Depression is a complex mental health condition that can affect many aspects of a person’s life, including how they feel, think, and handle daily activities. Knowledge graphs are a powerful tool for visualizing and understanding complex relationships between different entities in various domains, including mental health. For depression, a knowledge graph can help illustrate the relationships between symptoms, causes, treatments, comorbid conditions, and other relevant factors and provide an efficient approach for the semantic analysis on the literature on depression. In this paper, we present our extended work on the construction of Knowledge Graphs of Depression. It integrates a wide range of knowledge resources related to Depression, including metadata of medical literature, and their semantic annotations with well-known medical terminologies/ontologies such as SNOMED CT and UMLS. It provides a basic integration foundation of knowledge and data concerning depression for a comprehensive analysis. Furthermore, we present two case studies on the knowledge graphs of depression: i) exploration on the relations between depression and childhood trauma, and ii) investigation on non-suicidal self injury behaviors with depression. Based on those two case studies, we show that how Knowledge Graphs can be used for investigating various clinical scenarios with semantic analysis by the literature mining and gain a comprehensive analysis on those problems of depression patients.