Com-DNB: A novel method for identifying critical states of complex biological processes and its parallelization

Identifying critical states prior to critical transitions in complex biological processes is essential for disease forecasting and early interventional therapy. Due to the complexity of the underlying mechanisms, the currently proposed methods based on the dynamic network biomarker (DNB) theory cannot be well used to identify the critical state in a complex biological process like aging. In this paper, we propose the com-DNB method based on single-sample DNB community detection, which is capable of identifying adaptively-sized DNB communities and makes improvements in the quality of DNBs. Meanwhile, this paper substantially improves the computational efficiency of the original l-DNB part it contains based on the parallel strategy. Finally, using the gene expression data of PBMCs and classical monocytes from 130 healthy human donors, 61--65 years of age is successfully identified as the critical state of aging in men. Therefore, the com-DNB is able to identify single-sample personalized DNBs more efficiently and accurately, which is a significant advance in the field of critical state identification.