Automatic recognition of schizophrenia from brain-network features using graph convolutional neural network.

Schizophrenia is a severe mental illness that imposes considerable economic burden on families and society. However, its clinical diagnosis primarily relies on scales and doctors' clinical experience and lacks an objective and accurate diagnostic approach. In recent years, graph convolutional neural networks (GCN) have been used to assist in psychiatric diagnosis owing to their ability to learn spatial-association information. Therefore, this study proposes a schizophrenia automatic recognition model based on graph convolutional neural network. Herein, the resting-state electroencephalography (EEG) data of 103 first-episode schizophrenia patients and 92 normal controls (NCs) were obtained. The automatic recognition model was trained with a nodal feature matrix that comprised the time and frequency-domain features of the EEG signals and local features of the brain network. The most significant regions that contributed to the model classification were identified, and the correlation between the node topological features of each significant region and clinical evaluation metrics was explored. Experiments were conducted to evaluate the performance of the model using 10-fold cross-validation. The best performance in the theta frequency band with a 6 s epoch length and phase-locked value. The recognition accuracy was 90.01%. The most significant region for identifying with first-episode schizophrenia patients and NCs was located in the parietal lobe. The results of this study verify the applicability of the proposed novel method for the identification and diagnosis of schizophrenia.

Method for persistent topological features extraction of schizophrenia patients' electroencephalography signal based on persistent homology.

With the development of network science and graph theory, brain network research has unique advantages in explaining those mental diseases, the neural mechanism of which is unclear. Additionally, it can provide a new perspective in revealing the pathophysiological mechanism of brain diseases from the system level. The selection of threshold plays an important role in brain networks construction. There are no generally accepted criteria for determining the proper threshold. Therefore, based on the topological data analysis of persistent homology theory, this study developed a multi-scale brain network modeling analysis method, which enables us to quantify various persistent topological features at different scales in a coherent manner. In this method, the Vietoris-Rips filtering algorithm is used to extract dynamic persistent topological features by gradually increasing the threshold in the range of full-scale distances. Subsequently, the persistent topological features are visualized using barcodes and persistence diagrams. Finally, the stability of persistent topological features is analyzed by calculating the Bottleneck distances and Wasserstein distances between the persistence diagrams. Experimental results show that compared with the existing methods, this method can extract the topological features of brain networks more accurately and improves the accuracy of diagnostic and classification. This work not only lays a foundation for exploring the higher-order topology of brain functional networks in schizophrenia patients, but also enhances the modeling ability of complex brain systems to better understand, analyze, and predict their dynamic behaviors.

Transcript profiling of the meiotic drive phenotype in testis of Aedes aegypti using suppressive subtractive hybridization.

The meiotic drive gene in Aedes aegypti is tightly linked with the sex determination locus on chromosome 1, and causes highly male-biased sex ratios. We prepared cDNA libraries from testes from the Ae. aegypti T37 strain (driving) and RED strain (non-driving), and used suppressive subtraction hybridization techniques to enrich for T37 testes-specific transcripts. Expressed sequence tags (ESTs) were obtained from a total of 2784 randomly selected clones from the subtracted T37 (subT37) library as well as the primary libraries for each strain (pT37 and pRED). Sequence analysis identified a total of 171 unique genes in the subT37 library and 299 unique genes among the three libraries. The majority of genes enriched in the subT37 library were associated with signal transduction, development, reproduction, metabolic process and cell cycle functions. Further, as observed with meiotic drive systems in Drosophila and mouse, a number of these genes were associated with signaling cascades that involve the Ras superfamily of regulatory small GTPases. Differential expression of several of these genes was verified in Ae. aegypti pupal testes using qRT-PCR. This study increases our understanding of testes gene expression enriched in adult males from the meiotic drive strain as well as insights into the basic testes transcriptome in Ae. aegypti.