Predicting protein secondary structure

References:

Chou and Fasman, Ann. Rev. Biochem. 47, 251-276 (1978).
Wilmot and Thornton, J. Mol. Biol. 203, 221-232 (1988).
Qian and Sejnowski, J. Mol. Biol. 202, 865-884 (1988).
Biou, Gibrat, Levin, Robson, and Garnier, Protein Eng. 2, 185-191 (1988).
Holley and Karplus, Proc. Natl. Acad. Sci. USA 86, 152-156 (1989).

Elements of Protein Secondary Structure

alpha-helix:

beta-sheets:

beta-turns:

Protein Secondary Structure Prediction--Lab

Start with a protein sequence of interest. One good place to get a sequence is http://www3.ncbi.nlm.nih.gov/Entrez/ where you can Search the protein database for, e.g., "human taste receptor t1r2". From the results, I selected entry NP_689418. In FASTA format, it looks like this:

>gi|112789566|ref|NP_689418.2| taste receptor, type 1, member 2 [Homo sapiens]

                MGPRAKTISSLFFLLWVLAEPAENSDFYLPGDYLLGGLFSLHANMKGIVHLNFLQVPMCKEYEVKVIGYN

                LMQAMRFAVEEINNDSSLLPGVLLGYEIVDVCYISNNVQPVLYFLAHEDNLLPIQEDYSNYISRVVAVIG

                PDNSESVMTVANFLSLFLLPQITYSAISDELRDKVRFPALLRTTPSADHHIEAMVQLMLHFRWNWIIVLV

                SSDTYGRDNGQLLGERVARRDICIAFQETLPTLQPNQNMTSEERQRLVTIVDKLQQSTARVVVVFSPDLT

                LYHFFNEVLRQNFTGAVWIASESWAIDPVLHNLTELRHLGTFLGITIQSVPIPGFSEFREWGPQAGPPPL

                SRTSQSYTCNQECDNCLNATLSFNTILRLSGERVVYSVYSAVYAVAHALHSLLGCDKSTCTKRVVYPWQL

                LEEIWKVNFTLLDHQIFFDPQGDVALHLEIVQWQWDRSQNPFQSVASYYPLQRQLKNIQDISWHTINNTI

                PMSMCSKRCQSGQKKKPVGIHVCCFECIDCLPGTFLNHTEDEYECQACPNNEWSYQSETSCFKRQLVFLE

                WHEAPTIAVALLAALGFLSTLAILVIFWRHFQTPIVRSAGGPMCFLMLTLLLVAYMVVPVYVGPPKVSTC

                LCRQALFPLCFTICISCIAVRSFQIVCAFKMASRFPRAYSYWVRYQGPYVSMAFITVLKMVIVVIGMLAT

                GLSPTTRTDPDDPKITIVSCNPNYRNSLLFNTSLDLLLSVVGFSFAYMGKELPTNYNEAKFITLSMTFYF

             TSSVSLCTFMSAYSGVLVTIVDLLVTVLNLLAISLGYFGPKCYMILFYPERNTPAYFNSMIQGYTMRRD

You can select and copy this sequence to a word processing document for future use.

Internet servers for secondary structure prediction

NNPredict method: http://www.cmpharm.ucsf.edu/~nomi/nnpredict.html or, by e-mail, nnpredict-request@celeste.ucsf.edu, request help.
The Swiss Institute of Bioinformatics has links to several secondary structure prediction tools.
Secondary structure prediction server at Pole Bio-Informatique Lyonnais in Lyon, France. This site has a wide selection of prediction methods, as well as a consensus method.

Predicting Transmembrane Segments

Now let's take a sequence and try to predict transmembrane segments. The taste receptor belongs to a superfamily of proteins having seven transmembrane helical segments.
· Go to the TMPred site, http://www.ch.embnet.org/software/TMPRED_form.html
· Copy and paste the receptor sequence into the window. Click on the Run TMPred button.
· The algorithm will identify possible transmembrane segments, and then propose one or more models for the protein's membrane topology. Here is a sample result:

-----> STRONGLY prefered model: N-terminus inside
  10 strong transmembrane helices, total score : 14858
   # from   to length score orientation
   1    7   25 (19)     991 i-o
   2  143  165 (23)     921 o-i
   3  318  336 (19)     771 i-o
   4  384  405 (22)     696 o-i
   5  567  588 (22)    2572 i-o
   6  603  621 (19)    1912 o-i
   7  635  659 (25)    1566 i-o
   8  683  702 (20)    1769 o-i
   9  728  748 (21)    1606 i-o
  10  786  810 (25)    2054 o-i