An activation force-based affinity measure for analyzing complex networks

Comparisons of non-Gaussian statistical models in DNA methylation analysis

As a key regulatory mechanism of gene expression, DNA methylation patterns are widely altered in many complex genetic diseases, including cancer. DNA methylation is naturally quantified by bounded support data; therefore, it is non-Gaussian distributed. In order to capture such properties, we introduce some non-Gaussian statistical models to perform dimension reduction on DNA methylation data. Afterwards, non-Gaussian statistical model-based unsupervised clustering strategies are applied to cluster the data. Comparisons and analysis of different dimension reduction strategies and unsupervised clustering methods are presented. Experimental results show that the non-Gaussian statistical model-based methods are superior to the conventional Gaussian distribution-based method. They are meaningful tools for DNA methylation analysis. Moreover, among several non-Gaussian methods, the one that captures the bounded nature of DNA methylation data reveals the best clustering performance.

Inferring the temporal order of cancer gene mutations in individual tumor samples

The temporal order of cancer gene mutations in tumors is essential for understanding and treating the disease. Existing methods are unable to infer the order of mutations that are identified at the same time in individual tumor samples, leaving the heterogeneity of the order unknown. Here, we show that through a complex network-based approach, which is based on the newly defined statistic –carcinogenesis information conductivity (CIC), the temporal order in individual samples can be effectively inferred. The results suggest that tumor-suppressor genes might more frequently initiate the order of mutations than oncogenes, and every type of cancer might have its own unique order of mutations. The initial mutations appear to be dedicated to acquiring the function of evading apoptosis, and some order constraints might reflect potential regularities. Our approach is completely data-driven without any parameter settings and can be expected to become more effective as more data will become available.