Narrow Results

In the study of in silico functional genomics, improving the performance of protein function prediction is the ultimate goal for identifying proteins associated with defined cellular functions. The classical prediction approach is to employ pairwise sequence alignments. However this method often faces difficulties when no statistically significant homologous sequences are identified. An alternative way is to predict protein function from sequence-derived features using machine learning. In this case the choice of possible features which can be derived from the sequence is of vital importance to ensure adequate discrimination to predict function. In this paper we have successfully selected biologically significant features for protein function prediction. This was performed using a new feature selection method (FrankSum) that avoids data distribution assumptions, uses a data independent measurement (p-value) within the feature, identifies redundancy between features and uses an appropiate ranking criterion for feature selection. We have shown that classifiers generated from features selected by FrankSum outperforms classifiers generated from full feature sets, randomly selected features and features selected from the Wrapper method. We have also shown the features are concordant across all species and top ranking features are biologically informative. We conclude that feature selection is vital for successful protein function prediction and FrankSum is one of the feature selection methods that can be applied successfully to such a domain.

Intrinsically disordered proteins or proteins with disordered regions are very common in nature. These proteins have numerous biological functions which are complementary to the biological activities of traditional ordered proteins. A noticeable difference in the amino acid sequences encoding long and short disordered regions was found and this difference was used in the development of length-dependent predictors of intrinsic disorder. In this study, we analyze the scaling of intrinsic disorder in eukaryotic proteins and investigate the presence of length-dependent functions attributed to proteins containing long disordered regions.

THEMATICS (Theoretical Microscopic Titration Curves) is a simple, reliable computational predictor of the active sites of enzymes from structure. Our method, based on well-established Finite Difference Poisson–Boltzmann techniques, identifies the ionisable residues with anomalous predicted titration behavior. A cluster of two or more such perturbed residues is a very reliable predictor of the active site. The protein does not have to bear any resemblance in sequence or structure to any previously characterized protein, but the method does require the three-dimensional structure. We now present evidence that THEMATICS can also locate the active site in structures built by comparative modeling from similar structures. Results are given for a total of 21 sets of proteins, including 21 templates and 83 comparative model structures. Detailed results are presented for three sets of orthologous proteins (Triosephosphate isomerase, 6-Hydroxymethyl-7,8-dihydropterin pyrophosphokinase, and Aspartate aminotransferase) and for one set of human homologues of Aldose reductase with different functions. THEMATICS correctly locates the active site in the model structures. This suggests that the method can be applicable to a much larger set of proteins for which an experimentally determined structure is unavailable. With a few exceptions, the predicted active sites in the comparative model structures are similar to that of the corresponding template structure.

Protein–protein interactions play a defining role in protein function. Identifying the sites of interaction in a protein is a critical problem for understanding its functional mechanisms, as well as for drug design. To predict sites within a protein chain that participate in protein complexes, we have developed a novel method based on the Hidden Markov Model, which combines several biological characteristics of the sequences neighboring a target residue: structural information, accessible surface area, and transition probability among amino acids. We have evaluated the method using 5-fold cross-validation on 139 unique proteins and demonstrated precision of 66% and recall of 61% in identifying interfaces. These results are better than those achieved by other methods used for identification of interfaces.

Algorithms for geometric and chemical comparison of protein substructure can be useful for many applications in protein function prediction. These motif matching algorithms identify matches of geometric and chemical similarity between well-studied functional sites, motifs, and substructures of functionally uncharacterized proteins, targets. For the purpose of function prediction, the accuracy of motif matching algorithms can be evaluated with the number of statistically significant matches to functionally related proteins, true positives (TPs), and the number of statistically insignificant matches to functionally unrelated proteins, false positives (FPs).

Our earlier work developed cavity-aware motifs which use motif points to represent functionally significant atoms and C-spheres to represent functionally significant volumes. We observed that cavity-aware motifs match significantly fewer FPs than matches containing only motif points. We also observed that high-impact C-spheres, which significantly contribute to the reduction of FPs, can be isolated automatically with a technique we call Cavity Scaling.

This paper extends our earlier work by demonstrating that C-spheres can be used to accelerate point-based geometric and chemical comparison algorithms, maintaining accuracy while reducing runtime. We also demonstrate that the placement of C-spheres can significantly affect the number of TPs and FPs identified by a cavity-aware motif. While the optimal placement of C-spheres remains a diffcult open problem, we compared two logical placement strategies to better understand C-sphere placement.

Next-generation sequencing advances are rapidly expanding the number of human mutations to be analyzed for causative roles in genetic disorders. Our Human Protein Mutation Viewer (HPMV) is intended to explore the biomolecular mechanistic significance of non-synonymous human mutations in protein-coding genomic regions. The tool helps to assess whether protein mutations affect the occurrence of sequence-architectural features (globular domains, targeting signals, post-translational modification sites, etc.). As input, HPMV accepts protein mutations — as UniProt accessions with mutations (e.g. HGVS nomenclature), genome coordinates, or FASTA sequences. As output, HPMV provides an interactive cartoon showing the mutations in relation to elements of the sequence architecture. A large variety of protein sequence architectural features were selected for their particular relevance to mutation interpretation. Clicking a sequence feature in the cartoon expands a tree view of additional information including multiple sequence alignments of conserved domains and a simple 3D viewer mapping the mutation to known PDB structures, if available. The cartoon is also correlated with a multiple sequence alignment of similar sequences from other organisms. In cases where a mutation is likely to have a straightforward interpretation (e.g. a point mutation disrupting a well-understood targeting signal), this interpretation is suggested. The interactive cartoon can be downloaded as standalone viewer in Java jar format to be saved and viewed later with only a standard Java runtime environment. The HPMV website is: http://hpmv.bii.a-star.edu.sg/.

The concept that an unfolded protein could have biological function was proposed by Linus Pauling in 1940. The first experimental evidence for the existence of unfolded proteins, which are now often called intrinsically disordered proteins and regions (IDPs and IDRs), was published in the 1950s. From the 1950s to the 1990s, many additional articles describing IDPs or IDRs were published, and even several Nobel Prizes have been awarded for research on these proteins. The first computational biology/bioinformatics papers on these proteins were reported between 1998 and 2002, and these computational efforts rapidly increased shortly thereafter. In the late 1980s and mid-1990s, a few IDPs and IDRs involved with signaling and regulation were structurally characterized by nuclear magnet resonance (NMR), and these studies demonstrated a likely involvement of IDPs or IDRs in the associated biological regulation. Altogether, the computational biology/bioinformatics and NMR investigations were synergistic in stimulating the rapid increase in the research on IDPs and IDRs. Herein, we describe this history and recent developments in the IDP research arena. We also highlight a recent collection of papers on IDPs and IDRs that are driven by computational biology and bioinformatics efforts.

Intrinsically disorder regions (IDRs) lack a stable structure, yet perform biological functions. The functions of IDRs include mediating interactions with other molecules, including proteins, DNA, or RNA and entropic functions, including domain linkers. Computational predictors provide residue-level indications of function for disordered proteins, which contrasts with the need to functionally annotate the thousands of experimentally and computationally discovered IDRs. In this work, we investigate the feasibility of using residue-level prediction methods for region-level function predictions. For an initial examination of the multiple function region-level prediction problem, we constructed a dataset of (likely) single function IDRs in proteins that are dissimilar to the training datasets of the residue-level function predictors. We find that available residue-level prediction methods are only modestly useful in predicting multiple region-level functions. Classification is enhanced by simultaneous use of multiple residue-level function predictions and is further improved by inclusion of amino acids content extracted from the protein sequence. We conclude that multifunction prediction for IDRs is feasible and benefits from the results produced by current residue-level function predictors, however, it has to accommodate inaccuracy in functional annotations.

We propose a new kernel-based method for the classification of protein sequences and structures. We first represent each protein as a set of time series data using several structural, physicochemical, and predicted properties such as a sequence of consecutive dihedral angles, hydrophobicity indices, or predictions of disordered regions. A kernel function is then computed for pairs of proteins, exploiting the principles of vector quantization and subsequently used with support vector machines for protein classification. Although our method requires a significant pre-processing step, it is fast in the training and prediction stages owing to the linear complexity of kernel computation with the length of protein sequences. We evaluate our approach on two protein classification tasks involving the prediction of SCOP structural classes and catalytic activity according to the Gene Ontology. We provide evidence that the method is competitive when compared to string kernels, and useful for a range of protein classification tasks. Furthermore, the applicability of our approach extends beyond computational biology to any classification of time series data.