Recently biological sequence databases have grown much faster than the ability of researchers to annotate such sequences. Techniques for computational analysis of biological sequences have grown in importance as researchers attempt to understand some features of these sequences. These features are computationally predicted through correlation with the presence of signals, which are measurable characteristics of a sequence correlated with the sequence feature. This study describes a general methodology for combining the information from signals in order to predict the presence of sequence feature. The methodology is based on Genetic Algorithm, which are a class of computational techniques that borrow concepts from Genetics in order to solve complex problems. The problems of prokaryotic start site prediction and prediction of RNA editing in order to demonstrate these this methodology.

Library of Congress Subject Headings

Nucleotide sequence--Databases; Amino acid sequence--Databases; Genes--Analysis--Data processing; Proteins--Analysis--Data processing; Genetic algorithms

Publication Date


Document Type


Department, Program, or Center

Biomedical Sciences (CHST)


Skuse, Gary

Advisor/Committee Member

Frederick, G. Thomas


Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: QP625.N89 T46 2006


RIT – Main Campus