Once you have a trained HMM, you can use the program hmma (``HMM align'') to make a multiple sequence alignment of a set of sequences (either the training set, or another set of sequences). For instance, we can now take an entire database of globin sequences and align them to a globin model:
> hmma globin2.hmm globins630.fa
The alignment will be printed to your terminal. Alternatively, you can specify a save file with the option hmma -o <outfile>; the alignment will be written to the output file in SELEX format.
The HMM distinguishes between ``match'' states and ``insert'' states. Residues that get assigned to ``insert'' states are not aligned. Their positioning in the alignment is arbitrary. These positions are shown in lower case in the alignment. Aligned ``match'' state positions are in upper case.