Jian Shi

  • Designation: Associate Professor of Neurology at the University of California
  • Country: USA
  • Title: Machine Learning and Bioinformatics Approaches for Classification and Clinical Detection and Treatment of BVZ-Responsive Glioblastoma Subtypes Based on Multi-Omics


Jian Shi is an Associate Professor of Neurology at the University of California, San Francisco. Dr Shi received her PhD and completed post-doctoral training in China and at the UT Southwestern Medical Center. Dr Shi has published over 20 papers in Cell Death Diff., Sci—Rep., J Clin. Inv., and other journals. Dr Shi is primarily interested in research areas such as brain tumours, brain injury, and neurodegenerative diseases, for which she has used machine learning and bioinformatics skills, molecular and cellular techniques, and preclinical animal studies.


Abstract: Glioblastoma multiforme (GBM) is the most aggressive form of malignant brain tumor. In addition to surgery and radiation therapy, several types of chemotherapy are available for GBM patients. As an anti-angiogenesis treatment, bevacizumab (BVZ) therapy for GBM patients targets vascular endothelial growth factor A (VEGF), which can significantly reduce tumour sizes in the early stage based on imaging studies. However, only some GBM patients respond to this treatment.

To precisely treat GBM patients, we classified and detected BVZ-responsive subtypes of GBM and found their differential expression (DE) of miRNAs and mRNAs, clinical characteristics, and related functional pathways. Approximately 30% of GBM patients were classified as having the GBM BVZ-responsive subtype based on miR-21 and miR-10b expression z-scores. For this subtype, GBM patients had a significantly shorter survival time than other GBM patients (p = 0.014), and vascular endothelial growth factor A (VEGF) methylation was significantly lower than that in other GBM patients (p = 0.005). It also revealed 14 DE miRNAs and 7 DE mRNAs and showed functional characteristics between GBM BVZ subgroups.

 After comparing several machine learning algorithms, the construction and cross-validation of the support vector machine (SVM) classifier were performed based on the existing datasets and miRNA biomarkers. For clinical use, miR-197 was optimized and added to the miRNA panel for better classification. Afterwards, we validated the classifier with several GBM datasets and discovered some key-related issues. According to this study, GBM BVZ subtypes can be classified and detected by combining SVM classifiers and miRNA panels in existing GBM tissue datasets.  

Based on the comparison of before and after BVZ treatment, mRNA expression patterns and functional pathways were very different between GBM BVZ-responsive and nonresponsive subtypes after BVZ treatment. Most importantly, Ang1, SHID1A, and LEF1 expression levels were significantly suppressed in the tissues of GBM BVZ-responsive patients but not in the tissues of GMB BVZ-nonresponsive patients. Furthermore, this study also revealed that there might be ageing-related side effects caused by treatment for GBM BVZ-nonresponsive patients, strongly suggesting the need for pre-detection of GBM BVZ-responsive subtypes. However, further research in the field is needed to better understand the roles of these genes following BVZ treatment.

Don't miss our future updates!

Get in Touch