Hi Bharat Gupta,

Motifs may be conserved because of function or because of structure. Sometimes the active site environment is conserved, but most other regions are without clear homologies (e.g. triglyceride lipases). We have studied both the lipase family as well as the trypsin family of proteins and have been able to trace evolutionary aspects for 1-2 billion years, which we found very interesting. I list afew papers below that my be of interest to you. The last two papers addresses conservation aspects of amino acid pairs in proteins.

The origin of trypsin: Evidence for multiple gene duplications in trypsins

By: Baptista, AM; Jonson, PH; Hough, E; et al.

JOURNAL OF MOLECULAR EVOLUTION Volume: 47 Issue: 3 Pages: 353-362 Published: SEP 1998

A critical view on conservative mutations

By: Jonson, PH; Petersen, SB

PROTEIN ENGINEERING Volume: 14 Issue: 6 Pages: 397-402 Published: JUN 2001

Identification of conserved residues in family of esterase and lipase sequences

By: Drablos, F; Petersen, SB

Edited by: Rubin, B; Dennis, EA

LIPASES, PART A: BIOTECHNOLOGY Book Series: Methods in Enzymology Volume: 284 Pages: 28-61 Published: 1997

Scale-Free Behaviour of Amino Acid Pair Interactions in Folded Proteins

By: Petersen, Steffen B.; Neves-Petersen, Maria Teresa; Henriksen, Svend B.; et al.

PLOS ONE Volume: 7 Issue: 7 Article Number: e41322 Published: JUL 26 2012

Hyperdimensional Analysis of Amino Acid Pair Distributions in Proteins

By: Henriksen, Svend B.; Mortensen, Rasmus J.; Geertz-Hansen, Henrik M.; et al.

PLOS ONE Volume: 6 Issue: 12 Article Number: e25638 Published: DEC 9 2011

Hi Bharat

Nice to know your work. Lots of work is going on in the related field, most of them are trying to understand the evolution of motifs computationally. I have also done a work closely related to your field. my question was whether folding thermodynamics could be a tool to understand phylogenetic relationship among proteins, and I got some good results. you may see this.

Thanks Prof. Petersen for such a detailed response. I read the last two papers and in both the papers the data was analyzed for non-homologous proteins. However, in my case I have already analyzed the non-homologous data and wanted to see what happens to the conservation in the homologues of each protein structure from the non-homologous data set. For eg., l generated a set of some 500 non-homologous protein sequences from PDB to examine a certain structural motif. I found that it was conserved in some 70% of the structures. Next, to see whether the homologues of each of those 500 structures follows the same conservation pattern or not , I searched for the homologues for each non-homologous structure ? I hope I made my point clear. Does this kind of analyses makes sense or not ?? Please comment.

Dr. Sen, thank you for kind response. I have stated my query above and hope you could provide some insight ...

As much I understood, you have kept '70% homology' of 'the motifs' as the benchmark for choosing non-homologous proteins, and then your wish is to search homologous proteins to those 'non-homologous' proteins. then you want to compare the motifs of homologous proteins and correlate it with respective non-homologous proteins'. Idea is good, but you have to set clear logics, like why 70% homology benchmark? your choice of motifs etc. Hope you understand this.

It's not a benchmark... It is the % of conservedness of the motif. For eg. out of 100 examples studied, 70 had the motif where as other 30 were having some other motif.