Combining sequence-based prediction methods and circular dichroism and infrared spectroscopic data to improve protein secondary structure determinations

In this study we show that the sequence prediction methods have accuracies nearly comparable to those of spectroscopic methods. However, we also demonstrate that combining the spectroscopic and sequences techniques produces significant overall improvements in secondary structure determinations. In addition, combining the extra information content available from synchrotron radiation circular dichroism data with sequence methods also shows improvements.Combining sequence prediction with experimentally determined spectroscopic methods for protein secondary structure content significantly enhances the accuracy of the overall results obtained.With the growing availability of a large number of new genome sequences, there is an ongoing interest in the structures of the proteins represented by the open reading frames in those genomes. Although there has been a growth in the number of crystal structures of proteins as the result of Structural Protemics programmes worldwide, their production has not kept pace with the sequencing effort. Furthermore, such programmes often produce small amounts of any given protein which are not enough for crystal structure analysis. However, these quantities are often sufficient for other biophysical studies.Modern sequence-based prediction methods can provide information on the secondary structure content of these proteins without the need for producing any protein. Recent years have seen large improvements in the per-residue secondary structure prediction from sequence [1,2], although they are still limited in their accuracies, especially when applied to certain classes of protein, most notably those with high β-sheet content.Circular dichroism (CD) and Fourier transform infrared (FTIR) spectroscopies are commonly used techniques for secondary structure content determination, and these methods have also shown great advances in their accuracies [3], with the development of new reference data bases [4] and empirical methods [5-7]. The spectra