Prediction of cis/trans isomerization using feature selection and support vector machines

Detecting Proline and Non-Proline Cis Isomers in Protein Structures from Sequences Using Deep Residual Ensemble Learning

It has been long established that cis conformations of amino acid residues play many biologically important roles despite their rare occurrence in protein structure. Because of this rarity, few methods have been developed for predicting cis isomers from protein sequences, most of which are based on outdated datasets and lack the means for independent testing. In this work, using a database of >10000 high-resolution protein structures, we update the statistics of cis isomers and develop a sequence-based prediction technique using an ensemble of residual convolutional and long short-term memory bidirectional recurrent neural networks that allow learning from the whole protein sequence. We show that ensembling eight neural network models yields maximum Matthews correlation coefficient values of approximately 0.35 for cis-Pro isomers and 0.1 for cis-nonPro residues. The method should be useful for prioritizing functionally important residues in cis isomers for experimental validations and improving the sampling of rare protein conformations for ab initio protein structure prediction.

Sixty-five years of the long march in protein secondary structure prediction: the final stretch?

Protein secondary structure prediction began in 1951 when Pauling and Corey predicted helical and sheet conformations for protein polypeptide backbone even before the first protein structure was determined. Sixty-five years later, powerful new methods breathe new life into this field. The highest three-state accuracy without relying on structure templates is now at 82-84%, a number unthinkable just a few years ago. These improvements came from increasingly larger databases of protein sequences and structures for training, the use of template secondary structure information and more powerful deep learning techniques. As we are approaching to the theoretical limit of three-state prediction (88-90%), alternative to secondary structure prediction (prediction of backbone torsion angles and Cα-atom-based angles and torsion angles) not only has more room for further improvement but also allows direct prediction of three-dimensional fragment structures with constantly improved accuracy. About 20% of all 40-residue fragments in a database of 1199 non-redundant proteins have <6 Å root-mean-squared distance from the native conformations by SPIDER2. More powerful deep learning methods with improved capability of capturing long-range interactions begin to emerge as the next generation of techniques for secondary structure prediction. The time has come to finish off the final stretch of the long march towards protein secondary structure prediction.

Molecular Dynamics Information Improves cis-Peptide-Based Function Annotation of Proteins

cis-Peptide bonds, whose occurrence in proteins is rare but evolutionarily conserved, are implicated to play an important role in protein function. This has led to their previous use in a homology-independent, fragment-match-based protein function annotation method. However, proteins are not static molecules; dynamics is integral to their activity. This is nicely epitomized by the geometric isomerization of cis-peptide to trans form for molecular activity. Hence we have incorporated both static (cis-peptide) and dynamics information to improve the prediction of protein molecular function. Our results show that cis-peptide information alone cannot detect functional matches in cases where cis-trans isomerization exists but 3D coordinates have been obtained for only the trans isomer or when the cis-peptide bond is incorrectly assigned as trans. On the contrary, use of dynamics information alone includes false-positive matches for cases where fragments with similar secondary structure show similar dynamics, but the proteins do not share a common function. Combining the two methods reduces errors while detecting the true matches, thereby enhancing the utility of our method in function annotation. A combined approach, therefore, opens up new avenues of improving existing automated function annotation methodologies.

Detection of trans-cis flips and peptide-plane flips in protein structures

A coordinate-based method is presented to detect peptide bonds that need correction either by a peptide-plane flip or by a trans-cis inversion of the peptide bond. When applied to the whole Protein Data Bank, the method predicts 4617 trans-cis flips and many thousands of hitherto unknown peptide-plane flips. A few examples are highlighted for which a correction of the peptide-plane geometry leads to a correction of the understanding of the structure-function relation. All data, including 1088 manually validated cases, are freely available and the method is available from a web server, a web-service interface and through WHAT_CHECK.

Randomized Subspace Learning for Proline Cis-Trans Isomerization Prediction

Proline residues are common source of kinetic complications during folding. The X-Pro peptide bond is the only peptide bond for which the stability of the cis and trans conformations is comparable. The cis-trans isomerization (CTI) of X-Pro peptide bonds is a widely recognized rate-limiting factor, which can not only induces additional slow phases in protein folding but also modifies the millisecond and sub-millisecond dynamics of the protein. An accurate computational prediction of proline CTI is of great importance for the understanding of protein folding, splicing, cell signaling, and transmembrane active transport in both the human body and animals. In our earlier work, we successfully developed a biophysically motivated proline CTI predictor utilizing a novel tree-based consensus model with a powerful metalearning technique and achieved 86.58 percent Q2 accuracy and 0.74 Mcc, which is a better result than the results (70-73 percent Q2 accuracies) reported in the literature on the well-referenced benchmark dataset. In this paper, we describe experiments with novel randomized subspace learning and bootstrap seeding techniques as an extension to our earlier work, the consensus models as well as entropy-based learning methods, to obtain better accuracy through a precise and robust learning scheme for proline CTI prediction.

Intelligent Consensus Modeling for Proline Cis-Trans Isomerization Prediction

Proline cis-trans isomerization (CTI) plays a key role in the rate-determining steps of protein folding. Accurate prediction of proline CTI is of great importance for the understanding of protein folding, splicing, cell signaling, and transmembrane active transport in both the human body and animals. Our goal is to develop a state-of-the-art proline CTI predictor based on a biophysically motivated intelligent consensus modeling through the use of sequence information only (i.e., position specific scores generated by PSI-BLAST). The current computational proline CTI predictors reach about 70-73 percent Q2 accuracies and about 0.40 Matthew correlation coefficient (Mcc) through the use of sequence-based evolutionary information as well as predicted protein secondary structure information. However, our approach that utilizes a novel decision tree-based consensus model with a powerful randomized-metal earning technique has achieved 86.58 percent Q2 accuracy and 0.74 Mcc, on the same proline CTI data set, which is a better result than those of any existing computational proline CTI predictors reported in the literature.

Cis-trans isomerization of omega dihedrals in proteins

Peptide bonds in protein structures are mainly found in trans conformation with a torsion angle ω close to 180°. Only a very low proportion is observed in cis conformation with ω angle around 0°. Cis-trans isomerization leads to local conformation changes which play an important role in many biological processes. In this paper, we reviewed the recent discoveries and research achievements in this field. First, we presented some interesting cases of biological processes in which cis-trans isomerization is directly implicated. It is involved in protein folding and various aspect of protein function like dimerization interfaces, autoinhibition control, channel gating, membrane binding. Then we reviewed conservation studies of cis peptide bonds which emphasized evolution constraints in term of sequence and local conformation. Finally we made an overview of the numerous molecular dynamics studies and prediction methodologies already developed to take into account this structural feature in the research area of protein modeling. Many cis peptide bonds have not been recognized as such due to the limited resolution of the data and to the refinement protocol used. Cis-trans proline isomerization reactions represents a vast and promising research area that still needs to be further explored for a better understanding of isomerization mechanism and improvement of cis peptide bond predictions.

The alphabet of intrinsic disorder: I. Act like a Pro: On the abundance and roles of proline residues in intrinsically disordered proteins

A significant fraction of every proteome is occupied by biologically active proteins that do not form unique three-dimensional structures. These intrinsically disordered proteins (IDPs) and IDP regions (IDPRs) have essential biological functions and are characterized by extensive structural plasticity. Such structural and functional behavior is encoded in the amino acid sequences of IDPs/IDPRs, which are enriched in disorder-promoting residues and depleted in order-promoting residues. In fact, amino acid residues can be arranged according to their disorder-promoting tendency to form an alphabet of intrinsic disorder that defines the structural complexity and diversity of IDPs/IDPRs. This review is the first in a series of publications dedicated to the roles that different amino acid residues play in defining the phenomenon of protein intrinsic disorder. We start with proline because data suggests that of the 20 common amino acid residues, this one is the most disorder-promoting.

Extraction of consensus protein patterns in regions containing non-proline cis peptide bonds and their functional assessment

Background

In peptides and proteins, only a small percentile of peptide bonds adopts the cis configuration. Especially in the case of amide peptide bonds, the amount of cis conformations is quite limited thus hampering systematic studies, until recently. However, lately the emerging population of databases with more 3D structures of proteins has produced a considerable number of sequences containing non-proline cis formations (cis-nonPro).

Results

In our work, we extract regular expression-type patterns that are descriptive of regions surrounding the cis-nonPro formations. For this purpose, three types of pattern discovery are performed: i) exact pattern discovery, ii) pattern discovery using a chemical equivalency set, and iii) pattern discovery using a structural equivalency set. Afterwards, using each pattern as predicate, we search the Eukaryotic Linear Motif (ELM) resource to identify potential functional implications of regions with cis-nonPro peptide bonds. The patterns extracted from each type of pattern discovery are further employed, in order to formulate a pattern-based classifier, which is used to discriminate between cis-nonPro and trans-nonPro formations.

Conclusions

In terms of functional implications, we observe a significant association of cis-nonPro peptide bonds towards ligand/binding functionalities. As for the pattern-based classification scheme, the highest results were obtained using the structural equivalency set, which yielded 70% accuracy, 77% sensitivity and 63% specificity.

Computing Relative Free Energies of Solvation using Single Reference Thermodynamic Integration Augmented with Hamiltonian Replica Exchange

This paper introduces an efficient single-topology variant of Thermodynamic Integration (TI) for computing relative transformation free energies in a series of molecules with respect to a single reference state. The presented TI variant that we refer to as Single-Reference TI (SR-TI) combines well-established molecular simulation methodologies into a practical computational tool. Augmented with Hamiltonian Replica Exchange (HREX), the SR-TI variant can deliver enhanced sampling in select degrees of freedom. The utility of the SR-TI variant is demonstrated in calculations of relative solvation free energies for a series of benzene derivatives with increasing complexity. Noteworthy, the SR-TI variant with the HREX option provides converged results in a challenging case of an amide molecule with a high (13-15 kcal/mol) barrier for internal cis/trans interconversion using simulation times of only 1 to 4 ns.

PBOND: web server for the prediction of proline and non-proline cis/trans isomerization

PBOND is a web server that predicts the conformation of the peptide bond between any two amino acids. PBOND classifies the peptide bonds into one out of four classes, namely cis imide (cis-Pro), cis amide (cis-nonPro), trans imide (trans-Pro) and trans amide (trans-nonPro). Moreover, for every prediction a reliability index is computed. The underlying structure of the server consists of three stages: (1) feature extraction, (2) feature selection and (3) peptide bond classification. PBOND can handle both single sequences as well as multiple sequences for batch processing. The predictions can either be directly downloaded from the web site or returned via e-mail. The PBOND web server is freely available at http://195.251.198.21/pbond.html.

Detection of discriminative sequence patterns in the neighborhood of proline cis peptide bonds and their functional annotation

Background

Polypeptides are composed of amino acids covalently bonded via a peptide bond. The majority of peptide bonds in proteins is found to occur in the trans conformation. In spite of their infrequent occurrence, cis peptide bonds play a key role in the protein structure and function, as well as in many significant biological processes.

Results

We perform a systematic analysis of regions in protein sequences that contain a proline cis peptide bond in order to discover non-random associations between the primary sequence and the nature of proline cis/trans isomerization. For this purpose an efficient pattern discovery algorithm is employed which discovers regular expression-type patterns that are overrepresented (i.e. appear frequently repeated) in a set of sequences. Four types of pattern discovery are performed: i) exact pattern discovery, ii) pattern discovery using a chemical equivalency set, iii) pattern discovery using a structural equivalency set and iv) pattern discovery using certain amino acids' physicochemical properties. The extracted patterns are carefully validated using a specially implemented scoring function and a significance measure (i.e. log-probability estimate) indicative of their specificity. The score threshold for the first three types of pattern discovery is 0.90 while for the last type of pattern discovery 0.80. Regarding the significance measure, all patterns yielded values in the range [-9, -31] which ensure that the derived patterns are highly unlikely to have emerged by chance. Among the highest scoring patterns, most of them are consistent with previous investigations concerning the neighborhood of cis proline peptide bonds, and many new ones are identified. Finally, the extracted patterns are systematically compared against the PROSITE database, in order to gain insight into the functional implications of cis prolyl bonds.

Conclusion

Cis patterns with matches in the PROSITE database fell mostly into two main functional clusters: family signatures and protein signatures. However considerable propensity was also observed for targeting signals, active and phosphorylation sites as well as domain signatures.