During DNA sequence alignment by MEGA software, there are options like ClustalW and Muscle method. What is the difference in between them and when to choose what?
Almost all multiple sequence alignment programs (MAFFT, MUSCLE, ClustalW, ClustalOmega, etc...) do a pretty good job with most data. When the sequences become very difficult to align because of huge distances between them, there are other issues in addition to poor alignment quality to deal with. Some programs are better suited to protein-coding genes than others, and there is a huge difference in speed between various programs. So if your data set takes more than a few hours to align with CLUSTAL you should consider MAFFT or other methods.
Always check the quality of you alignment by eyebal with a good mutlitple sequence alignment viewer/editor such as BioEdit (free for Windows) or AliView (free for Mac OS).
I found this from a colleague's (Salvadore) response to a similar question "ClustalW implements an iterative algorithm so mistakes produce in earlier step are quite unlikely to be corrected in later step meanwhile muscle implements a progressive algorithm allowing re-optimizations of columns during the whole process."