Transmembrane Potential

Courtesy of Anders Krogh The Transmembrane Hidden Markov Model prediction of the topology of one photosynthetic reaction center chain. The five red bars are the predicted membrane helices and the histogram below represents the certainty with which they can be predicted. Membrane proteins do not reveal their structures easily. Because they are particularly hard to crystallize, such proteins make X-ray crystallography expensive and time-consuming. So, many investigators turn to theoretical, mathematical methods to predict transmembrane protein topologies. This Hot Paper was among the first to report the application of the hidden Markov model (HMM) to such a task. All the information and material exchange of the cell goes through transmembrane proteins. The stakes are potentially high in the hunt for transmembrane topologies. "People are doing data mining to find interesting proteins," says Hot Paper lead author Anders Krogh, a professor of bioinformatics at the University of Copenhagen. "And membrane proteins, for instance, are very important to the drug industry because receptors on the cell are membrane proteins." Named after Russian mathematician Andrei A. Markov, whose work initiated the theory of stochastic processes, HMMs are commonly used to identify members of protein families, predict gene structure, and find sequence motifs that, for example, may help locate transcription factor binding sites upstream of genes. The HMM underlying principle: The elements of interest are aligned to a best approximation. Data derived from the Science Watch/Hot Papers database and the Web of Science (ISI, Philadelphia) show that Hot Papers are cited 50 to 100 times more often than the average paper of the same type and age. A. Krogh et al., "Predicting transmembrane protein topology with a hidden Markov model: Application to complete genomes," J Mol Biol, 305:567-80, 2001. (Cited in 186 papers) In this Hot Paper,1 the HMM uses states, such as transmembrane regions, that model a particular part of the protein sequence. The HMM then makes predictions based on the experimentally derived probabilities that these regions would exist, given a specific amino acid sequence. Prior to the advent of HMMs for transmembrane protein topology prediction, other methods, such as hydrophobicity plots, were used to predict structure. Automation and better accuracy, important when searching through entire genomes, generally make the HMM a superior method. Two groups independently published similar HMM transmembrane protein- prediction methods in 1998.2,3 Krogh et al. first published their results in conference proceedings and were rarely cited. An updated version, the Hot Paper, obviously garnered more recognition. (Their prediction service is available at www.cbs. dtu.dk/services/ TMHMM.) The other group, led by Istvan Simon, a professor of biology at the Biological Research Center's Institute of Enzymology in Budapest, Hungary, also published an updated version in 2001,4 though it has been less cited than their original effort. (Their method is available at www.enzim.hu/hmmtop/index.html.) "In some tests, ours comes out as the best. In others, theirs comes out as the best," says Krogh. The Krogh et al. method can recognize signal peptides, which dictate the final location of a given protein, whereas the Simon et al method cannot; the more proteins with signal peptides, the greater the advantage to using the Krogh et al. method. HMM methods have many uses. Pharmaceutical companies often seek drug targets by first identifying cell receptor families with an appropriate number of transmembrane helices for the drug being studied. Also, investigators looking to construct three-dimensional models of proteins might combine predicted membrane helices with other information about binding sites, for example. Simon notes that his group's 2001 HMM version was the first to allow the incorporation of experimental information. A user could input the experimentally determined location of an amino acid to optimize better the predicted structure. This feature was added recently to the Krogh group's method as well.5 Simon emphasizes the importance of using multiple HMMs and other methods. Doing so, he says, doesn't just lead to more reliable solutions. "Different methods may give different answers," says Simon, "and that can pinpoint some important feature." Eugene Russo (erusso@the-scientist.com) is a freelance writer in Takoma Park, Md. References 1. A. Krogh et al., "Predicting transmembrane protein topology with a hidden Markov model: Application to complete genomes," J Mol Biol, 305:567-80, 2001. (Cited in 186 Papers) 2. E.L.L. Sonnhammer et al., "A hidden Markov model for predicting transmembrane helices in protein sequences," In: J. Glasgow et al., eds., Proceedings of the Sixth International Conference on Intelligent Systems for Molecular Biology, AAAI Press: Menlo Park, Calif., 1998, pp. 175-82. 3. G.E. Tusnady, I. Simon, "Principles governing amino acid composition of integral membrane proteins: Application to topology prediction," J Mol Biol, 283:489-506, 1998. 4. G.E. Tusnady, I. Simon, "The HMMTOP transmembrane topology prediction server," Bioinformatics, 17:849-50, 2001. 5. K. Melen et al., "Reliability measures for membrane protein topology prediction algorithms," J Mol Biol, 327:735-44, March 28, 2003. function sendData() { document.frm.pathName.value = location.pathname; result = false if (document.frm.score[0].checked) result = true; if (document.frm.score[1].checked) result = true; if (document.frm.score[2].checked) result = true; if (document.frm.score[3].checked) result = true; if (document.frm.score[4].checked) result = true; if (!result) alert("Please select") return result; } .myradio{background-color : #94bee5} Please indicate on a 1 - 5 scale how strongly you would recommend this article to your colleagues? Not recommended 1 2 3 4 5 Highly recommended Please register your vote

Transmembrane Potential

Interested in reading more?

Become a Member of