“School of Biological Sciences”

Back to Papers Home
Back to Papers of School of Biological Sciences

Paper   IPM / Biological Sciences / 15616
School of Biological Sciences
  Title:   A Network-based Comparison between Molecular Apocrine Breast Cancer Tumors and Basal and Luminal Tumors by Joint Graphical Lasso
  Author(s): 
1.  Maryam Shahdoust
2.  Hossein Mahjub
3.  Hamid Pezeshk
4.  Mehdi Sadeghi
  Status:   Published
  Journal: computational biology and bioinformatics
  Vol.:  DOI: 10.1109/TCBB.2019.2911074
  Year:  2019
  Supported by:  IPM
  Abstract:
Joint graphical lasso(JGL) approach is a Gaussian graphical model to estimate multiple graphical models corresponding to distinct but related groups. Molecular apocrine (MA) breast cancer tumor has similar characteristics to luminal and basal subtypes. Due to the relationship between MA tumor and two other subtypes, this paper investigates the similarities and differences between the MA genes association network and the ones corresponding to other tumors by taking advantageous of JGL properties. Two distinct JGL graphical models are applied to two sub-datasets including the gene expression information of the MA and the luminal tumors and also the MA and the basal tumors. Then, topological comparisons between the networks such as finding the shared edges are applied. In addition, several support vector machine (SVM) classification models are performed to assess the discriminating power of some critical nodes in the networks, like hub nodes, to discriminate the tumors sample. Applying the JGL approach prepares an appropriate tool to observe the networks of the MA tumor and other subtypes in one map. The results obtained by comparing the networks could be helpful to generate new insight about MA tumor for future studies.

Download TeX format
back to top
scroll left or right