A survey of machine learning methods for secondary and supersecondary protein structure prediction.

In this chapter we provide a survey of protein secondary and supersecondary structure prediction using methods from machine learning. Our focus is on machine learning methods applicable to ß-hairpin and ß-sheet prediction, but we also discuss methods for more general supersecondary structure prediction. We provide background on the secondary and supersecondary structures that we discuss, the features used to describe them, and the basic theory behind the machine learning methods used. We survey the machine learning methods available for secondary and supersecondary structure prediction and compare them where possible.

BetaSearch: a new method for querying ß-residue motifs.

BACKGROUND: Searching for structural motifs across known protein structures can be useful for identifying unrelated proteins with similar function and characterising secondary structures such as ß-sheets. This is infeasible using conventional sequence alignment because linear protein sequences do not contain spatial information. ß-residue motifs are ß-sheet substructures that can be represented as graphs and queried using existing graph indexing methods, however, these approaches are designed for general graphs that do not incorporate the inherent structural constraints of ß-sheets and require computationally-expensive filtering and verification procedures. 3D substructure search methods, on the other hand, allow ß-residue motifs to be queried in a three-dimensional context but at significant computational costs. FINDINGS: We developed a new method for querying ß-residue motifs, called BetaSearch, which leverages the natural planar constraints of ß-sheets by indexing them as 2D matrices, thus avoiding much of the computational complexities involved with structural and graph querying. BetaSearch exhibits faster filtering, verification, and overall query time than existing graph indexing approaches whilst producing comparable index sizes. Compared to 3D substructure search methods, BetaSearch achieves 33 and 240 times speedups over index-based and pairwise alignment-based approaches, respectively. Furthermore, we have presented case-studies to demonstrate its capability of motif matching in sequentially dissimilar proteins and described a method for using BetaSearch to predict ß-strand pairing. CONCLUSIONS: We have demonstrated that BetaSearch is a fast method for querying substructure motifs. The improvements in speed over existing approaches make it useful for efficiently performing high-volume exploratory querying of possible protein substructural motifs or conformations. BetaSearch was used to identify a nearly identical ß-residue motif between an entirely synthetic (Top7) and a naturally-occurring protein (Charcot-Leyden crystal protein), as well as identifying structural similarities between biotin-binding domains of avidin, streptavidin and the lipocalin gamma subunit of human C8.

PConPy--a Python module for generating 2D protein maps.

UNLABELLED: PConPy is an open-source Python module for generating protein contact maps, distance maps and hydrogen bond plots. These maps can be generated in a number of publication-quality vector and raster image formats. Contact maps can be annotated with secondary structure and hydrogen bond assignments. PConPy offers a more flexible choice of contact definition parameters than existing toolkits, most notably a greater choice of inter-residue distance metrics. PConPy can be used as a stand-alone application or imported into existing source code. A web-interface to PConPy is also available for use. AVAILABILITY: The PConPy web-interface and source code can be accessed from its website at http://www.csse.unimelb.edu.au/~hohkhkh1/pconpy/. CONTACT: hohkhkh1@csse.unimelb.edu.au