Skip to main content

Big Data Big Impact Grant: Deep Learning of Complex Genomics Data for Effective Clinical Decisions

Project Member(s): Kennedy, P., Catchpoole, D., Li, J., Leong, T.

Funding or Partner Organisation: Cancer Institute NSW

Start year: 2015

Summary: How can we identify patients destined to fail with current treatment approaches? Currently clinicians¿ group symptomatically similar patients into broad clusters and treat them with the same treatment regimens. Ideally this stratification strategy should reliably identify specific patients who are at greater risk of relapse who would then undergo more stringent treatment protocols. However, cancer is biologically complex and defies the apparent clinical groupings applied. It has become apparent that improved patient outcome requires the development of approaches that lead to ¿personalised medicine¿, treatment strategies which are tailored to specifically benefit individual patients. This requires approaches which capture all of the important genotypic and/or phenotypic variations among patients, not just a few apparent ones. By then allowing the clinician to visualise patients based on this information so that distances between patients reflect their relevant biological similarity, the physician will be able to compare patients which are similar based on common biology, rather than on just clinical presentation. Our ¿constructive¿ approach to develop a similarity space that grasps the ¿big picture¿ of biological relationships between and within patients and will form the basis of a patient specific "clinical decision support system". Uniquely, this study will specifically address the implementation of our proposed application within the clinic.

Publications:

Anaissi, A, Goyal, M, Catchpoole, DR, Braytee, A & Kennedy, PJ 2016, 'Ensemble Feature Learning of Genomic Data Using Support Vector Machine.', PLoS ONE, vol. 11, no. 6, pp. 1-17.
View/Download from: UTS OPUS or Publisher's site

Braytee, A, Liu, W & Kennedy, P 2016, 'A Cost-Sensitive Learning Strategy for Feature Extraction from Imbalanced Data', Springer International Publishing, International Conference on Neural Information Processing, Springer International Publishing, Kyoto, Japan, pp. 78-86.
View/Download from: UTS OPUS or Publisher's site

Tran, J, Nguyen, Q, Simoff, S & Huang, M 2016, 'Areas of Life Visualisation: Growing Data-Reliance', Lecture Notes in Computer Science, Cooperative Design, Visualization, and Engineering, Springer International Publishing, Sydney, pp. 227-234.
View/Download from: UTS OPUS or Publisher's site

Wang, S, Liu, W, Wu, J, Cao, L, Meng, Q & Kennedy, PJ 2016, 'Training deep neural networks on imbalanced data sets', Proceedings of the International Joint Conference on Neural Networks, IEEE International Joint Conference on Neural Networks, IEEE, Vancouver, Canada, pp. 4368-4374.
View/Download from: UTS OPUS or Publisher's site

Keywords: Data mining, deep learning, visualisation

FOR Codes: Pattern Recognition and Data Mining, Bioinformatics Software, Cancer and Related Disorders