Investigations on unconventional hydrogen bonds in RNA binding proteins: the role of CH...O=C interactions

We have investigated the roles played by C-H...O=C interactions in RNA binding proteins. There was an average of 78 CH...O=C interactions per protein and also there was an average of one significant CH...O=C interactions for every 6 residues in the 59 RNA binding proteins studied. Main chain-Main chain (MM) CH...O=C interactions are the predominant type of interactions in RNA binding proteins. The donor atom contribution to CH...O=C interactions was mainly from aliphatic residues. The acceptor atom contribution for MM CH...O=C interactions was mainly from Val, Phe, Leu, Ile, Arg and Ala. The secondary structure preference analysis of CH...O=C interacting residues showed that, Arg, Gln, Glu and Tyr preferred to be in helix, while Ala, Asp, Cys, Gly, Ile, Leu, Lys, Met, Phe, Trp and Val preferred to be in strand conformation. Most of the CH...O=C interacting polar amino acid residues were solvent exposed while, majority of the CH...O=C interacting non polar residues were excluded from the solvent. Long and medium-range CH...O=C interactions are the predominant type of interactions in RNA binding proteins. More than 50% of CH...O=C interacting residues had a higher conservation score. Significant percentage of CH...O=C interacting residues had one or more stabilization centers. Sixty-six percent of the theoretically predicted stabilizing residues were also involved in CH...O=C interactions and hence these residues may also contribute additional stability to RNA binding proteins.     

Role of a two-residue spacer in an alpha,beta-Didehydrophenylalanine containing hexapeptide: crystal and solution structure of Boc-val-deltaPhe-Leu-Ala-deltaPhe-Ala-OMe.

The peptide Boc-Val1-deltaPhe2-Leu3-Ala4-deltaPhe5-Ala6-OMe has been examined for the structural consequence of placing a two-residue segment between the deltaPhe residues. The peptide is stabilized by four consecutive beta-turns. The overall conformation of the molecule is a right-handed 3(10)-helix, with average (phi, psi) values (-67.7 degrees, -22.7 degrees), unwound at the C-terminus. The 1H NMR results also suggest that the peptide maintains its 3(10)-helical structure in solution as observed in the crystal state. The crystal structure is stabilized through head-to-tail hydrogen bonds and a repertoire of aromatic interactions laterally directed between adjacent helices, which are antiparallel to each other. The aromatic ring of deltaPhe5 forms the hub of multicentred interactions, namely as a donor in aromatic C-H...pi and aromatic C-H...O=C interactions and as an acceptor in a CH3...pi interaction. The present structure uniquely illustrates the unusual capability of a deltaPhe ring to host such concerted interactions and suggests its exploitation in introducing long-range interactions in the folding of supersecondary structures.     

Role of evolutionary information in prediction of aromatic-backbone NH interactions in proteins

In this study, an attempt has been made to develop a neural network-based method for predicting segments in proteins containing aromatic-backbone NH (Ar-NH) interactions using multiple sequence alignment. We have analyzed 3121 segments seven residues long containing Ar-NH interactions, extracted from 2298 non-redundant protein structures where no two proteins have more than 25% sequence identity. Two consecutive feed-forward neural networks with a single hidden layer have been trained with standard back-propagation as learning algorithm. The performance of the method improves from 0.12 to 0.15 in terms of Matthews correlation coefficient (MCC) value when evolutionary information (multiple alignment obtained from PSI-BLAST) is used as input instead of a single sequence. The performance of the method further improves from MCC 0.15 to 0.20 when secondary structure information predicted by PSIPRED is incorporated in the prediction. The final network yields an overall prediction accuracy of 70.1% and an MCC of 0.20 when tested by five-fold cross-validation. Overall the performance is 15.2% higher than the random prediction. The method consists of two neural networks: (i) a sequence-to-structure network which predicts the aromatic residues involved in Ar-NH interaction from multiple alignment of protein sequences and (ii) a structure-to structure network where the input consists of the output obtained from the first network and predicted secondary structure. Further, the actual position of the donor residue within the 'potential' predicted fragment has been predicted using a separate sequence-to-structure neural network. Based on the present study, a server Ar_NHPred has been developed which predicts Ar-NH interaction in a given amino acid sequence. The web server Ar_NHPred is available at and (mirror site).     

Quantum-chemical analysis of C-H...O and C-H...N interactions in RNA base pairs--H-bond versus anti-H-bond pattern

Geometries and interaction energies of unusual UU and AA base pairs with one standard hydrogen bond (H-bond) and additional C-H...O or C-H...N contacts have been determined by quantum-chemical methods taking into account electron correlation. Whereas the C-H bond length in the UU C-H...O contact increases upon complex formation (H-bond pattern), the C-H bond of the AA C-H....N interaction is shortened (anti-H-bond pattern). The same properties are found for model complexes between U or A and formaldehyde that have intermolecular C-H...acceptor contacts but no standard H-bonds. Both the C-H...acceptor H-bond and anti-H-bond interactions are attractive. A possible influence of the donor CH group charge distribution on the interaction pattern is discussed.     

The occurrence of C--H...O hydrogen bonds in alpha-helices and helix termini in globular proteins

A comprehensive database analysis of C--H...O hydrogen bonds in 3124 alpha-helices and their corresponding helix termini has been carried out from a nonredundant data set of high-resolution globular protein structures resolved at better than 2.0 A in order to investigate their role in the helix, the important protein secondary structural element. The possible occurrence of 5 --> 1 C--H...O hydrogen bond between the ith residue CH group and (i - 4)th residue C==O with C...O < or = 3.8 A is studied, considering as potential donors the main-chain Calpha and the side-chain carbon atoms Cbeta, Cgamma, Cdelta and Cepsilon. Similar analysis has been carried out for 4 --> 1 C--H...O hydrogen bonds, since the C--H...O hydrogen bonds found in helices are predominantly of type 5 --> 1 or 4 --> 1. A total of 17,367 (9310 of type 5 --> 1 and 8057 of type 4 --> 1) C--H...O hydrogen bonds are found to satisfy the selected criteria. The average stereochemical parameters for the data set suggest that the observed C--H...O hydrogen bonds are attractive interactions. Our analysis reveals that the Cgamma and Cbeta hydrogen atom(s) are frequently involved in such hydrogen bonds. A marked preference is noticed for aliphatic beta-branched residue Ile to participate in 5 --> 1 C--H...O hydrogen bonds involving methylene Cgamma 1 atom as donor in alpha-helices. This may be an enthalpic compensation for the greater loss of side-chain conformational entropy for beta-branched amino acids due to the constraint on side-chain torsion angle, namely, chi1, when they occur in helices. The preference of amino acids for 4 --> 1 C--H...O hydrogen bonds is found to be more for Asp, Cys, and for aromatic residues Trp, Phe, and His. Interestingly, overall propensity for C--H...O hydrogen bonds shows that a majority of the helix favoring residues such as Met, Glu, Arg, Lys, Leu, and Gln, which also have large side-chains, prefer to be involved in such types of weak attractive interactions in helices. The amino acid side-chains that participate in C--H...O interactions are found to shield the acceptor carbonyl oxygen atom from the solvent. In addition, C--H...O hydrogen bonds are present along with helix stabilizing salt bridges. A novel helix terminating interaction motif, X-Gly with Gly at C(cap) position having 5 --> 1 Calpha--H...O, and a chain reversal structural motif having 1 --> 5 Calpha-H...O have been identified and discussed. Our analysis highlights that a multitude of local C--H...O hydrogen bonds formed by a variety of amino acid side-chains and Calpha hydrogen atoms occur in helices and more so at the helix termini. It may be surmised that the main-chain Calpha and the side-chain CH that participate in C--H...O hydrogen bonds collectively augment the cohesive energy and thereby contribute together with the classical N--H...O hydrogen bonds and other interactions to the overall stability of helix and therefore of proteins.     

Prediction of continuous B-cell epitopes in an antigen using recurrent neural network

B-cell epitopes play a vital role in the development of peptide vaccines, in diagnosis of diseases, and also for allergy research. Experimental methods used for characterizing epitopes are time consuming and demand large resources. The availability of epitope prediction method(s) can rapidly aid experimenters in simplifying this problem. The standard feed-forward (FNN) and recurrent neural network (RNN) have been used in this study for predicting B-cell epitopes in an antigenic sequence. The networks have been trained and tested on a clean data set, which consists of 700 non-redundant B-cell epitopes obtained from Bcipep database and equal number of non-epitopes obtained randomly from Swiss-Prot database. The networks have been trained and tested at different input window length and hidden units. Maximum accuracy has been obtained using recurrent neural network (Jordan network) with a single hidden layer of 35 hidden units for window length of 16. The final network yields an overall prediction accuracy of 65.93% when tested by fivefold cross-validation. The corresponding sensitivity, specificity, and positive prediction values are 67.14, 64.71, and 65.61%, respectively. It has been observed that RNN (JE) was more successful than FNN in the prediction of B-cell epitopes. The length of the peptide is also important in the prediction of B-cell epitopes from antigenic sequences. The webserver ABCpred is freely available at www.imtech.res.in/raghava/abcpred/.     

GXXXG and AXXXA: common alpha-helical interaction motifs in proteins, particularly in extremophiles

The GXXXG motif is a frequently occurring sequence of residues that is known to favor helix-helix interactions in membrane proteins. Here we show that the GXXXG motif is also prevalent in soluble proteins whose structures have been determined. Some 152 proteins from a non-redundant PDB set contain at least one alpha-helix with the GXXXG motif, 41 +/- 9% more than expected if glycine residues were uniformly distributed in those alpha-helices. More than 50% of the GXXXG-containing alpha-helices participate in helix-helix interactions. In fact, 26 of those helix-helix interactions are structurally similar to the helix-helix interaction of the glycophorin A dimer, where two transmembrane helices associate to form a dimer stabilized by the GXXXG motif. As for the glycophorin A structure, we find backbone-to-backbone atomic contacts of the C alpha-H...O type in each of these 26 helix-helix interactions that display the stereochemical hallmarks of hydrogen bond formation. These glycophorin A-like helix-helix interactions are enriched in the general set of helix-helix interactions containing the GXXXG motif, suggesting that the inferred C alpha-H...O hydrogen bonds stabilize the helix-helix interactions. In addition to the GXXXG motif, some 808 proteins from the non-redundant PDB set contain at least one alpha-helix with the AXXXA motif (30 +/- 3% greater than expected). Both the GXXXG and AXXXA motifs occur frequently in predicted alpha-helices from 24 fully sequenced genomes. Occurrence of the AXXXA motif is enhanced to a greater extent in thermophiles than in mesophiles, suggesting that helical interaction based on the AXXXA motif may be a common mechanism of thermostability in protein structures. We conclude that the GXXXG sequence motif stabilizes helix-helix interactions in proteins, and that the AXXXA sequence motif also stabilizes the folded state of proteins.     

Tryptophan-containing peptide helices: interactions involving the indole side chain.

Two designed peptide sequences containing Trp residues at positions i and i + 5 (Boc-Leu-Trp-Val-Ala-Aib-Leu-Trp-Val-OMe, 1) as well as i and i + 6 (Boc-Leu-Trp-Val-Aib-Ala-Aib-Leu-Trp-Val-OMe, 2) containing one and two centrally positioned Aib residues, respectively, for helix nucleation, have been shown to form stable helices in chloroform solutions. Structures derived from nuclear magnetic resonance (NMR) data reveal six and seven intramolecularly hydrogen-bonded NH groups in peptides 1 and 2, respectively. The helical conformation of octapeptide 1 has also been established in the solid state by X-ray diffraction. The crystal structure reveals an interesting packing motif in which helical columns are stabilized by side chain-backbone hydrogen bonding involving the indole Nepsilon1H of Trp(2) as donor, and an acceptor C=O group from Leu(6) of a neighboring molecule. Helical columns also associate laterally, and strong interactions are observed between the Trp(2) and Trp(7) residues on neighboring molecules. The edge-to-face aromatic interactions between the indoles suggest a potential C-H...pi interaction involving the Czeta3H of Trp(2). Concentration dependence of NMR chemical shifts provides evidence for peptide association in solution involving the Trp(2) Nepsilon1H protons, presumably in a manner similar to that observed in the crystal.     

Stereospecific interactions of proline residues in protein structures and complexes

The constrained backbone torsion angle of a proline (Pro) residue has usually been invoked to explain its three-dimensional context in proteins. Here we show that specific interactions involving the pyrrolidine ring atoms also contribute to its location in a given secondary structure and its binding to another molecule. It is adept at participating in two rather non-conventional interactions, C-H...pi and C-H...O. The geometry of interaction between the pyrrolidine and aromatic rings, vis-à-vis the occurrence of the C-H...pi interactions has been elucidated. Some of the secondary structural elements stabilized by Pro-aromatic interactions are beta-turns, where a Pro can interact with an adjacent aromatic residue, and in antiparallel beta-sheet, where a Pro in an edge strand can interact with an aromatic residue in the adjacent strand at a non-hydrogen-bonded site. The C-H groups at the Calpha and Cdelta positions can form strong C-H...O interactions (as seen from the clustering of points) and such interactions involving a Pro residue at C' position relative to an alpha-helix can cap the hydrogen bond forming potentials of the free carbonyl groups at the helix C terminus. Functionally important Pro residues occurring at the binding site of a protein almost invariably engage aromatic residues (with one of them being held by C-H...pi interaction) from the partner molecule in the complex, and such aromatic residues are highly conserved during evolution.     

C-H...pi-interactions in proteins

A non-redundant set of 1154 protein structures from the Protein Data Bank was examined with respect to close interactions between C-H-donor and pi-acceptor groups. A total of 31,087 interactions were found to satisfy our selection criteria. Their geometric parameters suggest that these interactions can be classified as weak hydrogen bonds.A set of 12 interaction classes were defined based on the division of the donors into three groups and the acceptors into four groups. These classes were examined separately, and the respective interactions described in detail in each class. Most prominent were interactions between aliphatic C-H donors and aromatic pi-acceptors and interactions between aromatic C-H donors and aromatic pi-acceptors. About three-quarters of the Trp-rings, half of all Phe and Tyr-rings and a quarter of all His-rings were found to be involved as acceptors in C-H...pi-interactions. On the donor side, a preference for aromatic C-H groups was observed, but also for the aliphatic side-chains of the long, extended amino acid residues Lys, Arg and Met, and the Pro ring.The average distance between the C-donor and the center-of-mass of the pi-acceptor was observed to be significantly longer in the 174 protein structures determined at >2.5 A resolution. Also, the distribution is significantly wider. This resolution dependence suggests that the force fields commonly used for the refinement of protein structures may not be adequate. C-H...pi-interactions involving aromatic groups either as donor or as acceptor groups are found mostly in the interior of the protein. The more hydrophilic the participating groups are, the closer to the surface are the interactions located. About 40 % of all C-H...pi-interactions occur between amino acid residue side-chains that are separated by nine or less residues in sequence. Dependent on the interaction class, different preferences for secondary structure, residue type and side-chain conformation were observed.It is likely that the C-H...pi-interactions contribute significantly to the overall stability of a protein.     

Peptide design using alpha,beta-dehydro amino acids: from beta-turns to helical hairpins

Incorporation of alpha,beta-dehydrophenylalanine (DeltaPhe) residue in peptides induces folded conformations: beta-turns in short peptides and 3(10)-helices in larger ones. A few exceptions-namely, alpha-helix or flat beta-bend ribbon structures-have also been reported in a few cases. The most favorable conformation of DeltaPhe residues are (phi,psi) approximately (-60 degrees, -30 degrees ), (-60 degrees, 150 degrees ), (80 degrees, 0 degrees ) or their enantiomers. DeltaPhe is an achiral and planar residue. These features have been exploited in designing DeltaPhe zippers and helix-turn-helix motifs. DeltaPhe can be incorporated in both right and left-handed helices. In fact, consecutive occurrence of three or more DeltaPhe amino acids induce left-handed screw sense in peptides containing L-amino acids. Weak interactions involving the DeltaPhe residue play an important role in molecular association. The C--H.O==C hydrogen bond between the DeltaPhe side-chain and backbone carboxyl moiety, pi-pi stacking interactions between DeltaPhe side chains belonging to enantiomeric helices have shown to stabilize folding. The unusual capability of a DeltaPhe ring to form the hub of multicentered interactions namely, a donor in aromatic C--H.pi and C--H.O==C and an acceptor in a CH(3).pi interaction suggests its exploitation in introducing long-range interactions in the folding of supersecondary structures.     

Strong and weak hydrogen bonds in the protein-ligand interface

The characteristics of N--H...O, O--H...O, and C--H...O hydrogen bonds and other weak intermolecular interactions are analyzed in a large and diverse group of 251 protein-ligand complexes using a new computer program that was developed in-house for this purpose. The interactions examined in the present study are those which occur in the active sites, defined here as a sphere of 10 A radius around the ligand. Notably, N--H...O and O--H...O bonds tend towards linearity. Multifurcated interactions are especially common, especially multifurcated acceptors, and the average degree of furcation is 2.6 hydrogen bonds per furcated acceptor. A significant aspect of this study is that we have been able to assess the reliability of hydrogen bond geometry as a function of crystallographic resolution. Thresholds of 2.3 and 2.0 A are established for strong and weak hydrogen bonds, below which hydrogen bond geometries may be safely considered for detailed analysis. Interactions involving water as donor or acceptor, and C--H...O bonds with Gly and Tyr as donors are ubiquitous in the active site. A similar trend was observed in an external test set of 233 protein-ligand complexes belonging to the kinase family. Weaker interactions like X--H...pi (X = C, N, O) and those involving halogen atoms as electrophiles or nucleophiles have also been studied. We conclude that the strong and weak hydrogen bonds are ubiquitous in protein-ligand recognition, and that with suitable computational tools very large numbers of strong and weak intermolecular interactions in the ligand-protein interface may be analyzed reliably. Results confirm earlier trends reported previously by us but the extended nature of the present data set mean that the observed trends are more reliable.     

Investigations on C-H...pi interactions in RNA binding proteins

We have investigated the roles played by C-Hcdots, three dots, centeredpi interactions in RNA binding proteins. There was an average of 55 C-Hcdots, three dots, centeredpi interactions per protein and also there was an average of one significant C-Hcdots, three dots, centeredpi interaction for every nine residues in the 59 RNA binding proteins studied. Main-chain to side-chain C-Hcdots, three dots, centeredpi interactions is the predominant type of interactions in RNA binding proteins. The donor atom contribution to C-Hcdots, three dots, centeredpi interactions was mainly from Phe, Tyr, Trp, Pro, Gly, Lys, His and Ala residues. The acceptor atom contribution to main-chain to side-chain C-Hcdots, three dots, centeredpi and side-chain to side-chain C-Hcdots, three dots, centeredpi interactions was mainly from Phe and Tyr residues. On the contrary, the acceptor atoms of Trp residues contributed to all the four types of C-Hcdots, three dots, centeredpi interactions. Also, Trp contributed both donor and acceptor atoms in main-chain to side-chain, main-chain to side-chain five-member aromatic ring and side-chain to side-chain C-Hcdots, three dots, centeredpi interactions. The secondary structure preference analysis of C-Hcdots, three dots, centeredpi interacting residues showed that, Arg, Gln, Glu, His, Ile, Leu, Lys, Met, Phe and Tyr preferred to be in helix, while Ala, Asp, Cys, Gly, Trp and Val preferred to be in strand conformation. Long-range C-Hcdots, three dots, centeredpi interactions are the predominant type of interactions in RNA binding proteins. More than 50% of C-Hcdots, three dots, centeredpi interacting residues had a higher conservation score. Significant percentage of C-Hcdots, three dots, centeredpi interacting residues had one or more stabilization centers. Seven percent of the theoretically predicted stabilizing residues were also involved in C-Hcdots, three dots, centeredpi interactions and hence these residues may also contribute additional stability to RNA binding proteins.     

Aromatic-aromatic interactions in and around alpha-helices

To understand the role of aromatic-aromatic interactions in imparting specificity to the folding process, the geometries of four aromatic residues with different sequence spacing, located in alpha-helices or five residues from helical ends, interacting with each other have been elucidated. The geometry is found to depend on the sequence difference. Specific interactions (C-H...pi and N-H...pi) which result from this geometry may cause a given pair of residues (such as Phe-His) with a particular sequence difference to occur more than expected. The most conspicuous residue in an aromatic pair in the context of helix stability is His, which is found at the last (C1) position or the two positions (Ncap and Ccap) immediately flanking the helix. An alpha-helix and a contiguous 3(10)-helix or two helices separated by a non-helical residue can have interacting aromatic pairs, the geometry of interaction and the relative orientation between the helices being rather fixed. Short helices can also have interacting residues from either side.     

C-H...O hydrogen bonds in the nuclear receptor RARgamma--a potential tool for drug selectivity

Hydrogen bonds between polarized atoms play a crucial role in protein interactions and are often used in drug design, which usually neglects the potential of C-H...O hydrogen bonds. The 1.4 A resolution crystal structure of the ligand binding domain of the retinoic acid receptor RARgamma complexed with the retinoid SR11254 reveals several types of C-H...O hydrogen bonds. A striking example is the hydroxyl group of the ligand that acts as an H bond donor and acceptor, leading to a synergy between classical and C-H...O hydrogen bonds. This interaction introduces both specificity and affinity within the hydrophobic ligand pocket. The similarity of intraprotein and protein-ligand C-H...O interactions suggests that such bonds should be considered in rational drug design approaches.     

The contribution of C alpha-H...O hydrogen bonds to membrane protein stability depends on the position of the amide

Structural analyses of membrane proteins reveal a large number of C(alpha)-H...O contacts between transmembrane helices, presumed to be hydrogen bonds. Recent experiments produced conflicting results for the contribution of such hydrogen bonds to membrane protein stability. An FTIR study estimated an energy of -0.88 kcal/mol for the G79-C(alpha)-H...I76-O hydrogen bond in glycophorin A, whereas a mutagenesis study showed that the A51-C(alpha)-H...T24-O(gamma) hydrogen bond does not stabilize bacteriorhodopsin. Here, we reconcile these results using molecular mechanics calculations and an implicit membrane model (IMM1). With explicit hydrogen atoms, the potential energy of the G79-C(alpha)-H...I76-O interaction in GpA ranges from -0.54 to -0.9 kcal/mol and its contribution to stability (effective energy) from -0.49 to -0.83 kcal/mol, depending on the structural model used. The average values of these quantities in GpA-like motifs are similar. In bR, the corresponding numbers for the A51-C(alpha)-H...T24-O(gamma) interaction are +0.15 and +0.32 kcal/mol. The difference results from the different arrangement of the interacting groups and specifically the position of the acceptor with respect to the C(alpha) and N atoms. This conclusion likely applies to soluble proteins as well.     

Prediction of human mRNA donor and acceptor sites from the DNA sequence

Artificial neural networks have been applied to the prediction of splice site location in human pre-mRNA. A joint prediction scheme where prediction of transition regions between introns and exons regulates a cutoff level for splice site assignment was able to predict splice site locations with confidence levels far better than previously reported in the literature. The problem of predicting donor and acceptor sites in human genes is hampered by the presence of numerous amounts of false positives: here, the distribution of these false splice sites is examined and linked to a possible scenario for the splicing mechanism in vivo. When the presented method detects 95% of the true donor and acceptor sites, it makes less than 0.1% false donor site assignments and less than 0.4% false acceptor site assignments. For the large data set used in this study, this means that on average there are one and a half false donor sites per true donor site and six false acceptor sites per true acceptor site. With the joint assignment method, more than a fifth of the true donor sites and around one fourth of the true acceptor sites could be detected without accompaniment of any false positive predictions. Highly confident splice sites could not be isolated with a widely used weight matrix method or by separate splice site networks. A complementary relation between the confidence levels of the coding/non-coding and the separate splice site networks was observed, with many weak splice sites having sharp transitions in the coding/non-coding signal and many stronger splice sites having more ill-defined transitions between coding and non-coding.     

Detection of N-H...N hydrogen bonding in RNA via scalar couplings in the absence of observable imino proton resonances

Hydrogen bond networks stabilize RNA secondary and tertiary structure and are thus essentially important for protein recognition. During structure refinements using either NMR or X-ray techniques, hydrogen bonds were usually inferred indirectly from the proximity of donor and acceptor functional groups. Recently, quantitative heteronuclear J(N,N)-HNN COSY NMR experiments were introduced that allowed the direct identification of donor and acceptor nitrogen atoms involved in hydrogen bonds. However, protons involved in base pairing interactions in nucleic acids are often not observable due to exchange processes. The application of a modified quantitative J(N,N)-HNN COSY pulse scheme permits observation of ^2hJ(N,N) couplings via non-exchangeable protons. This approach allowed the unambiguous identification of the A27·U23 reverse Hoogsteen base pair involved in a U-A·U base triple in the HIV-2 transactivation response element–argininamide complex. Despite a wealth of NOE information, direct evidence for this interaction was lacking due to the rapid exchange of the U23 imino proton. The ability to directly observe hydrogen bonds, even in D₂O and in the presence of rapid exchange, should facilitate structural studies of RNA.     

Non-classical hydrogen bonds in interleukins: the role of C-H...O interactions

The role of classical hydrogen bonds in the structural stability of biological macro-molecules is well understood. In the present study, we explore the influence of C-H...O interactions in relation to other environmental preferences in interleukins. Main chain-main chain interactions are predominant. Pro residues might stabilize helices and strands by C-H...O H-bonds in interleukins. Majority of the C-H...O interacting residues were solvent exposed. 62% of C-H...O interactions was long-range interactions. The results presented in this study might be useful for structural stability studies in interleukins.     

Inference of genetic regulatory networks with recurrent neural network models using particle swarm optimization

Genetic regulatory network inference is critically important for revealing fundamental cellular processes, investigating gene functions, and understanding their relations. The availability of time series gene expression data makes it possible to investigate the gene activities of whole genomes, rather than those of only a pair of genes or among several genes. However, current computational methods do not sufficiently consider the temporal behavior of this type of data and lack the capability to capture the complex nonlinear system dynamics. We propose a recurrent neural network (RNN) and particle swarm optimization (PSO) approach to infer genetic regulatory networks from time series gene expression data. Under this framework, gene interaction is explained through a connection weight matrix. Based on the fact that the measured time points are limited and the assumption that the genetic networks are usually sparsely connected, we present a PSO-based search algorithm to unveil potential genetic network constructions that fit well with the time series data and explore possible gene interactions. Furthermore, PSO is used to train the RNN and determine the network parameters. Our approach has been applied to both synthetic and real data sets. The results demonstrate that the RNN/PSO can provide meaningful insights in understanding the nonlinear dynamics of the gene expression time series and revealing potential regulatory interactions between genes.