Question about freak structures used in prediction testing

VICTOR B. STRELETS STRELETS at SCRI.FSU.EDU
Tue Jul 18 11:55:39 EST 1995


Hi,

I'm finishing paper about prediction of secondary structure and find
some freak case(s) of accuracy (less than 20% per-residual in the
simulation of blind test). There are only two such structures,
both are tropomiosins (2TMAA and 2TMAB). They contain only one
secondary structure element: alpha helix of length more than
250 amino acids. It seems reasonable from the point of view of
their function, but I never thought before that it can be obstacle 
when used as a prediction input. And what more intriguing, they
were reported with resolution of 15 Angstroms (but this still
carries name of X-ray definition..).

Could somebody provide me with comment about possibility and
correctness of these proteins usage in testing of the structure
prediction algorithms? I feel that it would be better to exclude
them from consideration, but using what arguments? Because of
their action mechanism they should have such type of structure..
And how could one describe quality of these data in PDB, bearing
in mind 15A resolution?

Thanks,

V.Strelets



More information about the Proteins mailing list