With the available data at genetic, MRI, behavior and environmental level, we are in the position to first understand the data and then simulate the brain with the data constraints. Here we introduce our endeavors in recent years to analyze these data using machine learning approaches. After a brief introduction of our methodological development, we will concentrate on public health issues related to various measures of our brain such as sleeping duration, social isolation etc. We then turn our attentions to brain diseases including neurology and mental disorders with a special emphasize on subtyping brain diseases. Based upon DTI, grey matter and fMRI, finally, we will briefly introduce our efforts on simulating the whole human brain (digital twin brain, DTB) and current progresses.