Prediction of functional class of novel bacterial proteins without the use of sequence similarity by a statistical learning method

Novel immune-modulator identified by a rapid, functional screen of the parapoxvirus ovis (Orf virus) genome

Background

The success of new sequencing technologies and informatic methods for identifying genes has made establishing gene product function a critical rate limiting step in progressing the molecular sciences. We present a method to functionally mine genomes for useful activities in vivo, using an unusual property of a member of the poxvirus family to demonstrate this screening approach.

Results

The genome of Parapoxvirus ovis (Orf virus) was sequenced, annotated, and then used to PCR-amplify its open-reading-frames. Employing a cloning-independent protocol, a viral expression-library was rapidly built and arrayed into sub-library pools. These were directly delivered into mice as expressible cassettes and assayed for an immune-modulating activity associated with parapoxvirus infection. The product of the B2L gene, a homolog of vaccinia F13L, was identified as the factor eliciting immune cell accumulation at sites of skin inoculation. Administration of purified B2 protein also elicited immune cell accumulation activity, and additionally was found to serve as an adjuvant for antigen-specific responses. Co-delivery of the B2L gene with an influenza gene-vaccine significantly improved protection in mice. Furthermore, delivery of the B2L expression construct, without antigen, non-specifically reduced tumor growth in murine models of cancer.

Conclusion

A streamlined, functional approach to genome-wide screening of a biological activity in vivo is presented. Its application to screening in mice for an immune activity elicited by the pathogen genome of Parapoxvirus ovis yielded a novel immunomodulator. In this inverted discovery method, it was possible to identify the adjuvant responsible for a function of interest prior to a mechanistic study of the adjuvant. The non-specific immune activity of this modulator, B2, is similar to that associated with administration of inactivated particles to a host or to a live viral infection. Administration of B2 may provide the opportunity to significantly impact host immunity while being itself only weakly recognized. The functional genomics method used to pinpoint B2 within an ORFeome may be more broadly applicable to screening for other biological activities in an animal.

Prediction of functional class of proteins and peptides irrespective of sequence homology by support vector machines

Various computational methods have been used for the prediction of protein and peptide function based on their sequences. A particular challenge is to derive functional properties from sequences that show low or no homology to proteins of known function. Recently, a machine learning method, support vector machines (SVM), have been explored for predicting functional class of proteins and peptides from amino acid sequence derived properties independent of sequence similarity, which have shown promising potential for a wide spectrum of protein and peptide classes including some of the low- and non-homologous proteins. This method can thus be explored as a potential tool to complement alignment-based, clustering-based, and structure-based methods for predicting protein function. This article reviews the strategies, current progresses, and underlying difficulties in using SVM for predicting the functional class of proteins. The relevant software and web-servers are described. The reported prediction performances in the application of these methods are also presented.

Accurate prediction of secreted substrates and identification of a conserved putative secretion signal for type III secretion systems

The type III secretion system is an essential component for virulence in many Gram-negative bacteria. Though components of the secretion system apparatus are conserved, its substrates--effector proteins--are not. We have used a novel computational approach to confidently identify new secreted effectors by integrating protein sequence-based features, including evolutionary measures such as the pattern of homologs in a range of other organisms, G+C content, amino acid composition, and the N-terminal 30 residues of the protein sequence. The method was trained on known effectors from the plant pathogen Pseudomonas syringae and validated on a set of effectors from the animal pathogen Salmonella enterica serovar Typhimurium (S. Typhimurium) after eliminating effectors with detectable sequence similarity. We show that this approach can predict known secreted effectors with high specificity and sensitivity. Furthermore, by considering a large set of effectors from multiple organisms, we computationally identify a common putative secretion signal in the N-terminal 20 residues of secreted effectors. This signal can be used to discriminate 46 out of 68 total known effectors from both organisms, suggesting that it is a real, shared signal applicable to many type III secreted effectors. We use the method to make novel predictions of secreted effectors in S. Typhimurium, some of which have been experimentally validated. We also apply the method to predict secreted effectors in the genetically intractable human pathogen Chlamydia trachomatis, identifying the majority of known secreted proteins in addition to providing a number of novel predictions. This approach provides a new way to identify secreted effectors in a broad range of pathogenic bacteria for further experimental characterization and provides insight into the nature of the type III secretion signal.

Prediction of antibiotic resistance proteins from sequence-derived properties irrespective of sequence similarity

Increasing antibiotic resistance has become a worldwide challenge to the clinical treatment of infectious diseases. The identification of antibiotic resistance proteins (ARPs) would be helpful in the discovery of new therapeutic targets and the design of novel drugs to control the potential spread of antibiotic resistance. In this work, a support vector machine (SVM)-based ARP prediction system was developed using 1308 ARPs and 15587 non-ARPs. Its performance was evaluated using 313 ARPs and 7156 non-ARPs. The computed prediction accuracy was 88.5% for ARPs and 99.2% for non-ARPs. A potential application of this method is the identification of ARPs non-homologous to proteins of known function. Further genome screening found that ca. 3.5% and 3.2% of proteins in Escherichia coli and Staphylococcus aureus, respectively, are potential ARPs. These results suggest the usefulness of SVMs for facilitating the identification of ARPs. The software can be accessed at SARPI (Server for Antibiotic Resistance Protein Identification).

Recent progresses in the application of machine learning approach for predicting protein functional class independent of sequence similarity

Protein sequence contains clues to its function. Functional prediction from sequence presents a challenge particularly for proteins that have low or no sequence similarity to proteins of known function. Recently, machine learning methods have been explored for predicting functional class of proteins from sequence-derived properties independent of sequence similarity, which showed promising potential for low- and non-homologous proteins. These methods can thus be explored as potential tools to complement alignment- and clustering-based methods for predicting protein function. This article reviews the strategies, current progresses, and underlying difficulties in using machine learning methods for predicting the functional class of proteins. The relevant software and web-servers are described. The reported prediction performances in the application of these methods are also presented, which need to be interpreted with caution as they are dependent on such factors as datasets used and choice of parameters.

PROFEAT: a web server for computing structural and physicochemical features of proteins and peptides from amino acid sequence

Sequence-derived structural and physicochemical features have frequently been used in the development of statistical learning models for predicting proteins and peptides of different structural, functional and interaction profiles. PROFEAT (Protein Features) is a web server for computing commonly-used structural and physicochemical features of proteins and peptides from amino acid sequence. It computes six feature groups composed of ten features that include 51 descriptors and 1447 descriptor values. The computed features include amino acid composition, dipeptide composition, normalized Moreau-Broto autocorrelation, Moran autocorrelation, Geary autocorrelation, sequence-order-coupling number, quasi-sequence-order descriptors and the composition, transition and distribution of various structural and physicochemical properties. In addition, it can also compute previous autocorrelations descriptors based on user-defined properties. Our computational algorithms were extensively tested and the computed protein features have been used in a number of published works for predicting proteins of functional classes, protein-protein interactions and MHC-binding peptides. PROFEAT is accessible at http://jing.cz3.nus.edu.sg/cgi-bin/prof/prof.cgi.