Model for AD: by far most used are transgenic mice that express one or more mutant human proteins including APP and Tau, but also PS1, BACE, ... far less popular are transgenic rats of which only some models are available - a "best" model does not exist
@Kushal Biswas:please refrain from giving advice that is based on no hands-on experience!
@Kushl Biswas: there are plenty of reasons why only experienced people should give advice here - for one: in the AD field too much effort and money and carreers are waisted on wrong research in wrong models ... I take your youth for the explanation for your not so polite answer !
Colton has developed a model (mNos2 Deletion and Human NOS2 Replacement on an APPSwDI background) that provides the closest analogy to human sporadic AD pathophysiology of any of the mouse models available to date. See J Neuropathol Exp Neurol 73 (8): 752-769 (2014).