The future of protein secondary structure prediction accuracy

In order to investigate the path of medical education in Iran, indicators of medical education were searched from 1970 to 1994. There have been rises in the number of educational institutions from 10 to 46; student admissions in programmes of medical sciences from 1387 to 18,141; medical student admissions from 632 to 3630; teaching staff from 1573 to 7979; and teaching-bed to student ratio from 1.05 to 2.08. The numbers of students in clinical specialty and MS degrees have increased, and various programmes in clinical sub-specialty and PhD degrees have been initiated. The quality of medical education has improved with increasing field and ambulatory care training, with more emphasis on teaching preventive medicine and a significant rise in the research activities. Most qualitative and quantitative progress has been achieved following the establishment of a joint Ministry of Health and Medical Education in 1985. The results of this review demonstrate the success of Iran in upgrading medical education by the unification of health services and medical education in one ministry.     

Using evolutionary trees in protein secondary structure prediction and other comparative sequence analyses

Previously proposed methods for protein secondary structure prediction from multiple sequence alignments do not efficiently extract the evolutionary information that these alignments contain. The predictions of these methods are less accurate than they could be, because of their failure to consider explicitly the phylogenetic tree that relates aligned protein sequences. As an alternative, we present a hidden Markov model approach to secondary structure prediction that more fully uses the evolutionary information contained in protein sequence alignments. A representative example is presented, and three experiments are performed that illustrate how the appropriate representation of evolutionary relatedness can improve inferences. We explain why similar improvement can be expected in other secondary structure prediction methods and indeed any comparative sequence analysis method.     

Knowledge-based protein secondary structure assignment

We have developed an automatic algorithm STRIDE for protein secondary structure assignment from atomic coordinates based on the combined use of hydrogen bond energy and statistically derived backbone torsional angle information. Parameters of the pattern recognition procedure were optimized using designations provided by the crystallographers as a standard-of-truth. Comparison to the currently most widely used technique DSSP by Kabsch and Sander (Biopolymers 22:2577-2637, 1983) shows that STRIDE and DSSP assign secondary structural states in 58 and 31% of 226 protein chains in our data sample, respectively, in greater agreement with the specific residue-by-residue definitions provided by the discoverers of the structures while in 11% of the chains, the assignments are the same. STRIDE delineates every 11th helix and every 32nd strand more in accord with published assignments.     

Progress in protein structure prediction

If protein structure prediction methods are to make any impact on the impending onerous task of analyzing the large numbers of unknown protein sequences generated by the ongoing genome-sequencing projects, it is vital that they make the difficult transition from computational 'gedankenexperiments' to practical software tools. This has already happened in the field of comparative modelling and is currently happening in the threading field. Unfortunately, there is little evidence of this transition happening in the field of ab initio tertiary-structure prediction.     

JPred: a consensus secondary structure prediction server

An interactive protein secondary structure prediction Internet server is presented. The server allows a single sequence or multiple alignment to be submitted, and returns predictions from six secondary structure prediction algorithms that exploit evolutionary information from multiple sequences. A consensus prediction is also returned which improves the average Q3 accuracy of prediction by 1% to 72.9%. The server simplifies the use of current prediction algorithms and allows conservation patterns important to structure and function to be identified. AVAILABILITY: http://barton.ebi.ac.uk/servers/jpred.h tml CONTACT: geoff@ebi.ac.uk     

Protein secondary structure prediction using two-level case-based reasoning

We have developed a two-level case-based reasoning architecture for predicting protein secondary structure. The central idea is to break the problem into two levels: (i) reasoning at the object (protein) level and using the global information from this level to focus on a more restricted problem space; (ii) decomposing objects into pieces (segments) and reasoning at the level of internal structures. As a last step to the procedure, inferences from the parts of the internal structure are synthesized into predictions about global structure. The architecture has been developed and tested on a commonly used data set with 69.5% predictive accuracy. It was then tested on a new data set with 68.2% accuracy. With additional tuning, over 70% accuracy was achieved. In addition, a series of experiments were conducted to test various aspects of the method and the results are informative.     

A new approach to secondary structure evaluation: secondary structure prediction of porcine adenylate kinase and yeast guanylate kinase by CD spectroscopy of overlapping synthetic peptide segments

A new approach for evaluating the secondary structure of proteins by CD spectroscopy of overlapping peptide segments is applied to porcine adenylate kinase (AK1) and yeast guanylate kinase (GK3). One hundred seventy-six peptide segments of a length of 15 residues, overlapping by 13 residues and covering the complete sequences of AK1 and GK3, were synthesized in order to evaluate their secondary structure composition by CD spectroscopy. The peptides were prepared by solid phase multiple peptide synthesis method using the 9-fluorenylmethoxycarbonyl/tert-butyl strategy. The individual peptide secondary structures were studied with CD spectroscopy in a mixture of 30% trifluoroethanol in phosphate buffer (pH 7) and subsequently compared with x-ray data of AK1 and GK3. Peptide segments that cover alpha-helical regions of the AK1 or GK3 sequence mainly showed CD spectra with increasing and decreasing Cotton effects that were typical for appearing and disappearing alpha-helical structures. For segments with dominating beta-sheet conformation, however, the application of this method is limited due to the stability and clustering of beta-sheet segments in solution and due to the difficult interpretation of random-coiled superimposed beta-sheet CD signals. Nevertheless, the results of this method especially for alpha-helical segments are very impressive. All alpha-helical and 71% of the beta-sheet containing regions of the AK1 and GK3 could be identified. Moreover, it was shown that CD spectra of consecutive peptide content reveal the appearance and disappearance of alpha-helical secondary structure elements and help localizing them on the sequence string.     

Proteolytic cleavage sites of band 3 protein in alkali-treated membranes: fidelity of hydropathy prediction for band 3 protein

To assess the fidelity of hydropathy prediction for band 3 protein, we determined the cleavage sites of the protein and the portions of the protein tightly bound to the membrane lipid bilayer by means of in situ proteolytic digestion. For the removal of all anticipated hydrophilic connector loops from membranes, we had to denature the band 3 protein molecule in situ by alkali treatment. When the alkali-treated membranes were digested with trypsin, chymotrypsin, and pepsin, the majority of the anticipated transmembrane portions remained in the membrane fraction. However, five anticipated transmembrane portions were released into the supernatant fraction. Thus, the first, second, third, sixth and tenth anticipated transmembrane portions, in accordance with the hydropathy prediction, were released into the supernatant with the proteolytic digestion method. This indicates that these anticipated transmembrane portions are not bound with the boundary lipids although the hydrophobicity of these portions is comparable to that of the portions experimentally remaining in the membrane fraction. It is conceivable that the membrane peptide portions of band 3 protein could be classified into at least two categories, i.e. one bound to the boundary lipids and the other free from the boundary lipids. Approximately 90% of the transmembrane domain of the band 3 protein are recovered in either the supernatant fraction or the membrane fraction. The fidelity of hydropathy prediction for polytopic membrane proteins and the nature of the membrane embedded peptide portions are discussed.     

Gene structure prediction using information on homologous protein sequence

In this paper a new approach for the prediction of protein coding gene structures is described. The principal scheme of prediction is as follows: first, the exons with the best potential are predicted in a sequence with unknown functions and a list of potential amino acid fragments coded by these exons is formed. Second, testing the homology between each amino acid fragment from the list and proteins from the SWISS-PROT database of amino acid sequences. One protein with the best homology is chosen out of all the homologous sequences. Third, reconstruction of the exon-intron structure, basing it on its homology with the chosen protein sequences. The method was tested on an independent control set (20 genes). The results were as follows: 21% of real exons were lost and 3% of non-real exons were found. This system can be used to refine the results of gene prediction systems, especially if highly homologous proteins are found in the amino acid sequence database.     

Effects of proline cis-trans isomerization on TB domain secondary structure

The transforming growth factor beta (TGF-beta) binding protein-like (TB) domain is found principally in proteins localized to extracellular matrix fibrils, including human fibrillin-1, the defective protein in the Marfan syndrome. Analysis of the nuclear magnetic resonance (NMR) data for the sixth TB module from human fibrillin-1 has revealed the existence of two stable conformers that differ in the isomerization states of two proline residues. Unusually, the two isoforms do not readily interconvert and are stable on the time scale of milliseconds. We have computed independent structures of the major and minor conformers of TB6 to assess how the domain fold adjusts to incorporate alternatively cis- or trans-prolines. Based on previous observations, it has been suggested that multiple conformers can only be accommodated in flexible regions of protein structure. In contrast, P22, which exists in trans in the major form and cis in the minor form of TB6, is in a rigid region of the domain, which is confirmed by backbone dynamics measurements. Overall, the structures of the major and minor conformers are similar. However, the secondary structure topologies of the two forms differ as a direct consequence of the changes in proline conformation.     

Update on protein structure prediction: results of the 1995 IRBM workshop

Computational tools for protein structure prediction are of great interest to molecular, structural and theoretical biologists due to a rapidly increasing number of protein sequences with no known structure. In October 1995, a workshop was held at IRBM to predict as much as possible about a number of proteins of biological interest using ab initio prediction of fold recognition methods. 112 protein sequences were collected via an open invitation for target submissions. 17 were selected for prediction during the workshop and for 11 of these a prediction of some reliability could be made. We believe that this was a worthwhile experiment showing that the use of a range of independent prediction methods and thorough use of existing databases can lead to credible and useful ab initio structure predictions.     

FOLD: integrated analysis and display of protein secondary structure

FOLD, a computer program for the definition and analysis of protein secondary structure, is described. Algorithms implemented in the software are reviewed. These include methods for the identification of simple features such as hydrogen bonds, alpha helices, beta strands, beta bulges, and beta and psi turns. Techniques are also described for the definition and analysis of higher-order structures, such as beta hairpins, beta sheets and their topology, and beta barrels. In addition to considerable textual output the program supports visualization of protein secondary structure in either an atom-based display style or one reproducing the characteristics of a so-called ribbon drawing.     

Effects of stimulus intensity on signals from human somatosensory cortices

We recorded somatosensory evoked magnetic fields (SEFs) to left median nerve electric stimulation from seven healthy subjects. The stimulus intensity was varied in three sessions: sensory stimuli evoked a clear tactile sensation without any movement, weak motor stimuli exceeded the motor threshold, and strong motor stimuli caused a vigorous movement. Responses were modelled with sources in the contralateral primary somatosensory cortex (SI), the contralateral and ipsilateral secondary somatosensory cortices (SIIs) and the contralateral posterior parietal cortex (PPC). The amplitude of the 20 ms response from the SI cortex and the subjective magnitude estimations followed the stimulus intensity whereas signals from the three other areas saturated already at the level of the motor threshold. The results implicate differential roles for various somatosensory cortices in intensity coding.     

Protein secondary structure prediction by the analysis of variation and conservation in multiple alignments

A number of methods exist for the prediction of protein secondary structure from primary sequence. One method identifies variable charged and conserved hydrophobic residues within large multiple alignments as a means of indicating outside and inside sites respectively in the protein structure. These sites are then manually fitted to secondary structure templates to generate a secondary structure prediction. Using the existing theoretical bases of this method, we present an algorithm (STAMA) which automatically carries out the initial variation/conservation analysis of the alignment. We also test the accuracy of complete predictions carried out by manual fitting of the STAMA-derived assignments to structure templates, using five large multiple alignments each including a protein of known structure. The method was found on average to predict only 57% of residues in the correct secondary structure, and was only as accurate as predictions carried out using the established and automated method of Garnier, Osguthorpe and Robson (1978) applied to a single sequence. When used in conjunction with other secondary structure prediction methods, however, the resulting consensus predictions were found to be very accurate, with 78% of the elements (alpha helices or beta strands) for which a consensus could be obtained being predicted correctly. The algorithm presented here, plus the assessment of the accuracy of prediction generated by this method, should enable this predictive approach to receive informed general use.     

A mutational study of the peptide-binding domain of Hsc70 guided by secondary structure prediction

The abundant, cytoplasmic molecular chaperones of eukaryotic cells, of which mammalian Hsc70 is a member, have central roles in protein folding pathways in cells. Although substantial information is now available on substrate interactions and ATPase activity, neither the crystal structure of the intact Hsc70 molecule nor its isolated peptide-binding domain is known. Recently, the crystal structure of the isolated peptide-binding domain of an evolutionary relative of mammalian Hsc70, the DnaK protein of Escherichia coli, was solved. We have generated several rat Hsc70 mutants using site-directed and cassette mutagenesis guided by secondary structure predictions to test the hypothesis that the peptide-binding domains of mammalian Hsc70 and DnaK have similar molecular structures. Biochemical properties along with the ATPase and peptide binding activities of the resulting recombinant proteins were determined. Biochemical analyses included one- and two-dimensional gel electrophoresis, electrospray mass spectrometry, and N-terminal amino acid sequencing. The results of our study suggest that the DnaK molecular structure is a useful working model for the mammalian Hsc70 peptide-binding domain. Evidence is provided that (i) small additions to the N terminus of Hsc70 alter the function of the peptide-binding domain, (ii) alterations in the C-terminal tetrapeptide EEVD result in dramatic increases in basal ATPase activity, (iii) polyalanine substitution of a helical connector segment compensates for changes at the C terminus to restore near-normal function, (iv) specific side chain interactions involving this connector segment are not required for peptide-stimulated ATPase activity, and (v) disruption of the cap homologue region inhibits peptide binding by Hsc70.