Goh, Liang Kee

Research Interests:

Liang Kee Goh is an assistant professor at Duke-NUS Graduate Medical School. Her research interests are in computational biology; using machine learning and statistical inference to data mine for biomarkers and molecular profiles in the context of cancer and neural diseases. Her early works have focused on developing novel computational methods and mathematical modeling of microarray data, discovering genetic profiling for different types of cancer. While at Duke University, she got involved in the field of epigenetics, especially on hypermethylation and X-inactivation, particularly in understanding the locations and causes of chromosomal aberrations that are linked with genetic diseases. Her group has published a novel method cluster_boost, specifically for data mining in the context of unlabeled and grossly imbalanced data. Using cluster_boost, they have shown that genomic properties play a role in rendering a gene susceptible for DNA methylation in cancer. Since returning to Singapore in 2007, she has also started working in statistical genetics, employing genome-wide association studies (GWAS) to identify candidate genes in ocular and neural diseases. Her lab is working towards system biology, developing methods on genomic convergence and integrative analyses. The idea is to apply statistical machine learning methods to look at data synergistically.

Selected Publications:

Goh, L., S. Murphy, S. Mukherjee, T. Furey (2007). Genomic Sweeping for Hypermethylated Genes, Bioinformatics, Feb 1;23(3):281-8

Goh, L., N. Kasabov (2005). An Integrated Feature Selection and Classification Method to Select Minimum Number of Variables on the Case Study of Gene Expression Data, Journal of Bioinformatics and Computational Biology, Oct;3(5):1107-362005

N., Kasabov, Goh, L., M., Sullivan (2005). Integrated Prognostic Profiles: Combining Clinical and Gene Expression Information through Evolving Connectionist Approach. Book Chapter 10, Information Processing and Living Systems, World Scientific Publishing & Imperial College Press

Goh, L., Q. Song and N. Kasabov (2004). A Novel Feature Selection Method to Improve Classifiation of Gene Expression Data. Asia Pacific Bioinformatics Conference, Dunedin, New Zealand

Goh, L. and N. Kasabov (2003). Integrated Gene Expression Analysis of Multiple Microarray Data Sets Based on a Normalization Technique and on Adaptive Connectionist Model. International Joint Conference on Neural Networks