Prediction of MHC class I binding peptides, using SVMHC

Background  

T-cells are key players in regulating a specific immune response. Activation of cytotoxic T-cells requires recognition of specific peptides bound to Major Histocompatibility Complex (MHC) class I molecules. MHC-peptide complexes are potential tools for diagnosis and treatment of pathogens and cancer, as well as for the development of peptide vaccines. Only one in 100 to 200 potential binders actually binds to a certain MHC molecule, therefore a good prediction method for MHC class I binding peptides can reduce the number of candidate binders that need to be synthesized and tested.     

Support Vector Machines for predicting protein structural class

Background  

We apply a new machine learning method, the so-called Support Vector Machine method, to predict the protein structural class. Support Vector Machine method is performed based on the database derived from SCOP, in which protein domains are classified based on known structures and the evolutionary relationships and the principles that govern their 3-D structure.     

Predicting peptides binding to MHC class II molecules using multi-objective evolutionary algorithms

Background  

Peptides binding to Major Histocompatibility Complex (MHC) class II molecules are crucial for initiation and regulation of immune responses. Predicting peptides that bind to a specific MHC molecule plays an important role in determining potential candidates for vaccines. The binding groove in class II MHC is open at both ends, allowing peptides longer than 9-mer to bind. Finding the consensus motif facilitating the binding of peptides to a MHC class II molecule is difficult because of different lengths of binding peptides and varying location of 9-mer binding core. The level of difficulty increases when the molecule is promiscuous and binds to a large number of low affinity peptides.     

Prediction of MHC class II binding affinity using SMM-align,a novel stabilization matrix alignment method

Background  

Antigen presenting cells (APCs) sample the extra cellular space and present peptides from here to T helper cells, which can be activated if the peptides are of foreign origin. The peptides are presented on the surface of the cells in complex with major histocompatibility class II (MHC II) molecules. Identification of peptides that bind MHC II molecules is thus a key step in rational vaccine design and developing methods for accurate prediction of the peptide:MHC interactions play a central role in epitope discovery. The MHC class II binding groove is open at both ends making the correct alignment of a peptide in the binding groove a crucial part of identifying the core of an MHC class II binding motif. Here, we present a novel stabilization matrix alignment method, SMM-align, that allows for direct prediction of peptide:MHC binding affinities. The predictive performance of the method is validated on a large MHC class II benchmark data set covering 14 HLA-DR (human MHC) and three mouse H2-IA alleles.     

Characterization of an endogenous retrovirus class in elephants and their relatives

Background  

Endogenous retrovirus-like elements (ERV-Ls, primed with tRNA leucine) are a diverse group of reiterated sequences related to foamy viruses and widely distributed among mammals. As shown in previous investigations, in many primates and rodents this class of elements has remained transpositionally active, as reflected by increased copy number and high sequence diversity within and among taxa.     

Learning from positive examples when the negative class is undetermined- microRNA gene identification

Background

The application of machine learning to classification problems that depend only on positive examples is gaining attention in the computational biology community. We and others have described the use of two-class machine learning to identify novel miRNAs. These methods require the generation of an artificial negative class. However, designation of the negative class can be problematic and if it is not properly done can affect the performance of the classifier dramatically and/or yield a biased estimate of performance. We present a study using one-class machine learning for microRNA (miRNA) discovery and compare one-class to two-class approaches using naïve Bayes and Support Vector Machines. These results are compared to published two-class miRNA prediction approaches. We also examine the ability of the one-class and two-class techniques to identify miRNAs in newly sequenced species.

Results

Of all methods tested, we found that 2-class naive Bayes and Support Vector Machines gave the best accuracy using our selected features and optimally chosen negative examples. One class methods showed average accuracies of 70–80% versus 90% for the two 2-class methods on the same feature sets. However, some one-class methods outperform some recently published two-class approaches with different selected features. Using the EBV genome as and external validation of the method we found one-class machine learning to work as well as or better than a two-class approach in identifying true miRNAs as well as predicting new miRNAs.

Conclusion

One and two class methods can both give useful classification accuracies when the negative class is well characterized. The advantage of one class methods is that it eliminates guessing at the optimal features for the negative class when they are not well defined. In these cases one-class methods can be superior to two-class methods when the features which are chosen as representative of that positive class are well defined.

Availability

The OneClassmiRNA program is available at: [1]

     

Drug-target interaction prediction via class imbalance-aware ensemble learning

Background

Multiple computational methods for predicting drug-target interactions have been developed to facilitate the drug discovery process. These methods use available data on known drug-target interactions to train classifiers with the purpose of predicting new undiscovered interactions. However, a key challenge regarding this data that has not yet been addressed by these methods, namely class imbalance, is potentially degrading the prediction performance. Class imbalance can be divided into two sub-problems. Firstly, the number of known interacting drug-target pairs is much smaller than that of non-interacting drug-target pairs. This imbalance ratio between interacting and non-interacting drug-target pairs is referred to as the between-class imbalance. Between-class imbalance degrades prediction performance due to the bias in prediction results towards the majority class (i.e. the non-interacting pairs), leading to more prediction errors in the minority class (i.e. the interacting pairs). Secondly, there are multiple types of drug-target interactions in the data with some types having relatively fewer members (or are less represented) than others. This variation in representation of the different interaction types leads to another kind of imbalance referred to as the within-class imbalance. In within-class imbalance, prediction results are biased towards the better represented interaction types, leading to more prediction errors in the less represented interaction types.

Results

We propose an ensemble learning method that incorporates techniques to address the issues of between-class imbalance and within-class imbalance. Experiments show that the proposed method improves results over 4 state-of-the-art methods. In addition, we simulated cases for new drugs and targets to see how our method would perform in predicting their interactions. New drugs and targets are those for which no prior interactions are known. Our method displayed satisfactory prediction performance and was able to predict many of the interactions successfully.

Conclusions

Our proposed method has improved the prediction performance over the existing work, thus proving the importance of addressing problems pertaining to class imbalance in the data.

     

On the global CRISPR array behavior in class I systems

Background

Much effort is underway to build and upgrade databases and tools related to occurrence, diversity, and characterization of CRISPR-Cas systems. As microbial communities and their genome complements are unearthed, much emphasis has been placed on details of individual strains and model systems within the CRISPR-Cas classification, and that collection of information as a whole affords the opportunity to analyze CRISPR-Cas systems from a quantitative perspective to gain insight into distribution of CRISPR array sizes across the different classes, types and subtypes. CRISPR diversity, nomenclature, occurrence, and biological functions have generated a plethora of data that created a need to understand the size and distribution of these various systems to appreciate their features and complexity.

Results

By utilizing a statistical framework and visual analytic techniques, we have been able to test several hypotheses about CRISPR loci in bacterial class I systems. Quantitatively, though CRISPR loci can expand to hundreds of spacers, the mean and median sizes are 40 and 25, respectively, reflecting rather modest acquisition and/or retention overall. Histograms uncovered that CRISPR array size displayed a parametric distribution, which was confirmed by a goodness-of fit test. Mapping the frequency of CRISPR loci on a standardized chromosome plot revealed that CRISPRs have a higher probability of occurring at clustered locations along the positive or negative strand. Lastly, when multiple arrays occur in a particular system, the size of a particular CRISPR array varies with its distance from the cas operon, reflecting acquisition and expansion biases.

Conclusions

This study establishes that bacterial Class I CRISPR array size tends to follow a geometric distribution; these CRISPRs are not randomly distributed along the chromosome; and the CRISPR array closest to the cas genes is typically larger than loci in trans. Overall, we provide an analytical framework to understand the features and behavior of CRISPR-Cas systems through a quantitative lens.

Reviewers

This article was reviewed by Eugene Koonin (NIH-NCBI) and Uri Gophna (Tel Aviv University).

     

Xenorhodopsins, an enigmatic new class of microbial rhodopsins horizontally transferred between archaea and bacteria

Background

Currently a huge amount of protein-protein interaction data is available therefore extracting meaningful ones are a challenging task. In a protein-protein interaction network, hubs are considered as key proteins maintaining function and stability of the network. Therefore, studying protein-protein complexes from a structural perspective provides valuable information for predicted interactions.

Results

In this study, we have predicted by comparative modelling and docking methods protein-protein complexes of hubs of human NR-RTK network inferred from our earlier study. We found that some interactions are mutually excluded while others could occur simultaneously. This study revealed by structural analysis the key role played by Estrogen receptor (ESR1) in mediating the signal transduction between human Receptor Tyrosine kinases (RTKs) and nuclear receptors (NRs).

Conclusions

Although the methods require human intervention and judgment, they can identify the interactions that could occur together or ones that are mutually exclusive. This adds a fourth dimension to interaction network, that of time, and can assist in obtaining concrete predictions consistent with experiments.

Open peer review

This article was reviewed by Dr. Anthony Almudevar, Prof. James Faeder and Prof. Eugene Koonin. For the full reviews, please go to the Reviewers' comments.     

Paths of lateral gene transfer of lysyl-aminoacyl-tRNA synthetases with a unique evolutionary transition stage of prokaryotes coding for class I and II varieties by the same organisms

Background  

While the premise that lateral gene transfer (LGT) is a dominant evolutionary force is still in considerable dispute, the case for widespread LGT in the family of aminoacyl-tRNA synthetases (aaRS) is no longer contentious. aaRSs are ancient enzymes, guarding the fidelity of the genetic code. They are clustered in two structurally unrelated classes. Only lysine aminoacyl-tRNA synthetase (LysRS) is found both as a class 1 and a class 2 enzyme (LysRS1-2). Remarkably, in several extant prokaryotes both classes of the enzyme coexist, a unique phenomenon that has yet to receive its due attention.     

Phorbol myristate acetate and Bryostatin 1 rescue IFN-gamma inducibility of MHC class II molecules in LS1034 colorectal carcinoma cell line

Background  

The expression of major histocompatibility complex class II (MHCII) antigens in both mouse and human tumors is rare, and these antigens are not easily inducible by IFN-gamma (IFNg). Since MHCII may play an important role in the development of host antitumor immune response, we explored the possibility of restoring MHCII inducibility in several IFNg-resistant tumor cell lines using protein kinase C (PKC) agonists phorbol myristate acetate (PMA) or Bryostatin.     

NN-align. An artificial neural network-based alignment algorithm for MHC class II peptide binding prediction

Background  

The major histocompatibility complex (MHC) molecule plays a central role in controlling the adaptive immune response to infections. MHC class I molecules present peptides derived from intracellular proteins to cytotoxic T cells, whereas MHC class II molecules stimulate cellular and humoral immunity through presentation of extracellularly derived peptides to helper T cells. Identification of which peptides will bind a given MHC molecule is thus of great importance for the understanding of host-pathogen interactions, and large efforts have been placed in developing algorithms capable of predicting this binding event.     

Prenatal household size and composition are associated with infant fecal bacterial diversity in Cebu,Philippines

Objectives

The gut microbiome (GM) connects physical and social environments to infant health. Since the infant GM affects immune system development, there is interest in understanding how infants acquire microbes from mothers and other household members.

Materials and Methods

As a part of the Cebu Longitudinal Health and Nutrition Survey (CLHNS), we paired fecal samples (proxy for the GM) collected from infants living in Metro Cebu, Philippines at 2 weeks (N = 39) and 6 months (N = 36) with maternal interviews about prenatal household composition. We hypothesized that relationships between prenatal household size and composition and infant GM bacterial diversity (as measured in fecal samples) would vary by infant age, as well as by household member age and sex. We also hypothesized that infant GM bacterial abundances would differ by prenatal household size and composition.

Results

Data from 16 S rRNA bacterial gene sequencing show that prenatal household size was the most precise estimator of infant GM bacterial diversity, and that the direction of the association between this variable and infant GM bacterial diversity changed between the two time points. The abundances of bacterial families in the infant GM varied by prenatal household variables.

Conclusions

Results highlight the contributions of various household sources to the bacterial diversity of the infant GM, and suggest that prenatal household size is a useful measure for estimating infant GM bacterial diversity in this cohort. Future research should measure the effect of specific sources of household bacterial exposures, including social interactions with caregivers, on the infant GM.

     

Screening of microorganisms producing α-methylserine hydroxymethyltransferase, purification of the enzyme, gene cloning, and application to the enzymatic synthesis of α-methyl-l-serine

Through the screening of microorganisms capable of utilizing α-methylserine, three representative strains belonging to the bacterial genera Paracoccus, Aminobacter, and Ensifer were selected as potent producers of α-methylserine hydroxymethyltransferase, an enzyme that catalyzes the interconversion between α-methyl-l-serine and d-alanine via tetrahydrofolate. Among these strains, Paracoccus sp. AJ110402 was selected as the strain exhibiting the highest α-methylserine hydroxymethyltransferase activity. The enzyme was purified to homogeneity from a cell-free extract of this strain. The native enzyme is a homodimer with apparent molecular mass of 85 kDa and contains 1 mol of pyridoxal-5′-phosphate per mol of the subunit. The K_m for α-methyl-l-serine and tetrahydrofolate was 0.54 mM and 73 μM, respectively. The gene from Paracoccus sp. AJ110402 encoding α-methylserine hydroxymethyltransferase was cloned and expressed in Escherichia coli. Sequence analysis revealed an open reading frame of 1278 bp, encoding a polypeptide with a calculated molecular mass of 46.0 kDa. Using E. coli cells as whole-cell catalysts, 9.7 mmol of α-methyl-l-serine was stereoselectively obtained from 15 mmol of d-alanine and 13.2 mmol of formaldehyde.     

Evolution of the class C GPCR Venus flytrap modules involved positive selected functional divergence

Background  

Class C G protein-coupled receptors (GPCRs) represent a distinct group of the GPCR family, which structurally possess a characteristically distinct extracellular domain inclusive of the Venus flytrap module (VFTM). The VFTMs of the class C GPCRs is responsible for ligand recognition and binding, and share sequence similarity with bacterial periplasmic amino acid binding proteins (PBPs). An extensive phylogenetic investigation of the VFTMs was conducted by analyzing for functional divergence and testing for positive selection for five typical groups of the class C GPCRs. The altered selective constraints were determined to identify the sites that had undergone functional divergence via positive selection. In order to structurally demonstrate the pattern changes during the evolutionary process, three-dimensional (3D) structures of the GPCR VFTMs were modelled and reconstructed from ancestral VFTMs.     

Pooled PCR testing strategy and prevalence estimation of submicroscopic infections using Bayesian latent class models in pregnant women receiving intermittent preventive treatment at Machinga District Hospital,Malawi, 2010

Background

Metarhizium anisopliae is a naturally occurring fungal pathogen of mosquitoes. Recently, Metarhizium has been engineered to act against malaria by directly killing the disease agent within mosquito vectors and also effectively blocking onward transmission. It has been proposed that efforts should be made to minimize the virulence of the fungal pathogen, in order to slow the development of resistant mosquitoes following an actual deployment.

Results

Two mathematical models were developed and analysed to examine the efficacy of the fungal pathogen. It was found that, in many plausible scenarios, the best effects are achieved with a reduced or minimal pathogen virulence, even if the likelihood of resistance to the fungus is negligible. The results for both models depend on the interplay between two main effects: the ability of the fungus to reduce the mosquito population, and the ability of fungus‐infected mosquitoes to compete for resources with non‐fungus‐infected mosquitoes.

Conclusions

The results indicate that there is no obvious choice of virulence for engineered Metarhizium or similar pathogens, and that all available information regarding the population ecology of the combined mosquito‐fungus system should be carefully considered. The models provide a basic framework for examination of anti‐malarial mosquito pathogens that should be extended and improved as new laboratory and field data become available.     

Various hypotheses on MHC evolution suggested by the concerted evolution of CD94L and MHC class Ia molecules

Background  

In the accompanying paper by Virginie Rouillon and myself, our demonstration that homogenisation by gene conversion occurs readily among MHC class I genes was made possible because of the exceptional conservation of the CD94L locus between divergent species of separate taxa, suggesting that the molecules of this family are endowed with very important and well preserved biological functions. These results lead me to elaborate various hypotheses on several aspects of MHC evolution.