Unsupervised Feature Representation of Sleep EEG Data with Transient Deep Boltzmann Machine.

The supervised sleep staging methods are challenged by their strict requirements of a labelled and large dataset. This study considers an unsupervised dimensionality reduction method, the Deep Boltzmann Machine (DBM), trained to a transient state for binary classification of sleep stages. First, the joint time-frequency domain features from the polysomnographic recordings are extracted. Second, the extracted features are smoothed using 2 min rolling window to include contextual temporal information, and finally, they serve as an input for unsupervised training of DBM_transient. The results show that our method effectively separates the sleep stages in two-dimensional feature space with a large Fisher's discriminant value. The classification performance by the DBM_transient achieves a 96.1% F1 score, which is higher than DBM converged to an equilibrium state (95.2%), Principal Component Analysis (92.5%), Isometric Feature Mapping (95.9%), t-distributed Stochastic Neighbor Embedding (94.9%), and Uniform Manifold Approximation (95.0%) on the widely used sleep-EDF database. Additionally, Fisher's discriminant function demonstrates the superiority of the DBM_transient. The significance of the DBM transient lies in its ease of interpretability in two-dimensional space, and future multi-class implementation of the method may facilitate its usage in clinical applications.

Basis for Accurate Protein pKa Prediction with Machine Learning.

pH regulates protein structures and the associated functions in many biological processes via protonation and deprotonation of ionizable side chains where the titration equilibria are determined by pKa's. To accelerate pH-dependent molecular mechanism research in the life sciences or industrial protein and drug designs, fast and accurate pKa prediction is crucial. Here we present a theoretical pKa data set PHMD549, which was successfully applied to four distinct machine learning methods, including DeepKa, which was proposed in our previous work. To reach a valid comparison, EXP67S was selected as the test set. Encouragingly, DeepKa was improved significantly and outperforms other state-of-the-art methods, except for the constant-pH molecular dynamics, which was utilized to create PHMD549. More importantly, DeepKa reproduced experimental pKa orders of acidic dyads in five enzyme catalytic sites. Apart from structural proteins, DeepKa was found applicable to intrinsically disordered peptides. Further, in combination with solvent exposures, it is revealed that DeepKa offers the most accurate prediction under the challenging circumstance that hydrogen bonding or salt bridge interaction is partly compensated by desolvation for a buried side chain. Finally, our benchmark data qualify PHMD549 and EXP67S as the basis for future developments of protein pKa prediction tools driven by artificial intelligence. In addition, DeepKa built on PHMD549 has been proven an efficient protein pKa predictor and thus can be applied immediately to, for example, pKa database construction, protein design, drug discovery, and so on.

Unsupervised Feature Representation Based on Deep Boltzmann Machine for Seizure Detection.

The Electroencephalogram (EEG) pattern of seizure activities is highly individual-dependent and requires experienced specialists to annotate seizure events. It is clinically time-consuming and error-prone to identify seizure activities by visually scanning EEG signals. Since EEG data are heavily under-represented, supervised learning techniques are not always practical, particularly when the data is not sufficiently labelled. Visualization of EEG data in low-dimensional feature space can ease the annotation to support subsequent supervised learning for seizure detection. Here, we leverage the benefit of both the time-frequency domain features and the Deep Boltzmann Machine (DBM) based unsupervised learning techniques to represent EEG signals in a 2-dimensional (2D) feature space. A novel unsupervised learning approach based on DBM, namely DBM_transient, is proposed by training DBM to a transient state for representing EEG signals in a 2D feature space and clustering seizure and non-seizure events visually. The effectiveness of DBM_transient is demonstrated on a widely-used benchmark dataset from Bonn University (Bonn dataset) and a raw clinical dataset from Chinese 301 Hospital (C301 dataset), with a large fisher discriminant value, surpassing the abilities of other dimensionality reduction methods, including DBM converged to an equilibrium state, Kernel Principal Component Analysis, Isometric Feature Mapping, t-distributed Stochastic Neighbour Embedding, Uniform Manifold Approximation. Such feature representation and visualization can help physicians to understand better the normal versus epileptic brain activities of each patient and thus enhance their diagnosis and treatment abilities. The significance of our approach facilitates its future usage in clinical applications.