Prediction of Side Chain Orientations in Proteins by Statistical Machine Learning Methods

Abstract

We develop ways to predict the side chain orientations of residues within a protein structure by using several different statistical machine learning methods. Here side chain orientation of a given residue i is measured by an angle Ω_i between the vector pointing from the center of the protein structure to the C^α _i atom and the vector pointing from the C^α _i atom to the center of its side chain atoms. To predict the Ω_i angles, we construct statistical models by using several different methods such as general linear regression, a regression tree and bagging, a neural network, and a support vector machine. The root mean square errors for the different models range only from 36.67 to 37.60 degrees and the correlation coefficients are all between 30% and 34%. The performances of different models in the test set are, thus, quite similar, and show the relative predictive power of these models to be significant in comparison with random side chain orientations.     

Inter-individual variability and pattern recognition of surface electromyography in front crawl swimming

Variability of electromyographic (EMG) recordings is a complex phenomenon rarely examined in swimming. Our purposes were to investigate inter-individual variability in muscle activation patterns during front crawl swimming and assess if there were clusters of sub patterns present. Bilateral muscle activity of rectus abdominis (RA) and deltoideus medialis (DM) was recorded using wireless surface EMG in 15 adult male competitive swimmers. The amplitude of the median EMG trial of six upper arm movement cycles was used for the inter-individual variability assessment, quantified with the coefficient of variation, coefficient of quartile variation, the variance ratio and mean deviation. Key features were selected based on qualitative and quantitative classification strategies to enter in a k-means cluster analysis to examine the presence of strong sub patterns. Such strong sub patterns were found when clustering in two, three and four clusters. Inter-individual variability in a group of highly skilled swimmers was higher compared to other cyclic movements which is in contrast to what has been reported in the previous 50 years of EMG research in swimming. This leads to the conclusion that coaches should be careful in using overall reference EMG information to enhance the individual swimming technique of their athletes.     

The concept of channel permeability

The one-way fluxes of a certain species of ion through a passive channel in a fixed state are assumed to be well-behaved functions of seven variables only, all of which are macroscopically observable: the cis and trans concentrations of that ion, the charge of the ion, the voltage difference between the cis and trans solutions, the thermodynamic temperature, the ion concentration at half maximal conductance of the channel, and the channel permeability. It is then proved that, if these fluxes are independent and Nernstian, the absolute permeability is uniquely and explicitly defined in terms of the channel current near zero voltage; no assumptions about the detailed workings of the permeation process are necessary. It is also proved that, if the one-way fluxes of the species under consideration are merely unaffected by the other species present, then the channel's absolute permeability can still be uniquely and explicitly defined if a weak constraint on the flux ratio is assumed; again, no assumptions on the inner workings of the channel are necessary.