The tendency of an amino acid to adopt certain configurations in folded proteins is treated here as a statistical estimation problem. We model the joint distribution of the observed mainchain and sidechain dihedral angles (〈ϕ,ψ,χ1,χ2,…〉) of any amino acid by a mixture of a product of von Mises probability distributions. This mixture model maps any vector of dihedral angles to a point on a multi-dimensional torus. The continuous space it uses to specify the dihedral angles provides an alternative to the commonly used rotamer libraries. These rotamer libraries discretize the space of dihedral angles into coarse angular bins, and cluster combinations of sidechain dihedral angles (〈χ1,χ2,…〉) as a function of backbone 〈ϕ,ψ〉 conformations. A 'good' model is one that is both concise and explains (compresses) observed data. Competing models can be compared directly and in particular our model is shown to outperform the Dunbrack rotamer library in terms of model complexity (by three orders of magnitude) and its fidelity (on average 20% more compression) when losslessly explaining the observed dihedral angle data across experimental resolutions of structures. Our method is unsupervised (with parameters estimated automatically) and uses information theory to determine the optimal complexity of the statistical model, thus avoiding under/over-fitting, a common pitfall in model selection problems. Our models are computationally inexpensive to sample from and are geared to support a number of downstream studies, ranging from experimental structure refinement, de novo protein design, and protein structure prediction. We call our collection of mixture models as PhiSiCal (ϕψχal). AVAILABILITY AND IMPLEMENTATION: PhiSiCal mixture models and programs to sample from them are available for download at http://lcb.infotech.monash.edu.au/phisical.

MOTIVATION: Alignments are correspondences between sequences. How reliable are alignments of amino acid sequences of proteins, and what inferences about protein relationships can be drawn? Using techniques not previously applied to these questions, by weighting every possible sequence alignment by its posterior probability we derive a formal mathematical expectation, and develop an efficient algorithm for computation of the distance between alternative alignments allowing quantitative comparisons of sequence-based alignments with corresponding reference structure alignments. RESULTS: By analyzing the sequences and structures of 1 million protein domain pairs, we report the variation of the expected distance between sequence-based and structure-based alignments, as a function of (Markov time of) sequence divergence. Our results clearly demarcate the 'daylight', 'twilight' and 'midnight' zones for interpreting residue-residue correspondences from sequence information alone. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

SUMMARY: Sequences of proteins evolve by accumulating substitutions together with insertions and deletions (indels) of amino acids. However, it remains a common practice to disconnect substitutions and indels, and infer approximate models for each of them separately, to quantify sequence relationships. Although this approach brings with it computational convenience (which remains its primary motivation), there is a dearth of attempts to unify and model them systematically and together. To overcome this gap, this article demonstrates how a complete statistical model quantifying the evolution of pairs of aligned proteins can be constructed using a time-parameterized substitution matrix and a time-parameterized alignment state machine. Methods to derive all parameters of such a model from any benchmark collection of aligned protein sequences are described here. This has not only allowed us to generate a unified statistical model for each of the nine widely used substitution matrices (PAM, JTT, BLOSUM, JO, WAG, VTML, LG, MIQS and PFASUM), but also resulted in a new unified model, MMLSUM. Our underlying methodology measures the Shannon information content using each model to explain losslessly any given collection of alignments, which has allowed us to quantify the performance of all the above models on six comprehensive alignment benchmarks. Our results show that MMLSUM results in a new and clear overall best performance, followed by PFASUM, VTML, BLOSUM and MIQS, respectively, amongst the top five. We further analyze the statistical properties of MMLSUM model and contrast it with others. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

This paper presents a framework to characterize and identify local sequences of proteins that are statistically redundant under the measure of Shannon information content while accounting for variations in their occurrences over evolutionary insertions, deletions, and substitutions of amino acids. The identification of such local sequences provides insights for downstream studies on proteins. Here, we have applied our methods to amino acid sequence data sets derived from a database corresponding to 935,552 substructural regions of varying sizes, covering 113,724 proteins from the protein data bank. The results identify, among others, a surjective mapping between 110,598 local sequences (with an average length of 82 amino acids per sequence) and 1,493 topological shapes. The C++ source code and supporting material are available from https://lcb.infotech.monash.edu.au/bibm2021.

A universal code for the (positive) integers is a variable length code that can be used to store or compress a sequence of integers. It also implies a probability distribution on integers which can be a natural choice when the true distribution of a source of integers is unknown; such a code and distribution may be useful in statistical inference. This paper provides two improvements to the theory and practice of universal codes. First, it defines and examines a new universal code omega* (omega-star) that asymptotically beats the Elias omega code. Second, it analyses the properties of a code proposed by Wallace based on trees, and shows it to be a universal code, to have desirable properties for use in inference, and to beat the Elias omega code on almost all integers up to the 1697-bit code-word mark. Encoding and decoding routines for the codes described here are implemented and available for interactive use.11The codes may be tried at www.allisons.org/ll/MML/Discrete/Universal/ ← click.

MOTIVATION: The gene expression regulatory network in yeast controls the selective implementation of the information contained in the genome sequence. We seek to understand how, in different physiological states, the network reconfigures itself to produce a different proteome. RESULTS: This article analyses this reconfiguration, focussing on changes in the local structure of the network. In particular, we define, extract and compare the 1-neighbourhoods of each transcription factor, where a 1-neighbourhood of a node in a network is the minimal subgraph of the network containing all nodes connected to the central node by an edge. We report the similarities and differences in the topologies and connectivities of these neighbourhoods in five physiological states for which data are available: cell cycle, DNA damage, stress response, diauxic shift and sporulation. Based on our analysis, it seems apt to regard the components of the regulatory network as 'software', and the responses to changes in state, 'reprogramming'.

The information criterion of minimum message length (MML) provides a powerful statistical framework for inductive reasoning from observed data. We apply MML to the problem of protein sequence comparison using finite state models with Dirichlet distributions. The resulting framework allows us to supersede the ad hoc cost functions commonly used in the field, by systematically addressing the problem of arbitrariness in alignment parameters, and the disconnect between substitution scores and gap costs. Furthermore, our framework enables the generation of marginal probability landscapes over all possible alignment hypotheses, with potential to facilitate the users to simultaneously rationalize and assess competing alignment relationships between protein sequences, beyond simply reporting a single (best) alignment. We demonstrate the performance of our program on benchmarks containing distantly related protein sequences. AVAILABILITY AND IMPLEMENTATION: The open-source program supporting this work is available from: http://lcb.infotech.monash.edu.au/seqmmligner. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Comparison of protein sequences via alignment is an important routine in modern biological studies. Although the technologies for aligning proteins are mature, the current state of the art continues to be plagued by many shortcomings, chiefly due to the reliance on: (i) naive objective functions, (ii) fixed substitution scores independent of the sequences being considered, (iii) arbitrary choices for gap costs, and (iv) reporting, often, one optimal alignment without a way to recognise other competing sequence alignments. Here, we address these shortcomings by applying the compression-based Minimum Message Length (MML) inference framework to the protein sequence alignment problem. This grounds the problem in statistical learning theory, handles directly the complexity-vs-fit trade-off without ad hoc gap costs, allows unsupervised inference of all the statistical parameters, and permits the visualization and exploration of competing sequence alignment landscape.

Computational analyses of the growing corpus of three-dimensional (3D) structures of proteins have revealed a limited set of recurrent substructural themes, termed super-secondary structures. Knowledge of super-secondary structures is important for the study of protein evolution and for the modeling of proteins with unknown structures. Characterizing a comprehensive dictionary of these super-secondary structures has been an unanswered computational challenge in protein structural studies. This paper presents an unsupervised method for learning such a comprehensive dictionary using the statistical framework of lossless compression on a database comprised of concise geometric representations of protein 3D folding patterns. The best dictionary is defined as the one that yields the most compression of the database. Here we describe the inference methodology and the statistical models used to estimate the encoding lengths. An interactive website for this dictionary is available at http://lcb.infotech.monash.edu.au/proteinConcepts/scop100/dictionary.html.

Motivation: Structural molecular biology depends crucially on computational techniques that compare protein three-dimensional structures and generate structural alignments (the assignment of one-to-one correspondences between subsets of amino acids based on atomic coordinates). Despite its importance, the structural alignment problem has not been formulated, much less solved, in a consistent and reliable way. To overcome these difficulties, we present here a statistical framework for the precise inference of structural alignments, built on the Bayesian and information-theoretic principle of Minimum Message Length (MML). The quality of any alignment is measured by its explanatory power-the amount of lossless compression achieved to explain the protein coordinates using that alignment. Results: We have implemented this approach in MMLigner , the first program able to infer statistically significant structural alignments. We also demonstrate the reliability of MMLigner 's alignment results when compared with the state of the art. Importantly, MMLigner can also discover different structural alignments of comparable quality, a challenging problem for oligomers and protein complexes. Availability and Implementation: Source code, binaries and an interactive web version are available at http://lcb.infotech.monash.edu.au/mmligner . Contact: [email protected]. Supplementary information: Supplementary data are available at Bioinformatics online.

Comparing the top k elements between two or more ranked results is a common task in many contexts and settings. A few measures have been proposed to compare top k lists with attractive mathematical properties, but they face a number of pitfalls and shortcomings in practice. This work introduces a new measure to compare any two top k lists based on measuring the information these lists convey. Our method investigates the compressibility of the lists, and the length of the message to encode losslessly the lists gives a natural and robust measure of their variability. This information-theoretic measure objectively reconciles all the main considerations that arise when measuring (dis-)similarity between lists: the extent of their non-overlapping elements, the amount of disarray among overlapping elements, the measurement of displacement of actual ranks (positions) of their overlapping elements. We demonstrate that our measure is intuitively simple and superior to other commonly used measures. To the best of our knowledge, this is the first attempt to address the problem using information compression as its basis.

MOTIVATION: Progress in protein biology depends on the reliability of results from a handful of computational techniques, structural alignments being one. Recent reviews have highlighted substantial inconsistencies and differences between alignment results generated by the ever-growing stock of structural alignment programs. The lack of consensus on how the quality of structural alignments must be assessed has been identified as the main cause for the observed differences. Current methods assess structural alignment quality by constructing a scoring function that attempts to balance conflicting criteria, mainly alignment coverage and fidelity of structures under superposition. This traditional approach to measuring alignment quality, the subject of considerable literature, has failed to solve the problem. Further development along the same lines is unlikely to rectify the current deficiencies in the field. RESULTS: This paper proposes a new statistical framework to assess structural alignment quality and significance based on lossless information compression. This is a radical departure from the traditional approach of formulating scoring functions. It links the structural alignment problem to the general class of statistical inductive inference problems, solved using the information-theoretic criterion of minimum message length. Based on this, we developed an efficient and reliable measure of structural alignment quality, I-value. The performance of I-value is demonstrated in comparison with a number of popular scoring functions, on a large collection of competing alignments. Our analysis shows that I-value provides a rigorous and reliable quantification of structural alignment quality, addressing a major gap in the field. AVAILABILITY: http://lcb.infotech.monash.edu.au/I-value. SUPPLEMENTARY INFORMATION: Online supplementary data are available at http://lcb.infotech.monash.edu.au/I-value/suppl.html.

Proteins fold into complex three-dimensional shapes. Simplified representations of their shapes are central to rationalise, compare, classify, and interpret protein structures. Traditional methods to abstract protein folding patterns rely on representing their standard secondary structural elements (helices and strands of sheet) using line segments. This results in ignoring a significant proportion of structural information. The motivation of this research is to derive mathematically rigorous and biologically meaningful abstractions of protein folding patterns that maximize the economy of structural description and minimize the loss of structural information. We report on a novel method to describe a protein as a non-overlapping set of parametric three dimensional curves of varying length and complexity. Our approach to this problem is supported by information theory and uses the statistical framework of minimum message length (MML) inference. We demonstrate the effectiveness of our non-linear abstraction to support efficient and effective comparison of protein folding patterns on a large scale.

Proteins are biomolecules of life. They fold into a great variety of three-dimensional (3D) shapes. Underlying these folding patterns are many recurrent structural fragments or building blocks (analogous to 'LEGO® bricks'). This paper reports an innovative statistical inference approach to discover a comprehensive dictionary of protein structural building blocks from a large corpus of experimentally determined protein structures. Our approach is built on the Bayesian and information theoretic criterion of minimum message length. To the best of our knowledge, this work is the first systematic and rigorous treatment of a very important data mining problem that arises in the cross-disciplinary area of structural bioinformatics. The quality of the dictionary we find is demonstrated by its explanatory power - any protein within the corpus of known 3D structures can be dissected into successive regions assigned to fragments from this dictionary. This induces a novel one-dimensional representation of three-dimensional protein folding patterns, suitable for application of the rich repertoire of character-string processing algorithms, for rapid identification of folding patterns of newly determined structures. This paper presents the details of the methodology used to infer the dictionary of building blocks, and is supported by illustrative examples to demonstrate its effectiveness and utility.

MOTIVATION: Secondary structure underpins the folding pattern and architecture of most proteins. Accurate assignment of the secondary structure elements is therefore an important problem. Although many approximate solutions of the secondary structure assignment problem exist, the statement of the problem has resisted a consistent and mathematically rigorous definition. A variety of comparative studies have highlighted major disagreements in the way the available methods define and assign secondary structure to coordinate data. RESULTS: We report a new method to infer secondary structure based on the Bayesian method of minimum message length inference. It treats assignments of secondary structure as hypotheses that explain the given coordinate data. The method seeks to maximize the joint probability of a hypothesis and the data. There is a natural null hypothesis and any assignment that cannot better it is unacceptable. We developed a program SST based on this approach and compared it with popular programs, such as DSSP and STRIDE among others. Our evaluation suggests that SST gives reliable assignments even on low-resolution structures. AVAILABILITY: http://www.csse.monash.edu.au/~karun/sst.

UNLABELLED: Simple and concise representations of protein-folding patterns provide powerful abstractions for visualizations, comparisons, classifications, searching and aligning structural data. Structures are often abstracted by replacing standard secondary structural features-that is, helices and strands of sheet-by vectors or linear segments. Relying solely on standard secondary structure may result in a significant loss of structural information. Further, traditional methods of simplification crucially depend on the consistency and accuracy of external methods to assign secondary structures to protein coordinate data. Although many methods exist automatically to identify secondary structure, the impreciseness of definitions, along with errors and inconsistencies in experimental structure data, drastically limit their applicability to generate reliable simplified representations, especially for structural comparison. This article introduces a mathematically rigorous algorithm to delineate protein structure using the elegant statistical and inductive inference framework of minimum message length (MML). Our method generates consistent and statistically robust piecewise linear explanations of protein coordinate data, resulting in a powerful and concise representation of the structure. The delineation is completely independent of the approaches of using hydrogen-bonding patterns or inspecting local substructural geometry that the current methods use. Indeed, as is common with applications of the MML criterion, this method is free of parameters and thresholds, in striking contrast to the existing programs which are often beset by them. The analysis of results over a large number of proteins suggests that the method produces consistent delineation of structures that encompasses, among others, the segments corresponding to standard secondary structure. AVAILABILITY: http://www.csse.monash.edu.au/~karun/pmml.

MOTIVATION: Identification of functionally equivalent genes in different species is essential to understand the evolution of biological pathways and processes. At the same time, identification of strings of conserved orthologous genes helps identify complex genomic rearrangements across different organisms. Such an insight is particularly useful, for example, in the transfer of experimental results between different experimental systems such as Drosophila and mammals. RESULTS: Here, we describe the Encapsulated Gene-by-gene Matching (EGM) approach, a method that employs a graph matching strategy to identify gene orthologs and conserved gene segments. Given a pair of genomes, EGM constructs a global gene match for all genes taking into account gene context and family information. The Hungarian method for identifying the maximum weight matching in bipartite graphs is employed, where the resulting matching reveals one-to-one correspondences between nodes (genes) in a manner that maximizes the gene similarity and context. CONCLUSION: We tested our approach by performing several comparisons including a detailed Human versus Mouse genome mapping. We find that the algorithm is robust and sensitive in detecting orthologs and conserved gene segments. EGM can sensitively detect rearrangements within large and small chromosomal segments. The EGM tool is fully automated and easy to use compared to other more complex methods that also require extensive manual intervention and input. AVAILABILITY: The EGM software, Supplementary information and other tools are available online from http://vbc.med.monash.edu.au/ approximately kmahmood/EGM.

BACKGROUND: Protein structure comparison is a fundamental task in structural biology. While the number of known protein structures has grown rapidly over the last decade, searching a large database of protein structures is still relatively slow using existing methods. There is a need for new techniques which can rapidly compare protein structures, whilst maintaining high matching accuracy. RESULTS: We have developed IR Tableau, a fast protein comparison algorithm, which leverages the tableau representation to compare protein tertiary structures. IR tableau compares tableaux using information retrieval style feature indexing techniques. Experimental analysis on the ASTRAL SCOP protein structural domain database demonstrates that IR Tableau achieves two orders of magnitude speedup over the search times of existing methods, while producing search results of comparable accuracy. CONCLUSION: We show that it is possible to obtain very significant speedups for the protein structure comparison problem, by employing an information retrieval style approach for indexing proteins. The comparison accuracy achieved is also strong, thus opening the way for large scale processing of very large protein structure databases.

UNLABELLED: Comparison and classification of folding patterns from a database of protein structures is crucial to understand the principles of protein architecture, evolution and function. Current search methods for proteins with similar folding patterns are slow and computationally intensive. The sharp growth in the number of known protein structures poses severe challenges for methods of structural comparison. There is a need for methods that can search the database of structures accurately and rapidly. We provide several methods to search for similar folding patterns using a concise tableau representation of proteins that encodes the relative geometry of secondary structural elements. Our first approach allows the extraction of identical and very closely-related protein folding patterns in constant-time (per hit). Next, we address the hard computational problem of extraction of maximally-similar subtableaux, when comparing two tableaux. We solve the problem using Quadratic and Linear integer programming formulations and demonstrate their power to identify subtle structural similarities, especially when protein structures significantly diverge. Finally, we describe a rapid and accurate method for comparing a query structure against a database of protein domains, TableauSearch. TableauSearch is rapid enough to search the entire structural database in seconds on a standard desktop computer. Our analysis of TableauSearch on many queries shows that the method is very accurate in identifying similarities of folding patterns, even between distantly related proteins. AVAILABILITY: A web server implementing the TableauSearch is available from http://hollywood.bx.psu.edu/TabSearch.