Analysis of protein folds and alignment of secondary structure elements
If you have predicted that your protein will adopt a particular fold within the
database, then an important thing to consider to which fold your protein belongs,
and other proteins that adopt a similar fold.
To find out, look at one of the following databases:
- SCOP (MRC Cambridge)
- CATH (University College, London)
- FSSP (EBI, Cambridge)
- 3 Dee (EBI, Cambridge)
- HOMSTRAD (Biochemistry, Cambridge)
- VAST (NCBI, USA)
(Note that these databases don't always agree as to what constitutes a similar fold, so I would recommend looking
at as many of them as possible).
If your predicted fold has many "relatives", then have a look at what they are. Ask:
- Do any of members show functional similarity to your protein?
If there is any functional similarity between your protein and any members of the fold, then you may be
able to back up your prediction of fold (possibly by the conservation of active site residues, or the approximate
location of active site residues, etc.)
- Is this fold a superfold? If so, does this superfold contain a supersite? Certain folds show a tendancy to bind ligands in a common location, even in the absense of any functional or
clear evolutionary relationships. For an explanation of this, please see our work on supersites.
- Are there core secondary structure elements that should really be present in any member of the fold?
- Are there non-core secondary structure elements that might not be present in all members of the fold?
Core secondary structure elements, such as those comprising a beta-barrel, should really be present in a fold. If your
predicted secondary structures can't be made to match up with what you think is the core of the protein fold, then your
prediction of fold may be wrong (but be careful, since your secondary structure prediction may contain errors). You
can also use your prediction together with the core secondary structure elements to derive an alignment of of predicted and observed secondary structures.
For example, we predicted that the glutamyl tRNA reductases (hemA family) would adopt an alpha-beta barrel fold using a combination
of fold recognition and secondary structure prediction methods. We aligned the secondary structures of diverse members of the
alpha-beta barrel fold using a structural alignment program, and aligned the secondary structures to the core (boxed below)
secondary structure elements.
In the alignment above, each alpha and beta character refers to an entire secondary structure element. Those that
are boxed are core secondary structure elements found in most members of the fold. The alignment of
predicted secondary structures to the core elements appears at the bottom of the figure. Note that I have had to
delete several alpha helices and beta strands from our prediction to allow for alignment. This is not surprising, because
insertions or deletions of secondary structure elements are common across the diverse set of proteins that adopt this fold.
Next sequence to structure alignment.
Back to the Flowchart