|
|
|
|
|
|||
|
|||
|
|||
|
|
||
|
|||
|
|
 |
|||
|
  |
||||
| General Section People Events Program Details |
 |
Predoctoral Trainee
The recent availability of complete genomes has allowed researchers to develop techniques for identifying specific genetic features within a genome. However, novel computational techniques which are both accurate and computationally feasible are needed to help identify these features. Since what distinguishes a gene from an open reading frame which is not a gene is unknown, it is my belief that machine learning techniques must be used to model genomic regulatory elements and trained to learn how classified examples fit into the model. These systems can then be used to make predictions about unclassified examples. I will explore new models and variations on existing models and applying machine learning algorithms to them. In addition to trying to improve the models that represent specific genomic regions, I will explore the advantages of combining these models, perhaps in a Bayesian net framework, in order to expressly make use of all available information. Some of the genomic regions that I am interested in identifying using novel computational techniques include the gene regions themselves, intergenic regions, operons, promoters and terminators, and transcription factor binding sites. An ultimate goal is understanding and identifying the specific function of each gene in a genome as well as the statistical properties of a genome. CVK Noto CV (.pdf format; http://www.cs.wisc.edu/~noto/cv/) Publications |
|||
|
|
|
|||
| CIBM Home | UW Home | |||||
| Feedback,
questions or accessibility issues. |
|||||
