DISCOVERING GENETIC FUNCTIONS USING LINKED OPEN DATASETS ON THE WEB

Although the fast advances of bioinformatics research and technologies have led to the production of a vast amount of data, translating the genetic understandings obtained from studies of model organisms into human clinical efficacy remains of limited success. Until the analysis of the data deluge at large scale is computerised and automated, the productivity of genetic research and its application to drug development will lag behind the fast growth of data. The World Wide Web offers a scalable platform for publishing the fast growing research datasets. Its maturing standards and protocols ensure its linking capability and the interoperability between data shared on the Web. The Linking Open Data project (LOD, http://linkeddata.org/) is an emerging community that promotes the publication of data on the Web using an interoperable format. This provides a unique opportunity to investigate and prove the feasibility of supporting genetic research at a Web-scale using the interoperable, webs of linked research data.I propose a new practice of interacting with the segmented data on the Web that will improve the productivity of genetic research as applied to medicine, drug development, and epidemiology. The data will be automatically connected to form webs of linked datasets, building upon an interoperable representation format and machine-processable metadata. Analysis of the datasets will be automated, by a set of computer tools that I will create in response to concrete users' need. This will form a 'genetic platform' on top of the Web that will assist with the task of gathering and analysing research datasets at a Web-scale. Through ongoing collaborative interaction with biomedical researchers, the true barriers that presently hinder productive genetic research will be identified and the impediments will be removed in a user-led, incremental approach. By experimenting with current technologies, I will uncover their strength and shortcomings, and identify the missing technical, biological and social components in supporting data-intensive genetic studies.I will use two test beds, one for genomics studies and the other for proteomics studies. I will work together with Drosophila researchers from Oxford and Cardiff, and bioinformaticians from Amsterdam, and computer scientists from the LOD and the Semantic Web Health Care and Life Sciences (HCLS) (http://www.w3.org/2001/sw/hcls/) communities. Biological use cases will drive the research agenda, and new computational practices for interacting with large scale biological research data will be evaluated by the extent it removes the identified impediments of Drosophila genetic research productivity or improves the productivity compared to the conventional research practice. The efficiency of my new practice is built upon the linked research datasets, enabled by a new practice of data publication. A demonstration of its efficiency will drive adoption of this new data publication practice within the biomedical research community, assisted by the maturing tools and standards emerging from the web community. This will lead to a cultural shift! I will devote continuous efforts to transferring this cultural evolution to the wider community, in order to promote step changes to the convergence of biomedical research and its clinical application, such as the personalised medicine . The emergence of a new bioinformatics science is foreseeable, to pass the knowledge required for adopting this new practice of data publication and analysis to the young generations of biological and medical students, who will become the new scientists working with research data on a large scale and lead new innovations in biomedical research and its applications.