Date of Award

4-21-2008

Degree Type

Thesis

Degree Name

Master of Science (MS)

Department

Biology

First Advisor

Dr. Alexander Zelikovsky - Chair

Second Advisor

Dr. Robert Harrison - Co-Chair

Third Advisor

Dr. Irene Weber

Abstract

The network-mapping tool integrated with protein database search can be used for filling pathway holes. A metabolic pathway under consideration (pattern) is mapped into a known metabolic pathway (text), to find pathway holes. Enzymes that do not show up in the pattern may be a hole in the pattern pathway or an indication of alternative pattern pathway. We present a data-mining framework for filling holes in the pattern metabolic pathway based on protein function, prosite scan and protein sequence homology. Using this framework we suggest several fillings found with the same EC notation, with group neighbors (enzymes with same EC number in first three positions, different in the fourth position), and instances where the function of an enzyme has been taken up by the left or right neighboring enzyme in the pathway. The percentile scores are better when closely related organisms are mapped as compared to mapping distantly related organisms.

DOI

https://doi.org/10.57709/1059209

Included in

Biology Commons

Share

COinS