Abstract
Structural properties of proteins are being analysed with the help of feed forward neural networks of the perceptron type with hidden layers of neurons. After being trained on known structures the networks can predict local properties of new proteins on the basis of their sequence. Of the structural properties that the networks could predict with reasonable success were the surface structure, H-bond occurrence and secondary structure content. Prediction of the local surface properties means that the networks assign a number to each residue in the sequence signifying whether that residue is deeply burried in the protein or positioned on the surface of the protein. The networks were up to 10 % correct in predicting surface structures of proteins novel to the trained network. A similar score was obtained for the prediction of H-bond occurrence, e.g. if a certain residue participated in forming an H-bond, and for the prediction of secondary structures, i.e. if a certain residue was part of a particular secondary structure. A theoretical model for protein dynamics from an informational processing viewpoint will also be presented.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
N. Qian and T. J. Sejnowski J. Mol. Biol. V. 202, 865, (1988),
H. Bohr, J. Bohr, S. Brunak, R. M. J. Cotterill, B. Lautrup, L. Nørskov, O. Olsen and S. Petersen, FEBS Lett, V.241, 223, (1988)
and L. H. Holley and M. Karplus, Proc. Nat. Acad. Sci. USA, V.86, 152, (1989).
O. B. Ptitsyn et al., FEBS Lett, V.263, 54, (1990).
H. Bohr, “Determination of internal bondings and surface structures of complex protein molecules by neural networks. Abstract for the workshop on ”Complex dynamics in neural networks”, June 1991, IIASS, Vietri, Salerno, Italy.
W. Kabsch and C. Sander, Biopolymers, V.22, 2577 (1983).
H. G. Bohr, R. A. Goldstein and P. G. Wolynes, AMSE Periodicals, Modelling, Measurement and Control, C., Vol. 31, No 2, 55 (1992)
and A. Shrake and J. A. Rupley, J. Mol. Biol., V.79, 351 (1973)
and B. Lee and F. M. Richards, J. Mol. Biol., V.55, 379, (1971).
M. S. Friedrichs and P. G. Wolynes, Science, V.246, 371, (1989)
and M. S. Friedrichs, R. A. Goldstein and P. G. Wolynes, Generalized protein tertiary structure recognition using associative memory Hamiltonians, U. of I. preprint (1991), subm. to J. Mol. Biol.
F. M. Richards, Ann. Rev. Biophys. Bioeng. (1977).
UCSF MidasPlus Users Manual, Computer Graphics Lab. Univ. Of California, San Francisco (Nov. 1989).
H. Bohr and P. G. Wolynes: “Protein Folding: A Physical View of Neural Network Approaches”. Contribution to the proceedings of the workshop on “Neural Networks: From Biology to High Energy Physics”, Marciano Marina, Elba, Italy, June (1991).
H. Bohr, J. Bohr, S. Brunak, R. M. J. Cotterill, H. Fredholm, B. Lautrup and S. B. Petersen, FEBS Lett, V. 261, 43 (1990).
H. Bohr and P. G. Wolynes, The early stages of protein folding from an information processing viewpoint. (Submitted to J. Chem. Phys.) (May 1992).
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1992 Springer-Verlag London Limited
About this paper
Cite this paper
Bohr, H. (1992). Structural Properties of Proteins Predicted by Neural Networks. In: Taylor, J.G., Caianiello, E.R., Cotterill, R.M.J., Clark, J.W. (eds) Neural Network Dynamics. Perspectives in Neural Computing. Springer, London. https://doi.org/10.1007/978-1-4471-2001-8_20
Download citation
DOI: https://doi.org/10.1007/978-1-4471-2001-8_20
Publisher Name: Springer, London
Print ISBN: 978-3-540-19771-3
Online ISBN: 978-1-4471-2001-8
eBook Packages: Springer Book Archive