A web-server that compares protein-ligand interaction site

This project is a continuation of a previous ENS 491/2 graduation project. Only those students who worked therein will be accepted. So please do not apply unless this condition applies to you.
 
With the increase in structural genomics studies, the development of automated methods for predicting the molecular functions of proteins from their structural analysis has come into prominence over the years. In most cases, proteins exert their functions by interacting with other molecules, so-called ligands. These molecules generally interact with specific sites on the protein of interest forming pocket-like indentations called binding sites. The comparison of the binding sites pertaining to the proteins of interest, which can associate with the same/structurally similar ligands, is a key mediator to develop more effective systems in fine-tuning research areas, such as biosensor design as well as drug repurposing, and structure-based drug design. 
To this end, we have already developed a pipeline that classifies all proteins, to which a selected molecule and its derivatives are bound as ligand, according to the structural and sequential similarity of their binding sites as well as their physicochemical and geometric properties. This pipeline briefly includes the following steps:  
1. Finding all proteins to which the selected ligand and its structurally similar derivatives are bound, and subsequently extracting their information from the Protein Data Bank 1
2. Performing binding site similarity analysis via a server to better define the ligand-binding sites of these proteins as well as determine other proteins with the similar binding pocket 2
3. Identification of different proteins consisting of more than one chain by means of a server and rearrangement of the datasets containing unique protomers of these proteins 3
4. Further classification of the proteins with similar binding sites by considering common discerned proteins in the obtained dataset
Within the scope of this project, it is set out to  (i) prepare the interfaces of the pipeline described above, (ii) develop an open-source web-server that scans all molecules seen in nature as ligand, and  (iii) find related proteins with similar binding sites for each of these ligands -with a single click. After creating the web-server that automates all these steps, the established system will be studied for ligands of three different sizes. The dataset obtained for each ligand will be interpreted with respect to the evolutionary information. In addition, some amino acids in the binding region of a selected system will be modified according to the analysis obtained by the web-server. Thereafter, molecular dynamics simulations will be performed for the system. In this way, the effect of the modified binding site on protein-ligand interaction and protein dynamics will be scrutinized.
Note:The proposed project is ideally suited for a student with a double major in CS and BIO, or for a group of 1 CS and 1 BIO student. Those interested in this project are expected to (i) have a good knowledge of programming -preferably one of the following programming languages, Python, MATLAB, Java, and C++-,  (ii) be capable of using databases and creating interfaces, and, last but not least, (iii) have a basic knowledge of protein structure and function. When you contact us to express your keen interest in this project, please attach your detailed CV listing your experiences and also your transcript to the email.
References:
1         H. M. Berman, Nucleic Acids Res., 2000, 28, 235–242.
2         J. I. Ito, Y. Tabei, K. Shimizu, K. Tsuda and K. Tomii, Nucleic Acids Res., 2012, 40, 541–548.
 
3         E. D. Levy, J. B. Pereira-Leal, C. Chothia and S. A. Teichmann, PLoS Comput. Biol., 2006, 2, 1395–1406.