Analysing Patterns of Tripeptides Using Statistical Approach and Neural Network Paradigm
Advances in Bioinformatics and Its Applications / Proceedings of the International Conference Nova Southeastern University, Fort Lauderdale, Florida, USA
Tripeptides, Codon, Classification, Neural Networks
The goal of this research was the investigation of the relationships of the distances measured between the different amino acids in the protein of 7,964 tripeptides from the Protein Data Base (PDB). As a first step into this investigation, we developed a program capable of calculating the types of two triangles based on the twelve distances measured between the different amino acids in the protein. The second objective was to use an unsupervised neural network to cluster tripeptides based on the same input data. The selected for this purpose Self-Organizing Maps network was successful in categorizing the data in close approximation to the results achieved by the program.
Conference Proceeding Title
Series in Mathematical Biology and Medicine: Volume 8 - Advances in Bioinformatics and Its Applications
978-981-256-148-0 (hardcover) 978-981-4481-01-4 (ebook)
World Scientific Publishing Co Pte Ltd
Szabo, Raisa; He, Matthew; Burnham, Erick; and Jurani, Jessica, "Analysing Patterns of Tripeptides Using Statistical Approach and Neural Network Paradigm" (2004). Mathematics Faculty Proceedings, Presentations, Speeches, Lectures. 362.