The study is performed in 20 different hospitals from 19 cities in China. Three sub-projects are included. About sub-project 1, we build a clinical database system and a biological sample bank for data and samples management, which is applicable in other hospitals in this project. 1800 first-episode schizophrenia patients will be recruited in 19 sites and randomized into 6 treatment groups (olanzapine, risperidone, aripiprazole, ziprasidone, amisulpride, haloperidol). Through 8-week treatment and follow-up, we collect multidimensional indexes from psychopathology, neuropsychology, brain imaging, physiology, biochemistry, and life stress data. The summarized data is analyzed to screen potential biomarkers or biomarker panel that may predict the antipsychotic response, and ultimately to establish a prediction model.Sub-project 2, as an extension of sub-project 1, includes verification of the prediction model established in sub-project 1 and optimization of the current therapy with add-on treatment. Firstly, the validation process of the prediction model undergoes with an independent patient cohort. Next, we apply the add-on treatment to the patients who don't have ideal response to antipsychotic treatment after 8-week treatment. According to the results above, we manage to construct an optimized and individualized therapy for schizophrenia.In the end,We tend to conduct a randomized double-blind controlled trial to assess the safety and efficacy of the combination strategy for antipsychotic-induced metabolism syndrome, which includes metformin and lifestyle intervention. In the meanwhile, for schizophrenia patients at high-risk of metabolic syndrome, we tend to establish a prevention strategy expected to reduce or delay the occurrence of metabolic syndrome, which includes low-dose metformin and lifestyle intervention. We hope to successfully construct a comprehensive intervention strategy on metabolic syndrome induced by antipsychotic medications.