This new version has several capabilities not available in previous versions, including repeat detection, inversion and palindrome detection, local multiple alignment with re-arrangements and frame-shift detection. It achieves the highest accuracy scores so far reported on BALIBASE, PREFAB and BRALIBASE.

Version 3.7 is still being maintained and supported as it offers faster speed and reduced memory use on large numbers of sequences and has several capabilities not yet implemented in v4. Feedback from users interested in trying the new version is welcomed.

Currently there is only minimal documentation for the new program; see the README.TXT file in the download. Source code and a build script is provided for Unix and Linux systems, plus a Microsoft Visual Studio project file. No binaries are currently provided.

Download MUSCLE v4 source code.

UCLUST is a very efficent sequence clustering algorithm that can generate clusters of millions of sequences in minutes on a laptop at a wide range of identities. This may be of interest to users who wish to align very large sets.

Running MUSCLE via the web
Here is the MUSCLE web server.

Running MUSCLE locally
You can run MUSCLE locally as a command-line program. It has been compiled for Windows, Mac and Linux, and should be readily portable to other platforms. The source code is freely available. It is very simple to install, all you have to do is copy the binary file. Any standard PC that is used to run bioinformatics software should be able to run MUSCLE. I am happy to help port to other platforms.

Downloads
You can download MUSCLE here. 

Citing MUSCLE
Please cite:

Edgar, Robert C. (2004), MUSCLE: multiple sequence alignment with high accuracy and high throughput, Nucleic Acids Research 32(5), 1792-97.

Complete description of the algorithm
The NAR paper above gives only a brief overview of the algorithm and implementation details. For a full discussion of the method and many of the non-default options that it offers, please see:

Edgar, R.C. (2004) MUSCLE: a multiple sequence alignment method with reduced time and space complexity. BMC Bioinformatics 5(1): 113.