HMMER

This tutorial follows the HMMER User’s Guide written by Sean R. Eddy, Travis J. Wheeler and the HMMER development team. Thanks.

First of all, we download HMMERv3.1b2 from http://hmmer.org and unpack it.

wget http://eddylab.org/software/hmmer3/3.1b2/hmmer-3.1b2.tar.gz
tar xfv hmmer-3.1b2.tar.gz
cd hmmer-3.1b2

./configure
make
make check

export PATH=/path/to/hmmer-3.1b2/src:$PATH

- Build models and align sequences (DNA or protein)

1. hmmbuild - Build a profile HMM from an input multiple alignment.
2. hmmalign - Make a multiple alignment of many sequences to a common profile HMM

- Search protein queries against protein database

1. phmmer - Search a single protein sequence against a protein sequence database. (BLASTP-like)
2. jackhmmer - Iteratively search a protein sequence against a protein sequence database. (PSIBLAST-like)
3. hmmsearch - Search a protein profile HMM against a protein sequence database.
4. hmmscan - Search a protein sequence against a protein profile HMM database.
5. hmmpgmd - Search daemon used for hmmer.org website.

And many others….

Common use of HMMER is to search a sequence database for a protein family of interest (homologues).

Look at the file tutorial/globins4.sto

Construct a HMM from an alignment

hmmbuild globins4.hmm tutorial/globins4.sto

Run your example search against tutorial/globins45.fa.

hmmsearch globins4.hmm tutorial/globins45.fa > globins4.out

As BLASTP or FASTA, the phmmer is for searching a single sequence query against a sequence database.

phmmer tutorial/HBB_HUMAN tutorial/globins45.fa > phmmer.out

The HMM database can be Pfam, SMART, or TIGRFams, or another database of your choice.

A flatfile is just a concatenation of individual HMM files. Given this, we firstly build individual hmm files using hmmbuild and concatenate them using cat. Let's create a small database from the files of tutorial dir.

hmmbuild globins4.hmm tutorial/globins4.sto
hmmbuild fn3.hmm tutorial/fn3.sto
hmmbuild Pkinase.hmm tutorial/Pkinase.sto
cat globins4.hmm fn3.hmm Pkinase.hmm > minifam

hmmpress minifam

Now we can analyze sequences using our HMM database and hmmscan.

hmmscan minifam tutorial/7LESS_DROME

In the previous example, we used three records in the stockhold format: globins4.sto, fn3.sto, and Pkinase.sto. However, What can we do wheather we do not have any source records?

Your task is following:

1. Choose a profile HMMs database.
2. Find three protein families: globin, fn3, and Pkinase. Download their alignments as a seed file (contains representative members of the family which are judged to be well aligned) in the stockholm format.
3. Construct the HMM database (see the previous example).
4. Analyse tutorial/7LESS_DROME using our HMM database.