AGILE: A Cloud Approach to Automatic Gene Expression Pattern Recognition and Annotation Over Large-Scale Images

Modern biomedical research makes significant use of large datasets. Cloud computing is emerging as a cost-effective solution by providing virtual computers and storage disks on demand to store and process massive data efficiently without large upfront costs.

Despite some progress made, the use of cloud computing in the biomedical research is still at the very early stage. There exist various concerns on how to best utilise the cloud for accelerating large-scale biomedical applications. Especially, can a biomedical application be directly migrated to the cloud without requiring any modification? How to develop a cloud-based biomedical application? What are the performance and the cost of an application in the cloud? Are the performance and the cost acceptable? Do we have optimal methods to keep both performance and the overall cost of applications within the acceptable range in the cloud?

This project will develop a cloud approach for a real biomedical data intensive task for effective gene expression pattern recognition and annotation over large-scale image data through addressing the concerns above. This task is chosen largely for its importance in the biomedical research. This type of intensive data-analysis task is increasingly common in the biomedical sciences. This particular task concerns developmental anatomy of mouse embryo: it is of great interest to identify gene interactions and networks that are associated with developmental and physiological functions in the embryo by using anatomical annotation. The gene expression pattern recognition and annotation represents labelling embryo images with anatomical terms for mouse development. If an image is tagged with a term, it means the corresponding anatomical component shows expression of that gene. Currently, this task is mainly taken manually by domain experts. However, with the availability of the vast amount of data, a manual annotation is expensive and time consuming. Additionally, the manual annotation may also produce the inconsistency of labels across images introduced by the human annotators as it proves to be highly subjective. To alleviate issues with the manual annotation, we have employed data mining techniques to automatically identify an anatomical component in the embryo image and annotate the image using the provided terms. As this task involves the use of very large-scale images, we intend to exploit cloud computing for this task to address the massive data problems.

It is expected that the successful completion of this project will provide a typical exemplar for accessing and exploiting cloud computing technologies to analyse large-scale image-based biomedical data. An important, and novel, aspect of this proposal is that the major concerns that limit the more widespread use of cloud computing for biomedical applications will be addressed. The theoretical component of the work aims to provide (1) a practical user-friendly biomedical data-mining tool based on the cloud for effective gene expression pattern recognition and annotation and (2) a set of standard services (e.g. image processing algorithms, data mining algorithms) and a novel automatic data reuse mechanism for performance enhancement and cost reduction, which can be reused and plugged into the class of similar biomedical applications.

Biomedical research is moving to a data-centric paradigm, which requires large computing resources to scale with the increasing volume and complexity of biological data. Cloud computing is emerging as a cost-effective solution to process and store massive data on demand. However, there are numbers of challenges that limit its widespread use in the biomedical data-driven applications. Particularly, how to develop a biomedical application based on the cloud? How to optimise the performance and reduce the overall cost of the application within the acceptable range?

With these questions in mind, this project proposes a cloud approach to develop a cloud-based biomedical application tool for efficient gene expression pattern recognition and identification, which involves the use of large-scale images. The project addresses the challenges hindering the wider adoption of cloud computing for biomedical research. The theoretic component of the work aims to 1) develop a user-friendly, cloud-based gene expression pattern recognition and annotation tool by using parallel processing models to enable the application to run on the cloud; and 2) provide a novel data reuse mechanism to automatically determine which data should be deleted or stored for optimising the performance and reducing the overall cost of applications and a set of standard services, which can be reused and plugged in the class of similar applications.

These consist of main contribution and the novelty of the project.

The potential beneficiaries this work might be of interest to can be broadly classified into two groups including specific users and wider users.

1) Specific users
1.1) Biomedical researchers who use the gene-expression data resources, such as EMAGE that have major genomic-wide high-throughput data. Those resources represent massive untapped "data-mines" that will be used for decades to extract information using novel image-processing and data-mining algorithms. It is unlikely that existing resource providers will be able to satisfy this demand. This research will propose a cloud approach to develop a practical cloud-based tool for gene expression recognition and annotation. On the other hand, it will also test the feasibility of using cloud computing for facilitating data-intensive biomedical research.
1.2) Researchers in areas of other disciplines such as environmental, engineering and physical sciences, those who face a "data deluge" problem that requires support from high performance infrastructures, and would like to use and/or build infrastructures based on cloud computing. This work will inform good practice for researchers. They can learn and benefit from the development of this project to solve domain-specific data intensive computing problems. To date, a few leading international research institutions such as Cornell, Pudue Universities, CERN and Harvard Medical School have taken steps to move their computing infrastructures into the cloud. The output of this research will contribute to raising the international standing of UK research in the use and exploration of cloud computing as a platform.
1.3) A skilled researcher, fluent in domain-specific applications of cloud computing will be produced by the unique interdisciplinary nature of this project, which brings together researchers from the biomedical and computer sciences in the context of modern computational biomedical practice.
1.4) University students and lecturers in biomedical sciences can use the tools developed in the project for teaching and learning.
1.5) Information service management teams in universities which build computer infrastructures. One result of the project is the provision of good practice advice for exploiting cloud computing and information in order to manage the increasing complexity of the infrastructures for data intensive applications, and so achieve operational and environmental efficiency.
1.6) Practitioners in industry who design and implement cloud computing. The results of this work will provide feedback to cloud providers for optimization of cloud computing technologies and quality enhancement of cloud services offered by providers.
1.7) Policy makers and research councils. Research councils can use the outcomes as evidence to inform and influence policy-makers so that the policy makers can formulate policies and strategies for future investment and adoption of cloud computing.

2) Wider users
General public and public sectors, who may use cloud services but without awareness of that fact, and who have little knowledge about the cloud and gene expression annotation. One outcome of the research will be to educate public audiences and raise awareness of the cloud computing and gene expression annotation such as concepts and underlying technologies by distributing flyers, posters and involving in outreach activities.

In order to ensure that these potential beneficiaries have the opportunity to benefit from this research, we will adopt different impact activities. These activities include the dissemination of project deliverables and software through the project website, research publications in prestigious journals (e.g., Bioinformatics, BMC biology, PlosOne) and appropriate conferences (e.g., IEEE-eScience, HPDC), seminars for biomedical researchers and researchers from other disciplines, and public engagement activities for general audience. The detailed plan is described in the document of pathway to impact.