sequence homology

Steven Brenner brenner at
Thu Aug 3 14:13:31 EST 1995

stardog at (Kevin Lease) writes:

>     Can someone point me in the direction of a good program for finding 
>the most conserved regions of a protein by lining up all the a.a. 
>sequences for a given protein from different species?

There are two questions here: (1) making an alignment of the sequences
and (2) doing some analysis of the most conserved positions.  

Along the lines of (1), making a sequence alignment, probably the most
commonly used software is ClustalW.  You can find it at:
among other locations.   Also of interest for doing alingments at
hidden Markov models (HMMs).  You can obtain information about these at: (Eddy & Durbin's system) (Haussler group system)

To do the analysis (2), you can either look at the positions yourself,
your use any of a number of different systems for aiding the analysis.
HMMs automatically give you the level of conservation at each
position.  There are a large number of programs which will aid you in
coloring an alignment according to conservation, and these can be
helpful in a general sort of way.  More intersting is AMAS, a system
intended to do all sorts of various analyses on multiply aligned
sequences.  An interface to this is available at:

Finally, with apologizies for being pedantic, I just want to remind
one of the other people who followed up this thread that proteins the
sequences can not be 'more' or 'less' homologous.  Homology means that
two things stem from a common ancestor; to proteins are either
homolgous or they are not homologous (just as something is either
unique or not unique; it can't be 'more' unique).  Typically the
appropriate word is 'similarity,' and a high level of similarity
typically implies homology.


Steven E. Brenner                    | S.E.Brenner at 
MRC Laboratory of Molecular Biology  | 
Hills Road                           | Office:   +44 1223 248011
Cambridge CB2 2QH, UK                | Fax:      +44 1223 213556

More information about the Bio-soft mailing list