Reoccurring Patterns in Hierarchical Protein Materials and Music: The Power of Analogies

We have developed an automatic algorithm STRIDE for protein secondary structure assignment from atomic coordinates based on the combined use of hydrogen bond energy and statistically derived backbone torsional angle information. Parameters of the pattern recognition procedure were optimized using designations provided by the crystallographers as a standard-of-truth. Comparison to the currently most widely used technique DSSP by Kabsch and Sander (Biopolymers 22:2577-2637, 1983) shows that STRIDE and DSSP assign secondary structural states in 58 and 31% of 226 protein chains in our data sample, respectively, in greater agreement with the specific residue-by-residue definitions provided by the discoverers of the structures while in 11% of the chains, the assignments are the same. STRIDE delineates every 11th helix and every 32nd strand more in accord with published assignments.

Spider dragline silk is one of the strongest, most extensible and toughest biological materials known, exceeding the properties of many engineered materials including steel. Silk features a hierarchical architecture where highly organized, densely H-bonded beta-sheet nanocrystals are arranged within a semiamorphous protein matrix consisting of 3(1)-helices and beta-turn protein structures. By using a bottom-up molecular-based approach, here we develop the first spider silk mesoscale model, bridging the scales from Angstroms to tens to potentially hundreds of nanometers. We demonstrate that the specific nanoscale combination of a crystalline phase and a semiamorphous matrix is crucial to achieve the unique properties of silks. Our results reveal that the superior mechanical properties of spider silk can be explained solely by structural effects, where the geometric confinement of beta-sheet nanocrystals, combined with highly extensible semiamorphous domains, is the key to reach great strength and great toughness, despite the dominance of mechanically inferior chemical interactions such as H-bonding. Our model directly shows that semiamorphous regions govern the silk behavior at small deformation, unraveling first when silk is being stretched and leading to the large extensibility of the material. Conversely, beta-sheet nanocrystals play a significant role in defining the mechanical behavior of silk at large-deformation. In particular, the ultimate tensile strength of silk is controlled by the strength of beta-sheet nanocrystals, which is directly related to their size, where small beta-sheet nanocrystals are crucial to reach outstanding levels of strength and toughness. Our results and mechanistic insight directly explain recent experimental results, where it was shown that a significant change in the strength and toughness of silk can be achieved solely by tuning the size of beta-sheet nanocrystals. Our findings help to unveil the material design strategy that enables silk to achieve superior material performance despite simple and inferior material constituents. This concept could lead to a new materials design paradigm, where enhanced functionality is not achieved using complex building blocks but rather through the utilization of simple repetitive constitutive elements arranged in hierarchical structures from nano to macro.

We have introduced a new method of protein secondary structure prediction which is based on the theory of support vector machine (SVM). SVM represents a new approach to supervised pattern classification which has been successfully applied to a wide range of pattern recognition problems, including object recognition, speaker identification, gene function prediction with microarray expression profile, etc. In these cases, the performance of SVM either matches or is significantly better than that of traditional machine learning approaches, including neural networks.The first use of the SVM approach to predict protein secondary structure is described here. Unlike the previous studies, we first constructed several binary classifiers, then assembled a tertiary classifier for three secondary structure states (helix, sheet and coil) based on these binary classifiers. The SVM method achieved a good performance of segment overlap accuracy SOV=76.2 % through sevenfold cross validation on a database of 513 non-homologous protein chains with multiple sequence alignments, which out-performs existing methods. Meanwhile three-state overall per-residue accuracy Q(3) achieved 73.5 %, which is at least comparable to existing single prediction methods. Furthermore a useful "reliability index" for the predictions was developed. In addition, SVM has many attractive features, including effective avoidance of overfitting, the ability to handle large feature spaces, information condensing of the given data set, etc. The SVM method is conveniently applied to many other pattern classification tasks in biology. Copyright 2001 Academic Press.