Abstract: A Drug Target Protein (DTP) Database has been developed having mainly in silico predicted potential drug target proteins and non human homologous genes in bacterial human pathogens. The drug targets of pathogens available in literature are selected to store in the database. Currently 10 bacterial pathogens are considered as initials of the database with their general information (disease caused, common symptoms, available drugs etc). Secondly, the two statistical applications mainly (A) Amino Acid Dynamics and (B) amino acid composition were added in the database for analysis of sequence of drug target proteins. The webpage and data connectivity was maintained by PHP codes. The utility of database is for drug designing and vaccine development for selected pathogen.
INTRODUCTION
To strengthen the work bench of finding, analysis of drug targets in human pathogen is current challenge for pharmaceutical and medical research. Huge availability of protein sequence and computational approaches for identifying target proteins/genes has been a great support for further experimental work. Many drug targets have been identified in such pathogens by various computational tools (Suthar et al., 2009). Another extensive study related to identification of drug targets of selected foodborne pathogens by our group is currently under way (unpublished data). Up to date there is no database to structure these data; therefore a Drug Target Protein (DTP) Database has been developed which contains mainly proteins in pathogens identified as potential drug targets. List of non human homologous genes and general information (disease, symptoms, preventive measures etc.) of the pathogens are also included. Simultaneously, to analyse the annoted drug target protein sequence a PHP based web tool Statistical analysis has been interpreted in the database. Currently, there are few drug targets database are available but they list human proteins as drug targets for various human diseases (http://www.sciclips.com/sciclips/drug-targets-main.do). These databases do not include drug targets of pathogens. Kumar et al. (2007) have predicted drug targets for Brugia malayi, causing lymphatic filariasis. Moreover, there are few other drug target prediction reports which are complied as The TDR Targets Database (Aguero et al., 2008). However, TDR Targets Database includes drug targets for pathogen causing tropical diseases only.
The amino acids sequence and frequency of various amino acids dictates properties of the protein (Kong et al., 1992; Yoon et al., 2000). The database server facilitates two kind of sequence analysis (a) amino acid dynamics and (b) amino acid composition in the target protein. Amino acid dynamics measures frequency of occurrence of a particular amino acid along the protein sequence whereas second application is simply finding composition of various amino acids in whole protein sequence. The plot of frequency along the selected sequence is useful to analyze segment wise characteristic profile of the protein as well as nature of its core. The relative trend of occurrence and composition of amino acids gives idea about nature of the protein. In case of enzyme protein, dynamics of specific amino acid plays major role in maintaining 3D structure of protein and further enzyme specificity and its activity.
MATERIALS AND METHODS
The experiments were done at Indian Institute of Technology Guwahati by Mr. Sushil Kumar Shakyawar as part of his B.Tech project during May 2009 to July 2009.
Data collection: Available general information (disease caused, common symptoms, available drugs etc.) about pathogens in the database were collected from Wikipedia or literature. Some of the drug targets included in the database are already published / communicated by our group. The protein sequences of selected pathogens were downloaded from NCBI and analyzed as described earlier (Suthar et al., 2009). In brief, the gene sequence of Homo sapiens and various pathogens were downloaded from the National Center for Biotechnology Information (NCBI, ftp://ftp.ncbi.nlm.nih.gov/genomes/). The essential gene sequences were downloaded from the Database of Essential Genes (Zhang and Lin, 2009; Zhang et al., 2004) using the open GENparser (https://launchpad.net/opengenparser). Pathogens paralog sequences at 60% identity were purged using CD-HIT and excluded from further analysis (Li et al., 2001). Subsequently, the results were subjected to BLASTP analysis against human protein sequences to identify non-homologous genes in Leishmania infantum at an expectation E-value cutoff of 10-3. The non-homologs were aligned with sequences acquired from DEG using BLASTP at an E-value cutoff of 10-10 and 30% identity.
RESULTS AND DISCUSSION
Pathogenic infections still remain a leading cause of global disease burden. The impact of these diseases in India and across the globe has been tremendous at socioeconomic and public health levels. As most of the pathogens are developing resistance against the available drugs, If the new drug development efforts are stopped, current trends suggest that some diseases will have no effective therapies within the next ten years (Levin, 2001; Johnsen et al., 2009). Thus the development of new drug against bacterial disease is most urgent and continued requirement. As in addition to strong public health structure, the strategy to combat these diseases needs a, effective communication among the general public, we will include some other useful details like common symptoms, available treatment etc in the database which may be a valuable resource. Few studies have been reported about identification of drug targets by experimental as well as in-silico methods from pathogens (Smith, 2004; Chong et al., 2006; Suthar et al., 2009) but currently there are not database available listing these drug targets proteins. A Drug Target Protein (DTP) Database was prepared using data available in literature or published/submitted by our group. Additionally, sequence analysis tool is included in the web server.
A simple text based web model a (Fig. 1) of the database was developed to store and avail all the information and data generated which is stored in .csv files with connectivity of files by PHP programming.
| Fig. 1: | Screenshot of the database |
Further in utilization of the database the two tools were interpreted within the database for analysis of the protein sequence especially those which were annoted as potential drug targets. A completely user defined input based tool was developed in the database to study variation of amino acid along the sequence of protein to study its residual nature. The algorithm of the tool is based on frequency calculation in residues of 10 amino acids along the sequence. The whole protein sequence was first chopped into group of 10 residues along the sequence. In each residues the frequency of user selected amino acid was calculated as following-
The frequency along length of sequence has been used to study dynamics of amino acid. Secondly the compositional analysis of all the 20 amino acids in selected protein sequence were performed in single step and histogram was plotted as output file. Further search application developed in database facilitate user to search a gene ID mainly in drug target protein and nonhuman homologous database.
The developed database is useful resources for designing more effective vaccines and drugs for pathogens which are more resilient or for which drugs are yet to be discovered. The final proteins annoted as potential drug targets can be retrieved from Using the Database, Statistical analysis for two tools developed for residual and sequential analysis of the selected protein and Search for querying the gene ID whether it belongs to drug target of non human homologous.
For validation of the tools developed, DNA polymerase III subunit beta protein of Mycobacterium tuberculosis was considered for dynamic analysis of amino acid along its sequence as well as compositional analysis. Dynamic analysis all four amino acids viz. Phe, Lle, Ile and Leu along the sequence were studied by the tool in our database (Fig. 2a, b). Commonly, protein function is predicted by aligning the sequence with other protein sequence of known function.
| Fig. 2: | (a) Dynamics of phenylalanine and isoleucine in DNA polymerase
III subunit beta protein of Mycobacterium tuberculosis. (B) Dynamics
of leucine and Isoleucine in DNA polymerase III subunit beta protein of
Mycobacterium tuberculosis |
However, this method fails when the query sequence does not show significant sequence similarity with other protein. We are further trying to understand if function of a protein can be predicted by analysis of various amino acids frequencies along the protein sequence. The database is a valuable resource of researcher for planning experimental studies.
ACKNOWLEDGMENT
Assistance of Mr. Nanu Alan Kachari (Scientific Officer Gr.II, Department of CSE, IIT Guwahati) in preparation of the database is highly acknowledged. Infrastructural facility provided by IIT Guwahati is acknowledged.