Agent-Based Data Extraction in Bioinformatics

Abstract

Bioinformatics is an active and important research discipline in which molecular data is exponentially growing in complex nature. Because of the substantial research in this field, researchers are faced with critical issues such as bandwidth, storage, and complexity in order to retrieve molecular data. It becomes very difficult to conduct research using low computational devices such as Internet of things and sensors. We are employing migration of the agent technique to decrease network traffic and to mitigate the client's limited resource problem by utilizing server-side resources to perform large-scale computation. Our proposed solution does not necessitate additional storage or processing power on the client's side which makes it cost effective. In the proposed solution, (i) an agent visits service provider containing biological data, say sequences requested by the client, (ii) agent fetches the required data, and on the server side it will manipulate the data, and (iii) returns along with the required results to its source platform. Thus, it solves the bandwidth, storage, and computational issues without involving the low resources of the client. For the proof of concept, Java Agent Development (JADE) framework is used as an implementation tool and the results are compared with Java Remote Method Invocation (RMI). It is important to note that our findings reveal that our strategy saves the user up to 16.25% of average time with respect to bandwidth. On the other hand, our approach takes 46.82% less time than the other with respect to data that the agent carries. In addition to the previous contributions, our approach acts as a mashup, to collect data in different format from several service providers, and converts it in any required format. Thus, it solves the problem of complexity hidden in the nature of the data to increase the researchers' productivity.