Computational methods in drug discovery have increasingly gained attention of researchers within the last decades, as the tools for facilitating drug discovery process, for it allow rapid screening of huge databases, the ligand-abased virtual screening (LBVS) methods continue to be developed and improved as one of the important tool of drug discovery process, and it is become one of the most interested area in Chemo informatics, this review discuss various methods that are applied for LBVS, the paper focus deeply in the various of machine learning techniques that have been applied for LBVS, and also it will demonstrate and discuss the effectiveness of data fusion in LBVS, and how these techniques provided effectiveness tools could enhance the LBVS.