N-terminal N-myristoylation of proteins: prediction of substrate proteins from amino acid sequence

Myristoylation by the myristoyl-CoA:protein N-myristoyltransferase (NMT) is an important lipid anchor modification of eukaryotic and viral proteins. Automated prediction of N-terminal N-myristoylation from the substrate protein sequence alone is necessary for large-scale sequence annotation projects but it requires a low rate of false positive hits in addition to a sufficient sensitivity.Our previous analysis of substrate protein sequence variability, NMT sequences and 3D structures has revealed motif properties in addition to the known PROSITE motif that are utilized in a new predictor described here. The composite prediction function (with separate ad hoc parameterization (a) for queries from non-fungal eukaryotes and their viruses and (b) for sequences from fungal species) consists of terms evaluating amino acid type preferences at sequences positions close to the N terminus as well as terms penalizing deviations from the physical property pattern of amino acid side-chains encoded in multi-residue correlation within the motif sequence. The algorithm has been validated with a self-consistency and two jack-knife tests for the learning set as well as with kinetic data for model substrates. The sensitivity in recognizing documented NMT substrates is above 95 % for both taxon-specific versions. The corresponding rate of false positive prediction (for sequences with an N-terminal glycine residue) is close to 0.5 %; thus, the technique is applicable for large-scale automated sequence database annotation. The predictor is available as public WWW-server with the URL http://mendel.imp.univie.ac.at/myristate/. Additionally, we propose a version of the predictor that identifies a number of proteolytic protein processing sites at internal glycine residues and that evaluates possible N-terminal myristoylation of the protein fragments.A scan of public protein databases revealed new potential NMT targets for which the myristoyl modification may be of critical importance for biological function. Among others, the list includes kinases, phosphatases, proteasomal regulatory subunit 4, kinase interacting proteins KIP1/KIP2, protozoan flagellar proteins, homologues of mitochondrial translocase TOM40, of the neuronal calcium sensor NCS-1 and of the cytochrome c-type heme lyase CCHL. Analyses of complete eukaryote genomes indicate that about 0.5 % of all encoded proteins are apparent NMT substrates except for a higher fraction in Arabidopsis thaliana ( approximately 0.8 %).     

Generation of deviation parameters for amino acid singlets, doublets and triplets from three-dimensional structures of proteins and its implications for secondary structure prediction from amino acid sequences

We present a new method, secondary structure prediction by deviation parameter (SSPDP) for predicting the secondary structure of proteins from amino acid sequence. Deviation parameters (DP) for amino acid singlets, doublets and triplets were computed with respect to secondary structural elements of proteins based on the dictionary of secondary structure prediction (DSSP)-generated secondary structure for 408 selected nonhomologous proteins. To the amino acid triplets which are not found in the selected dataset, a DP value of zero is assigned with respect to the secondary structural elements of proteins. The total number of parameters generated is 15,432, in the possible parameters of 25,260. Deviation parameter is complete with respect to amino acid singlets, doublets, and partially complete with respect to amino acid triplets. These generated parameters were used to predict secondary structural elements from amino acid sequence. The secondary structure predicted by our method (SSPDP) was compared with that of single sequence (NNPREDICT) and multiple sequence (PHD) methods. The average value of the percentage of prediction accuracy for αhelix by SSPDP, NNPREDICT and PHD methods was found to be 57%, 44% and 69% respectively for the proteins in the selected dataset. For Β-strand the prediction accuracy is found to be 69%, 21% and 53% respectively by SSPDP, NNPREDICT and PHD methods. This clearly indicates that the secondary structure prediction by our method is as good as PHD method but much better than NNPREDICT method.     

Reading the three-dimensional structure of lattice model-designed proteins from their amino acid sequence.

While all the information required for the folding of a protein is contained in its amino acid sequence, one has not yet learned how to extract this information to predict the detailed, biological active, three-dimensional structure of a protein whose sequence is known. Using insight obtained from lattice model simulations of the folding of small proteins (fewer than 100 residues), in particular of the fact that this phenomenon is essentially controlled by conserved contacts (Mirny et al., Proc Natl Acad Sci USA 1995;92:1282) among (few) strongly interacting ("hot") amino acids (Tiana et al., J Chem Phys 1998;108:757-761), which also stabilize local elementary structures formed early in the folding process and leading to the (postcritical) folding core when they assemble together (Broglia et al., Proc Natl Acad Sci USA 1998;95:12930, Broglia & Tiana, J Chem Phys 2001;114:7267), we have worked out a successful strategy for reading the three-dimensional structure of lattice model-designed proteins from the knowledge of only their amino acid sequence and of the contact energies among the amino acids.     

A possible way for prediction of domain boundaries in globular proteins from amino acid sequence

A simple approach to domain border prediction in globular proteins is outlined relying on the amino acid sequence only. Statistically determined sequential and association preference data of amino acids were combined to generate short range preference profiles along the polypeptide chains. Domain boundaries correlate with the minima of preference profiles, but some false minima also exist. Possibilities are discussed to exclude the false minima and to further improve the efficiency of the algorithm.     

Ab initio prediction of the three-dimensional structure of a de novo designed protein: a double-blind case study

Ab initio structure prediction and de novo protein design are two problems at the forefront of research in the fields of structural biology and chemistry. The goal of ab initio structure prediction of proteins is to correctly characterize the 3D structure of a protein using only the amino acid sequence as input. De novo protein design involves the production of novel protein sequences that adopt a desired fold. In this work, the results of a double-blind study are presented in which a new ab initio method was successfully used to predict the 3D structure of a protein designed through an experimental approach using binary patterned combinatorial libraries of de novo sequences. The predicted structure, which was produced before the experimental structure was known and without consideration of the design goals, and the final NMR analysis both characterize this protein as a 4-helix bundle. The similarity of these structures is evidenced by both small RMSD values between the coordinates of the two structures and a detailed analysis of the helical packing.     

Protein location prediction using atomic composition and global features of the amino acid sequence

Subcellular location of protein is constructive information in determining its function, screening for drug candidates, vaccine design, annotation of gene products and in selecting relevant proteins for further studies. Computational prediction of subcellular localization deals with predicting the location of a protein from its amino acid sequence. For a computational localization prediction method to be more accurate, it should exploit all possible relevant biological features that contribute to the subcellular localization. In this work, we extracted the biological features from the full length protein sequence to incorporate more biological information. A new biological feature, distribution of atomic composition is effectively used with, multiple physiochemical properties, amino acid composition, three part amino acid composition, and sequence similarity for predicting the subcellular location of the protein. Support Vector Machines are designed for four modules and prediction is made by a weighted voting system. Our system makes prediction with an accuracy of 100, 82.47, 88.81 for self-consistency test, jackknife test and independent data test respectively. Our results provide evidence that the prediction based on the biological features derived from the full length amino acid sequence gives better accuracy than those derived from N-terminal alone. Considering the features as a distribution within the entire sequence will bring out underlying property distribution to a greater detail to enhance the prediction accuracy.     

Predicting DNA-binding sites of proteins from amino acid sequence

Background  

Understanding the molecular details of protein-DNA interactions is critical for deciphering the mechanisms of gene regulation. We present a machine learning approach for the identification of amino acid residues involved in protein-DNA interactions.     

Never born proteins as a test case for ab initio protein structures prediction

The number of natural proteins although large is significantly smaller than the theoretical number of proteins that can be obtained combining the 20 natural amino acids, the so-called “never born proteins” (NBPs). The study of the structure and properties of these proteins allows to investigate the sources of the natural proteins being of unique characteristics or special properties. However the structural study of NPBs can also been intended as an ideal test for evaluating the efficiency of software packages for the ab initio protein structure prediction. In this research, 10.000 three-dimensional structures of proteins of completely random sequence generated according to ROSETTA and FOD model were compared. The results show the limits of these software packages, but at the same time indicate that in many cases there is a significant agreement between the prediction obtained.     

Generation of deviation parameters for amino acid singlets, doublets and triplets from three-dimentional structures of proteins and its implications for secondary structure prediction from amino acid sequences

We present a new method, secondary structure prediction by deviation parameter (SSPDP) for predicting the secondary structure of proteins from amino acid sequence. Deviation parameters (DP) for amino acid singlets, doublets and triplets were computed with respect to secondary structural elements of proteins based on the dictionary of secondary structure prediction (DSSP)-generated secondary structure for 408 selected non-homologous proteins. To the amino acid triplets which are not found in the selected dataset, a DP value of zero is assigned with respect to the secondary structural elements of proteins. The total number of parameters generated is 15,432, in the possible parameters of 25,260. Deviation parameter is complete with respect to amino acid singlets, doublets, and partially complete with respect to amino acid triplets. These generated parameters were used to predict secondary structural elements from amino acid sequence. The secondary structure predicted by our method (SSPDP) was compared with that of single sequence (NNPREDICT) and multiple sequence (PHD) methods. The average value of the percentage of prediction accuracy for a helix by SSPDP, NNPREDICT and PHD methods was found to be 57%, 44% and 69% respectively for the proteins in the selected dataset. For b-strand the prediction accuracy is found to be 69%, 21% and 53% respectively by SSPDP, NNPREDICT and PHD methods. This clearly indicates that the secondary structure prediction by our method is as good as PHD method but much better than NNPREDICT method.     

Real value prediction of solvent accessibility from amino acid sequence

The solvent accessibility of amino acid residues has been predicted in the past by classifying them into exposure states with varying thresholds. This classification provides a wide range of values for the accessible surface area (ASA) within which a residue may fall. Thus far, no attempt has been made to predict real values of ASA from the sequence information without a priori classification into exposure states. Here, we present a new method with which to predict real value ASAs for residues, based on neighborhood information. Our real value prediction neural network could estimate the ASA for four different nonhomologous, nonredundant data sets of varying size, with 18.0-19.5% mean absolute error, defined as per residue absolute difference between the predicted and experimental values of relative ASA. Correlation between the predicted and experimental values ranged from 0.47 to 0.50. It was observed that the ASA of a residue could be predicted within a 23.7% mean absolute error, even when no information about its neighbors is included. Prediction of real values answers the issue of arbitrary choice of ASA state thresholds, and carries more information than category prediction. Prediction error for each residue type strongly correlates with the variability in its experimental ASA values.     

Analysis of protein function and its prediction from amino acid sequence

Understanding protein function is one of the keys to understanding life at the molecular level. It is also important in the context of human disease because many conditions arise as a consequence of alterations of protein function. The recent availability of relatively inexpensive sequencing technology has resulted in thousands of complete or partially sequenced genomes with millions of functionally uncharacterized proteins. Such a large volume of data, combined with the lack of high-throughput experimental assays to functionally annotate proteins, attributes to the growing importance of automated function prediction. Here, we study proteins annotated by Gene Ontology (GO) terms and estimate the accuracy of functional transfer from protein sequence only. We find that the transfer of GO terms by pairwise sequence alignments is only moderately accurate, showing a surprisingly small influence of sequence identity (SID) in a broad range (30-100%). We developed and evaluated a new predictor of protein function, functional annotator (FANN), from amino acid sequence. The predictor exploits a multioutput neural network framework which is well suited to simultaneously modeling dependencies between functional terms. Experiments provide evidence that FANN-GO (predictor of GO terms; available from http://www.informatics.indiana.edu/predrag) outperforms standard methods such as transfer by global or local SID as well as GOtcha, a method that incorporates the structure of GO.     

Membrane proteins: amino acid sequence and membrane penetration

A computer study shows that the membrane-penetrating portion of the erythrocyte surface MN-glycoprotein (Winzler, 1969; Marchesi et al., 1972) is distinguishable by informal cluster analysis from other segments of globular proteins when sequence length is plotted against hydrophobicity This analysis further suggests the possibility that other membrane-penetrating segments of proteins can be identified in the same way.     

Ab initio structure prediction of two alpha-helical oligomers with a multiple-chain united-residue force field and global search

A hierarchical methodology for ab initio structure prediction is extended to treat oligomeric proteins. Modifications are made to a united-residue (UNRES) force field and a Conformational Space Annealing (CSA) global search method. The computational cost of including additional chains and the increase in speed from symmetry optimizations are evaluated. The native structures of two oligomeric proteins from the CASP3 exercise, the retro-GCN4 leucine zipper and the synthetic domain-swapped dimer, were identified as the lowest-energy families resulting from the search of the proteins when rotational symmetry was imposed. Additional searches in different symmetries and oligomerization states were carried out, and the results indicate some problems in the thoroughness of the search and in the search of packing arrangements if symmetry constraints are not imposed.     

Prediction of post-translational glycosylation and phosphorylation of proteins from the amino acid sequence

Post-translational modifications (PTMs) occur on almost all proteins analyzed to date. The function of a modified protein is often strongly affected by these modifications and therefore increased knowledge about the potential PTMs of a target protein may increase our understanding of the molecular processes in which it takes part. High-throughput methods for the identification of PTMs are being developed, in particular within the fields of proteomics and mass spectrometry. However, these methods are still in their early stages, and it is indeed advantageous to cut down on the number of experimental steps by integrating computational approaches into the validation procedures. Many advanced methods for the prediction of PTMs exist and many are made publicly available. We describe our experiences with the development of prediction methods for phosphorylation and glycosylation sites and the development of PTM-specific databases. In addition, we discuss novel ideas for PTM visualization (exemplified by kinase landscapes) and improvements for prediction specificity (by using ESS--evolutionary stable sites). As an example, we present a new method for kinase-specific prediction of phosphorylation sites, NetPhosK, which extends our earlier and more general tool, NetPhos. The new server, NetPhosK, is made publicly available at the URL http://www.cbs.dtu.dk/services/NetPhosK/. The issues of underestimation, over-prediction and strategies for improving prediction specificity are also discussed.     

Computer prediction of potential immunogenic determinants from protein amino acid sequence

In proteins, immunogenic determinants that can induce protein-reactive antipeptide antibodies reside mostly in those parts of the molecule that have a high tendency to form beta-turns. A program for an IBM personal computer which predicts protein immunogenic determinants is described. The program predicts potential immunogenic determinants from protein amino acid sequences according to a Chou-Fasman-based probability of a beta-turn occurrence, p greater than 1.5 X 10(-4)(P. Y. Chou and G. D. Fasman, 1978, Adv. Enzymol. 47, 46-148). Oncopeptides (whose efficacy in generating protein-reactive antipeptide antibodies has been described) with a beta-turn probability of p greater than 1.5 X 10(-4) elicited antipeptide antibodies that reacted with the parent oncoprotein at a rate of 96%, thus showing a surprisingly good correlation between the tendency to form a beta-turn and the protein reactivity of antipeptide antibodies. Potential immunogenic determinants were predicted on myohemerythrin and myoglobin.     

The prediction of the conformation of membrane proteins from the sequence of amino acids.

The methods of Chou & Fasman [Biochemistry (1974) 13, 211-222, 222-245] and of Lim [J. Mol. Biol. (1974)88, 857-872, 873-894] for predicting secondary structure from amino acid sequence have been applied to five predominantly helical membrane-associated peptides. The predictions from the method of Lim (1974a,b) are consistent with the experimental observations, whereas those from Chou & Fasman (1974a,b), although not inconsistent with alpha-helix, favour a beta-structure for several very hydrophobic regions. The results may be rationalized in terms of the effect of the solvent on the conformation of a polypeptide.     

Prediction of RNA binding sites in proteins from amino acid sequence

RNA-protein interactions are vitally important in a wide range of biological processes, including regulation of gene expression, protein synthesis, and replication and assembly of many viruses. We have developed a computational tool for predicting which amino acids of an RNA binding protein participate in RNA-protein interactions, using only the protein sequence as input. RNABindR was developed using machine learning on a validated nonredundant data set of interfaces from known RNA-protein complexes in the Protein Data Bank. It generates a classifier that captures primary sequence signals sufficient for predicting which amino acids in a given protein are located in the RNA-protein interface. In leave-one-out cross-validation experiments, RNABindR identifies interface residues with >85% overall accuracy. It can be calibrated by the user to obtain either high specificity or high sensitivity for interface residues. RNABindR, implementing a Naive Bayes classifier, performs as well as a more complex neural network classifier (to our knowledge, the only previously published sequence-based method for RNA binding site prediction) and offers the advantages of speed, simplicity and interpretability of results. RNABindR predictions on the human telomerase protein hTERT are in good agreement with experimental data. The availability of computational tools for predicting which residues in an RNA binding protein are likely to contact RNA should facilitate design of experiments to directly test RNA binding function and contribute to our understanding of the diversity, mechanisms, and regulation of RNA-protein complexes in biological systems. (RNABindR is available as a Web tool from http://bindr.gdcb.iastate.edu.).     

Prediction of the structural class of globular proteins from their amino acid sequence

Two-dimensional representation of consequence of 32 proteins with known three-dimensional structure has been obtained on 20 X 20 matrix of the distribution of amino acid pairs (nearest neighbours). Prediction algorithm of the structural class of globular proteins has been worked out on the basis of the comparison of 20 X 20 matrix of the distribution of amino acid pairs for the proteins of different structural classes. The accuracy of structural class predictions of 32 proteins has been carried out (all the proteins are taken from numerous ones used to obtain the algorithm).     

Uneven distribution of amino acid substitutions throughout the amino acid sequence of homologous proteins]

A set of aligned homologous protein sequences is divided into two groups consisting of m and n most related sequences. The value of position variability for homologous protein sequences is defined as a number of failures to coincide in the intergroup comparison of all possible m*n pairs of amino acid residues in that position divided by m*n. The position variability value plotted versus the sequence position number with a window of 10 positions gives the intergroup local variability profile. Area S of the figure included between the local variability profile and the straight line corresponding to the mean local variability value is compared with the average area Sr for 1000 random homologous protein families. If S is greater than Sr by more than 2 standard deviation units sigma r, the local variability profile is assumed to contain peaks and hollows corresponding to significant variable and conservative regions of the sequences. The profile extrema containing the area surplus delta S = S-(Sr+ 2 sigma r) are cut off by two straight lines to locate significant regions. The difference (S-Sr) given in standard deviation units sigma r is believed to be the amino acid substitution overall irregularity along the homologous protein sequences OI = (S-Sr)/sigma r. The significant conservative and variable regions of six homologous sequence families (phospholipase A2, cytochromes b, alpha-subunits of Na,K-ATPase, L- and M-subunits of photosynthetic bacteria photoreaction centre and human rhodopsins) were identified. It was shown that for artificial homologous protein sequences derived by k-fold lengthening of natural protein sequences, the OI value rises as square root of k. To compare the degree of substitution irregularity in homologous protein sequence families of different lengths L the value of standard substitution overall irregularity for L = 250 is proposed.