Jump to - publications
Link to - CV

I am interested in machine learning applied to the genetics and systems biology domains. In the midst of numerous data sources such as SNP, expression and next-generation sequencing data, it still remains an unsolved problem of narrowing down the causes of complex diseases.

From a machine learning perspective, there are many challenges in this area including working with the high dimensional setting, obtaining statistical power amidst small sample sizes, and learning models that capture the highly structured data. I’m interested in navigating this area and seeing whether variations of regression such as multitask learning, clustering, graphical models or other approaches can be applied.

Currently, I am focusing on using graphical models for network estimation to uncover the underlying regulatory relationships amongst functional genomic elements. In addition, I am building models that can integrate next-generation sequencing data in order to harness the increased resolution of the data and the new types of information (e.g. Genome Sequencing, ChIP-Seq, DNase-Seq etc.).

Previous Work

Prior to starting my PhD, I worked in the area of medical image computing and computer-assisted intervention. I worked with the team at Queen’s University (D. Gobbi, P. Mousavi and P. Abolmaesumi) on SimITK, a project to integrate the ITK image processing libraries into Matlab Simulink to allow for rapid prototyping with drag and drop (link).

After moving to UBC, I began working on image registration of ultrasound (US) and computed tomography (CT). I worked on a method that exploited the geometry of the 3D curvilinear US probe and the physics of US image production to generate a simulated US from the CT image. The intuition is that trying to align two images that look similar is easier than registering two images that appear very different. (link).


  • C. Lippert, J. Xiang, D. Horta, C. Widmer, C. Kadie, D. Heckerman and J. Listgarten. (2014). Greater power and computational efficiency for kernel-based association testing of sets of genetic variants. Bioinformatics. link
  • J. Xiang and S. Kim. (2013). A* Lasso for Learning a Sparse Bayesian Network Structure for Continuous Variables. Advances in Neural Information Processing System (NIPS). pdf poster software
  • J. Listgarten, C. Lippert, E. Y. Kang, J. Xiang, C. M. Kadie, and D. Heckerman. (2013). A powerful and efficient set test for genetic markers that handles confounders. Bioinformatics, 29(12): 1526-33. pdf
  • R. E. Curtis, J. Xiang, A. Parikh, P. Kinnaird, and E. P. Xing. (2012). Enabling dynamic network analysis through visualization in TVNViewer. BMC Bioinformatics, 13(1): 204. pdf
  • J. Xiang, S. Gill, C. Nguan, P. Abolmaesumi, and R. N. Rohling. (2010). Registration of ultrasound to CT angiography of kidneys: a porcine phantom study. SPIE Medical Imaging Conference. 7625: 762518-1-8. pdf slides
  • J. Xiang. (2010). Registration of 3D ultrasound to computed tomography images of the kidney. Masters Thesis. University of British Columbia, Vancouver, Canada. pdf slides
  • A. Crisan and J. Xiang. (2009). Comparison of Hidden Markov Models and Sparse Bayesian Learning for Detection of Copy Number Alterations. Canadian Student Conference on Biomedical Computing. pdf slides Received an Honourable Mention!
  • D. G. Gobbi, P. Mousavi, K. Li, J. Xiang, A. Campigotto, G. Fichtinger, and P. Abolmaesumi. (2008). Simulink libraries for visual programming of VTK and ITK. Workshop on Systems and Architectures for Computer Assisted Interventions, Medical Image Computing and Computer Assisted Intervention (MICCAI). pdf