact: Kathleen Dickey, 617 253 3037, kvdickey@mit.edu
Relevant URL: http://www-math.mit.edu/compbiosem/
Many applications in modern biotechnology require an accurate prediction of hybridization, partial hybridization or the competition between folding and hybridization of two single stranded nucleic acid molecules in solution. Current methods often employ simple two-state models to estimate melting temperatures for single stranded nucleic acids or for dimers. Similarly, database searching for likely hybridization sites of a DNA oligonucleotide or putative targets of a ncRNA (non-coding RNA) is often performed using existing tools such as BLAST.
We have undertaken a research program to develop more sophisticated and therefore, we hope, more accurate algorithms and software to predict not just melting temperatures, but entire melting profiles. In parallel, we have also developed a database search method that uses nearest neighbor free energy parameters as the basis for hybridization prediction.
The new UNAFold package implements, in part, the detailed hybridization model. It considers the competition between folding and hybridization, and also allows homo-dimer as well as hetero-dimer formation. Partition function calculations are performed for each single- or double-stranded molecular species and ensemble free energies as well as base pair probabilities are computed as a function of temperature. Ensemble free energies lead directly to heat capacity (Cp) predictions that can be compared directly with measurements from DSC (Differential Scanning Calorimetry) experiments. Similarly, computed probabilities, together with published extinction coefficients, lead to predictions of UV absorbance as a function of temperature.
In contrast, our database search algorithm, named FASTH, considers only the formation of a hetero-dimer between a relatively short oligonucleotide and a very long target, the database. It computes a hash table for the sequence database using a two letter, {puRine, pYrimidene}, alphabet, and quickly identifies regions of potential hybridization ranked by a crude score that is a sum of stacking free energies.
See other events that are part of Bioinformatics Seminar Series 2005/2006
See other events happening in April 2006
