AN INTRODUCTION TO PROTEIN SECONDARY STRUCTURE PREDICTION (PSSP) 

Dr. Anita Wasilewska 
Computer Science Department 
SUNY at Stony Brook, 
Stony Brook, NY, USA 

Dr. Victor Robles Forcada 
Department of Computer Architecture and Technology, 
Technical University of Madrid 
Madrid, Spain 

ABSTRACT 

We provide an intuitive introduction to the protein secondary structure prediction (PSSP) research preliminaries and problems. 
We explain what are the proteins in real life and provide a simple mathematical model of a protein. 

In our talk we discuss the most modern PSSP datasets, and first, second and third generation of PSSP algorithms. 
We evaluate and compare predictive accuracy of nine of them with our recently built metaclassifier. The metaclassifier 
uses evolutionary information and combines the predictions from the nine classifiers. It gives a 2-3% improvement in 
accuracy over the best single method and 15-20% over the worst.