Search

Plasma protein biomarkers have been considered promising tools for diagnosing dementia subtypes due to their low variability, cost-effectiveness, and minimal invasiveness in diagnostic procedures. Machine learning (ML) methods have been applied to enhance accuracy of the biomarker discovery. However, previous ML-based studies often overlook interactions between proteins, which are crucial in complex disorders like dementia. While protein–protein interactions (PPIs) have been used in network models, these models often fail to fully capture the diverse properties of PPIs due to their local awareness. This drawback increases the chance of neglecting critical components and magnifying the impact of noisy interactions. In this study, we propose a novel graph-based ML model for dementia subtype diagnosis, the graph propagational network (GPN). By propagating the independent effect of plasma proteins on PPI network, the GPN extracts the globally interactive effects between proteins. Experimental results showed that the interactive effect between proteins yielded to further clarify the differences between dementia subtype groups and contributed to the performance improvement where the GPN outperformed existing methods by 10.4% on average. [ABSTRACT FROM AUTHOR]

Background: The amount of gray matter can measure the structural characteristics of the brain, and the deterioration of brain function can be evaluated accordingly. In particular, the lower the amount of gray matter in areas involved in memory, the lower the cognitive function, eventually affecting the onset of dementia, such as Alzheimer's disease (AD). Although there are various factors for changes in gray matter volume, genetic factors are also closely related. In addition, genetic factors are not only independently involved in biological activities, but also have complex interactions. Method: In this study, we construct a SNP network that reflects interactions between SNPs and propose a model that can predict the risk score for AD and the gray matter volume of the brain by region from the SNP network. First, in the SNP network, interactions between SNPs are measured as similarities based on the linkage disequilibrium score. Then, the proposed method constructs a regression model to predict brain volume and risk scores for AD using feature propagation, which is conducted to refine existing genetic variant information. Result: The proposed method was applied to data including both genetic and brain imaging information among UK Biobank (UKBB; 8,273 cases and 25,353 controls, total 33,626 participants) and Alzheimer's Disease Neuroimaging Initiative (ADNI; 202 cases and 720 controls, total 922 participants) data. The results were cross‐validated from the two datasets and compared with other models including the conventional polygenic risk score with pruning and thresholding (PRS‐PT) model. Compared to the other methods, the proposed method showed outstanding performance and the discovery of genetic variants and brain regions affecting Alzheimer's disease. Conclusion: We proposed the framework for predicting the volume of the brain with refined features of SNP that reflects the interactions between elements. The results revealed that genetic relations and inferring brain volumes are crucial for enhancing AD risk prediction. Furthermore, through the proposed method, it is expected that understanding and interpreting the phased mechanism of the onset of AD through brain lesions from genetic variants will be possible. [ABSTRACT FROM AUTHOR]

Background: Although drug discoveries can provide meaningful insights and significant enhancements in pharmaceutical field, the longevity and cost that it takes can be extensive where the success rate is low. In order to circumvent the problem, there has been increased interest in 'Drug Repositioning' where one searches for already approved drugs that have high potential of efficacy when applied to other diseases. To increase the success rate for drug repositioning, one considers stepwise screening and experiments based on biological reactions. Given the amount of drugs and diseases, however, the one-by-one procedure may be time consuming and expensive.Methods: In this study, we propose a machine learning based approach for efficiently selecting candidate diseases and drugs. We assume that if two diseases are similar, then a drug for one disease can be effective against the other disease too. For the procedure, we first construct two disease networks; one with disease-protein association and the other with disease-drug information. If two networks are dissimilar, in a sense that the edge distribution of a disease node differ, it indicates high potential for repositioning new candidate drugs for that disease. The Kullback-Leibler divergence is employed to measure difference of connections in two constructed disease networks. Lastly, we perform repositioning of drugs to the top 20% ranked diseases.Results: The results showed that F-measure of the proposed method was 0.75, outperforming 0.5 of greedy searching for the entire diseases. For the utility of the proposed method, it was applied to dementia and verified 75% accuracy for repositioned drugs assuming that there are not any known drugs to be used for dementia.Conclusion: This research has novelty in that it discovers drugs with high potential of repositioning based on disease networks with the quantitative measure. Through the study, it is expected to produce profound insights for possibility of undiscovered drug repositioning. [ABSTRACT FROM AUTHOR]