Predicting families of coding sequences
Protea is a software devoted to protein-coding sequences identification. The input is a set of DNA sequences that need not to be aligned. The method takes advantage of the specific substitution pattern of coding sequences together with the consistency of reading frames.
You can use Protea via the web interface. It is also possible to download (protea-0.09.tar.gz) and install it locally. You need a C compiler and some freely available librairies (GMP and MPFR) and UNIX tools (Lex, Yacc). You also need to install ClustalW (credits).