The scope of this meeting is the infrastructure, software, and algorithms needed to analyze large data sets in biological research.

INFO

Discussion Themes:

• Single Cell
• Personalized Medicine and Biomarkers
• Imaging
• Machine Learning
• Algorithmics
• Tools, Infrastructure, and Visualization

Poster Presentation

Large-scale cancer genome sequencing programs have revealed mutational heterogeneity and identified of genomic alterations driving cancer. However, a single cancer-driving mutation or mutated gene alone may not be sufficient to cause cancer and how various driver genes or mutations interact to cause cancer is still under investigation. Here, we estimate model-independent mutational interactions across 34,674 tumours by integrating samples from 3 major cancer databases, TCGA, ICGC and COSMIC, representing 24 types of cancer. By analysing the interaction patterns amongst driver genes, we uncover important molecular mechanisms underlying cancer progression and find that the interactions may depend on the tissue or cell of origin. We also investigate the interaction at the mutational level, which shows different features with that at the gene level. Our results suggest that it is important to consider the interaction amongst specific mutational sites as it may have distinctive properties and provide more personalised medical care to patients. Finally, we demonstrate the presence of third-order interactions in cancer, which significantly affect the interactions between two genes and have distinctive prognostic impacts. In summary, these findings may have important clinical implications for improving survival prediction, indicating therapeutic strategies and stratifying tumour subtypes.

More information