Background:
Our present investigation was conducted to explore the computational algorithm
for the protein secondary structure prediction as per the property of evolutionary transient and
large number (each 50) of homologous mesophilic-thermophilic proteins.
</P><P>
Objectives: These mesophilic-thermophilic proteins were used for numerical measurement of helix-sheetcoil
and turn tendency for which each amino-acid residue is screened to build up the propensity-table.
Methods:
In the current study, two different propensity windows have been introduced that allowed
predicting the secondary structure of protein more than 80% accuracy.
Results:
Using this propensity matrix and dynamic algorithm-based programme, a significant and decisive
outcome in the determination of protein (both thermophilic and mesophilic) secondary structure
was noticed over the previous algorithm based programme. It was demonstrated after comparison with
other standard methods including DSSP adopted by PDB with the help of multiple comparisons
ANOVA and Dunnett’s t-test.
Conclusion:
The PSSD is of great importance in the prediction of structural features of any unknown,
unresolved proteins. It is also useful in the studies of proteins structure-function relationship.
SSpro/ACCpro 5: almost perfect prediction of protein secondary structure and relative solvent accessibility using profiles, machine learning and structural similarity
