Identification of fundamental dynamics from large-scale time-series data for causal interaction networks

Revealing the complex operation principles of organisms and deciphering the information hidden within biological data have always been crucial topics in systems biology. With the advancement of modern biotechnologies, increasingly robust biological detection techniques have generated vast amounts of data. Time-series biological data, compared to static data, better capture the dynamic nature of biological processes. However, effectively integrating and analyzing diverse layers of biological time-series data remains a complex challenge.

This thesis aims to propose distinct analysis methods for two different types of biological time-series data. Firstly, for mRNA-seq time-series data, we present a comprehensive analysis pipeline that includes the identification of differentially expressed genes, gene clustering, and the construction of gene regulatory networks. Through these methods, we can uncover potential regulatory relationships among gene sets, preparing the ground for subsequent experimental validation.

Secondly, we investigate immunomics time-series data from individuals undergoing specific immunotherapy for various allergies. To provide personalized treatment strategies, we perform causal analysis and predictions on immunomics data, with the goal of identifying reliable immunotherapeutic biomarkers. These biomarkers can help determine the most suitable treatment approaches for individual patients, enhancing treatment effectiveness and personalization.

In summary, this thesis aims to utilize different analysis approaches to unravel the intricacies of two distinct types of biological time-series data, shedding light on complex biological processes and information within organisms. This work provides valuable support for further biological research and medical applications.