Advisor(s) - Committee Chair
Dr. Claire A. Rinehart (Director); Dr. Doug McElroy; Dr. Sigrid Jacobshagen
Department of Biology
Master of Science
A METHOD BASED ON CONSERVED MULTIPLE AMINO ACID PROPERTIES TO PREDICT AMINO ACID SUBSTITUTIONS WHICH MAINTAIN THE PROTEIN STRUCTURE Avinash M. Baktula September 16, 2004 1-117 Directed by: Claire A. Rinehart, Doug McElroy and Sigrid Jacobshagen Department of Biology Western Kentucky University Proteins often contain several domains, each with a distinct structure. Such domains have evolved as units that, when combined in various arrangements, produce proteins of unique structure. This study was conducted to identify amino acid substitutions that don’t change structure. Amino acid properties which were conserved in proteins with identical structures were used to predict a set of amino acids profiles at each sequence position that can serve as viable substitutions. To test this analysis ten different protein sets were taken from the Conserved Domain Database of National Center for Biotechnology Information (NCBI). An amino acid index database of numerical indices representing various physicochemical and biochemical properties of amino acids were mapped onto the amino acid sequences in each dataset and these were used to select properties common to the proteins with the same structure. Based on these conserved properties, a substitution index percentage (SI%) was calculated to represent the relative ability of an amino acid to substitute at a given position and still maintain a protein structure. Amino acid profiles from different SI% ranges were used to create a set of substitutions into the consensus sequence of each dataset (AASCS). The AASCS from each SI% range were submitted to two validation programs, RPS-BLAST and PSI-PRED. The number of matches between the AASCS and the primary data set sequences for each SI% range was used to select the substitution profile ranges that best maintained the structure. It was concluded that amino acid, substitutions with SI% greater than 90% consistently conserved the structure of the protein. This method may prove useful in predicting the structure of proteins with induced mutations (site-directed mutagenesis), and in studies pertaining to protein engineering.
Biology | Biotechnology
Baktula, Avinash M., "A Method Based on Conserved Multiple Amino Acid Properties to Predict Amino Acid Substitutions Which Maintain the Protein Structure" (2004). Masters Theses & Specialist Projects. Paper 1107.