Working at the interface of linguistics, computational methods and life sciences
Profs Bala and Mahe developed a linguistic based algorithm to crunch a huge text to list out a series of alphabets; ranked in order of their random dispersion. When carried out over a series of random text passages, the program often ended giving out vowels which was hardly surprising given their regular occurrence in English to allow proper pronunciation.
The question that now arose was, what will the test yield when carried out on a series of peptide sequences. The relevance of the question arose from the fact that twenty characters represent the entire protein world; and a juxtaposition of these is what gives the protein molecule its structure and function. Our group carried out a series of tests on enzymes spanning the six major classes from the IUBMB classes. A series of amino acids emerged in a strikingly unique pattern that possibly represented the vowels of the protein structure and function. A series of questions now arose
What are properties of these amino acids?
Do the amino acids represent key junctions of protein folds?
Is it possible to tweak the algorithm to identify the family to which a protein belongs
Work Team...
Fr.
Solomon Sara, Prof. Mahe,
Prof. Bala and
Dr. Natarajan Ganesan
Last edited: Tuesday January 17, 2006 01:27:49 PM
© copyright 2005-. All rights reserved. Created and maintained by Natarajan Ganesan