An Overview of Several Inhibitors for Alzheimer’s Disease: Characterization and Failure

Amyloid beta (Aβ) oligomers are the most neurotoxic aggregates causing neuronal death and cognitive damage. A detailed elucidation of the aggregation pathways from oligomers to fibril formation is crucial to develop therapeutic strategies for Alzheimer’s disease (AD). Although experimental techniques rely on the measure of time- and space-average properties, they face severe difficulties in the investigation of Aβ peptide aggregation due to their intrinsically disorder character. Computer simulation is a tool that allows tracing the molecular motion of molecules; hence it complements Aβ experiments, as it allows to explore the binding mechanism between metal ions and Aβ oligomers close to the cellular membrane at the atomic resolution. In this context, integrated studies of experiments and computer simulations can assist in mapping the complete pathways of aggregation and toxicity of Aβ peptides. Aβ oligomers are disordered proteins, and due to a rapid exploration of their intrinsic conformational space in real-time, they are challenging therapeutic targets. Therefore, no good drug candidate could have been identified for clinical use. Our previous investigations identified two small molecules, M30 (2-Octahydroisoquinolin-2(1H)-ylethanamine) and Gabapentin, capable of Aβ binding and inhibiting molecular aggregation, synaptotoxicity, intracellular calcium signaling, cellular toxicity and memory losses induced by Aβ. Thus, we recommend these molecules as novel candidates to assist anti-AD drug discovery in the near future. This review discusses the most recent research investigations about the Aβ dynamics in water, close contact with cell membranes, and several therapeutic strategies to remove plaque formation.

A High Accurate Machine Learning Meta-Strategy for the Prediction of Intrinsically Disorder Proteins

Background: Many proteins or partial regions of proteins do not have stable and well-defined three-dimensional structures in vitro. Understanding Intrinsically Disorder Proteins (IDPs) is significant for interpreting biological function as well as studying many diseases. Although more than 70 disorder predictors have been invented, many existing predictors are limited on the characteristics of proteins and do not have very high accuracy. Therefore, it is critical to formulate new strategies on disorder protein prediction. Results: Here, we propose a machine learning meta-strategy to improve the accuracy of disordered proteins and disordered regions prediction. We first use logistic forward parameter selection to select eight most significant predictors from the current available IDP predictors. Then we design a novel meta-strategy using several machine learning models, including Decision-tree based algorithm, Naive Bayes, Random forest, and Convolutional Neural Network (CNN). By applying different strategies, the results suggest Random forest can improve the predicted single amino acid accuracy significantly to 93.35%. Using the combination vector data of eight most significant predictors as input, the Convolution Neural Network can improve the whole protein prediction to 95.62%. Conclusion: According to the performance of our machine learning meta-strategy, the Random forest and CNN models can improve the accuracy to predict IDPs.

A High Accurate Machine Learning Meta-Strategy for the Prediction of Intrinsically Disorder Proteins

Background: Many proteins or partial regions of proteins do not have stable and well-defined three-dimensional structures in vitro. Understanding intrinsically disorder proteins (IDPs) is significant for interpreting biological function as well as studying many diseases. Although more than 70 disorder predictors have been invented, many existing predictors are limited on the characteristics of proteins and do not have very high accuracy. Therefore, it is critical to formulate new strategies on disorder protein prediction. Results: Here, we propose a machine learning meta-strategy to improve the accuracy of disordered proteins and disordered regions prediction. We first use logistic forward parameter selection to select eight most significant predictors from the current available IDP predictors. Then we design a novel meta-strategy using several machine learning models, including Decision-tree based algorithm, Naive Bayes, Random forest, and Convolutional Neural Network (CNN). By applying different strategies, the results suggest Random forest can improve the predicted single amino acid accuracy significantly to 93.35%. Using the combination vector data of eight most significant predictors as input, the Convolution Neural Network can improve the whole protein prediction to 95.62%. Conclusion: According to the performance of our machine learning meta-strategy, the Random forest and CNN models can improve the accuracy to predict intrinsically disorder proteins.

Computational Disorder Analysis in Ethylene Response Factors Uncovers Binding Motifs Critical to Their Diverse Functions

APETALA2/ETHYLENE RESPONSE FACTOR transcription factors (AP2/ERFs) play crucial roles in adaptation to stresses such as those caused by pathogens, wounding and cold. Although their name suggests a specific role in ethylene signalling, some ERF members also co-ordinate signals regulated by other key plant stress hormones such as jasmonate, abscisic acid and salicylate. We analysed a set of ERF proteins from three divergent plant species for intrinsically disorder regions containing conserved segments involved in protein–protein interaction known as Molecular Recognition Features (MoRFs). Then we correlated the MoRFs identified with a number of known functional features where these could be identified. Our analyses suggest that MoRFs, with plasticity in their disordered surroundings, are highly functional and may have been shuffled between related protein families driven by selection. A particularly important role may be played by the alpha helical component of the structured DNA binding domain to permit specificity. We also present examples of computationally identified MoRFs that have no known function and provide a valuable conceptual framework to link both disordered and ordered structural features within this family to diverse function.

Electronegativity and intrinsic disorder of preeclampsia-related proteins

Preeclampsia, hemorrhage, and infection are the leading causes of maternal death in underdeveloped countries. Since several proteins associated with preeclampsia are known, we conducted a computational study in which evaluated the commonness and potential functionality of intrinsic disorder in these proteins and also made an attempt to characterize their origin. To this end, we used a several supervised techniques, as a Polarity Index Method (PIM), which evaluates the electronegativity of proteins from their sequence alone. Peculiarities of resulting polar profile of the group of preeclampsia-related proteins were then compared with profiles of a group of lipoproteins, antimicrobial peptides, angiogenesis-related proteins, and the intrinsically disorder proteins. Our results showed a high graphical correlation between preeclampsia proteins, lipoproteins, and the angiogenesis proteins. These results lead us to strongly assume that the preeclampsia proteins are lipoproteins. We also show that several preeclampsia-related proteins contain significant amounts of functional disorder.