Magnetoencephalogram-based brain-computer interface for hand-gesture decoding using deep learning.

Advancements in deep learning algorithms over the past decade have led to extensive developments in brain-computer interfaces (BCI). A promising imaging modality for BCI is magnetoencephalography (MEG), which is a non-invasive functional imaging technique. The present study developed a MEG sensor-based BCI neural network to decode Rock-Paper-scissors gestures (MEG-RPSnet). Unique preprocessing pipelines in tandem with convolutional neural network deep-learning models accurately classified gestures. On a single-trial basis, we found an average of 85.56% classification accuracy in 12 subjects. Our MEG-RPSnet model outperformed two state-of-the-art neural network architectures for electroencephalogram-based BCI as well as a traditional machine learning method, and demonstrated equivalent and/or better performance than machine learning methods that have employed invasive, electrocorticography-based BCI using the same task. In addition, MEG-RPSnet classification performance using an intra-subject approach outperformed a model that used a cross-subject approach. Remarkably, we also found that when using only central-parietal-occipital regional sensors or occipitotemporal regional sensors, the deep learning model achieved classification performances that were similar to the whole-brain sensor model. The MEG-RSPnet model also distinguished neuronal features of individual hand gestures with very good accuracy. Altogether, these results show that noninvasive MEG-based BCI applications hold promise for future BCI developments in hand-gesture decoding.

Identifying complex motifs in massive omics data with a variable-convolutional layer in deep neural network.

Motif identification is among the most common and essential computational tasks for bioinformatics and genomics. Here we proposed a novel convolutional layer for deep neural network, named variable convolutional (vConv) layer, for effective motif identification in high-throughput omics data by learning kernel length from data adaptively. Empirical evaluations on DNA-protein binding and DNase footprinting cases well demonstrated that vConv-based networks have superior performance to their convolutional counterparts regardless of model complexity. Meanwhile, vConv could be readily integrated into multi-layer neural networks as an 'in-place replacement' of canonical convolutional layer. All source codes are freely available on GitHub for academic usage.

Spatial-linked alignment tool (SLAT) for aligning heterogenous slices.

Spatially resolved omics technologies reveal the spatial organization of cells in various biological systems. Here we propose SLAT (Spatially-Linked Alignment Tool), a graph-based algorithm for efficient and effective alignment of spatial slices. Adopting a graph adversarial matching strategy, SLAT is the first algorithm capable of aligning heterogenous spatial data across distinct technologies and modalities. Systematic benchmarks demonstrate SLAT's superior precision, robustness, and speed over existing state-of-the-arts. Applications to multiple real-world datasets further show SLAT's utility in enhancing cell-typing resolution, integrating multiple modalities for regulatory inference, and mapping fine-scale spatial-temporal changes during development. The full SLAT package is available at https://github.com/gao-lab/SLAT .

Machine Learning Applied to Omics Datasets Predicts Mortality in Patients with Alcoholic Hepatitis.

Alcoholic hepatitis is a major health care burden in the United States due to significant morbidity and mortality. Early identification of patients with alcoholic hepatitis at greatest risk of death is extremely important for proper treatments and interventions to be instituted. In this study, we used gradient boosting, random forest, support vector machine and logistic regression analysis of laboratory parameters, fecal bacterial microbiota, fecal mycobiota, fecal virome, serum metabolome and serum lipidome to predict mortality in patients with alcoholic hepatitis. Gradient boosting achieved the highest AUC of 0.87 for both 30-day mortality prediction using the bacteria and metabolic pathways dataset and 90-day mortality prediction using the fungi dataset, which showed better performance than the currently used model for end-stage liver disease (MELD) score.

[Construction of lentiviral mediated CyPA siRNA and its functions in non-small cell lung cancer].

OBJECTIVE: To construct a lentiviral-vector-mediated CyPA small interference RNA (siRNA) and study its function in non-small cell lung cancer. METHODS: First, four target sequences were selected according to CyPA mRNA sequence, the complementary DNA contained both sense and antisense oligonucleotides were designed, synthesized and cloned into the pGCL-GFP vector, which contained U6 promoter and green fluorescent protein (GFP). The resulting lentiviral vector containing CyPA shRNA was named Lv-shCyPA, and it was confirmed by PCR and sequencing. Next, it was cotransfected by Lipofectamine 2000 along with pHelper1.0 and pHelper 2.0 into 293T cells to package lentivirus particles. At the same time, the packed virus infected non-small cell lung cancer cell (A549), the level of CyPA protein at 5 d after infection was detected by Western Blot to screen the target of CyPA. A549 were infected with Lv-shCyPA and grown as xenografts in severe combined immunodeficient mice. Cell cycle and apoptosis were measured by FCM. RESULTS: It was confirmed by PCR and DNA sequencing that lentiviral-vector-mediated CyPA siRNA (Lv-shCyPA) producing CyPA shRNA was constructed successfully. The titer of concentrated virus were 1 x 10(7) TU/ml. Flow cytometric analysis demonstrated G2-M phase (11.40% +/- 0.68%) was decreased relatively in A549/LvshCyPA compared with control groups (14.52% +/- 1.19%) (P<0.05). The apoptosis rate of A549/Lv-shCyPA (5.01% +/- 0.5%) was higher than control groups (0.35% +/- 0.17%) (P<0.05). Visible tumors were only detectable at 6th day after inoculated by A549/Lv-shCyPA. The xenograft tumors of A549/Lv-shCyPA remarkably delayed tumor growth and remained at a similarly small average size at 38th days after inoculation compared with the control group (P < 0.05). CONCLUSION: Lentiviral-vector-mediated siRNA technique effectively inhibits the expression of CyPA, induces the NSCLC cell apoptosis, inhibits the tumor growth. Elucidation of the precise role of CypA in these pathways may lead to new targeted therapies for non-small cell lung cancer.

Suicides and medically serious attempters are of the same population in Chinese rural young adults.

BACKGROUND: Suicide rates in China are among the highest in the world, although there has been a decreasing trend in the past few years. One practical approach to study the characteristics and risk factors of suicide is to interview the suicide attempters. It was to compare completed suicides with serious attempters that may shed lights on suicide prevention strategies. METHOD: This is a combination of two case control studies for suicide completers and suicide attempters respectively. After a sample of suicides (n=392) and community living controls (n=416) were obtained and studied in rural China, we collected in the same rural areas data of suicide attempt and studied 507 medically serious attempters and 503 community counterparts. RESULTS: Characteristics and previously observed risk factors were compared between the suicides and the attempters, and we found that the demographic characteristics and risk factors for the suicides were also for the medically serious attempters but at some lesser degrees for the attempters than for the suicides. It was especially true of suicide intent, deficient coping, negative life events, and impulsivity. While most of the demographic characteristics were not significantly different between the suicides and the attempters, most of the clinical variables could distinguish the two groups. CONCLUSIONS: The suicide victims and the serious attempters could be of the same group of people who were at the edge of fatal self-injury, and the same clinical risk factors but of different degrees have divided them into the life and death groups.

Risk factors for rural young suicide in China: a case-control study.

BACKGROUND: Suicide is the fifth most important cause of death in China and the leading cause of death among young adults aged 15-34 years. The suicide rate in rural China is three times higher than the urban suicide rate, and the rate in women is higher than in men. METHODS: Sixteen counties from three provinces were selected as sampling sites in which 392 suicide cases and 416 community living controls were obtained. For each suicide case and control there were two informants who provided the target person's information. A structured questionnaire including demographics, social and familial environments, and personal characteristics was administered to the informants. RESULTS: Mental disorders and high hopelessness were found to be strongly related to suicide among Chinese rural young adults. Other suicide risk factors among this population were negative life events, never married but dating, suicide history in family, lack of positive coping skills, lack of social support, dysfunctional impulsivity, and not being a Communist Party/League member. CONCLUSIONS: The prevalence of mental disorders, although the strongest risk factor among rural young adult suicides in this study, was markedly lower than that in Western countries. Some of the risk factors found in the comprehensive analyses are specific to Chinese culture. "Being a Communist Party/League member" as a protective factor for suicide among Chinese rural youths requires further study and appropriate interpretation.

Psychological strains and youth suicide in rural China.

The strain theory of suicide postulates that suicide is usually preceded by psychological strains. A strain can be a consequence of any of four conflicts: differential values, aspiration and reality, relative deprivation, and lack of coping skills for a crisis. This study, with a blend of psychiatric and social predictors of suicide, identified correlates of suicide that are relevant to Chinese culture and tested the strain theory of suicide with Chinese data. We sampled 392 suicides and 416 living controls (both aged 15-34 years) from 16 rural counties in China in 2008 and interviewed two informants for each suicide and each control. We found that marriage and religion/religiosity did not distinguish the suicides from the living controls among Chinese rural young women. Religion/religiosity tended to be stronger for suicides than for controls. Psychological strains in the forms of relative deprivation, unrealized aspiration, and lack of coping skills were significantly associated with suicide, even after accounting for the role of mental illness. The strain theory of suicide forms a challenge to the psychiatric model popular in the West, at least in explaining the Chinese suicide.

[Preventive and therapeutic effects of recombinant IFN-alpha2b nasal spray on SARS-CoV infection in Macaca mulata].

OBJECTIVE: To study the preventive and therapeutic effects of recombinant IFN-alpha2b for nasal spray on SARS-CoV infection in Macaca mulata (rhesus monkey). METHODS: Ten rhesus monkeys were divided into two groups, 5 in interferon group, and 5 in control group. Before and after SARS-CoV attack, the virus was detected in samples such as pharyngeal swab in all the two groups by Real-time PCR (RT-PCR) and virus isolation was performed. RESULTS: After virus attack, the level of SARS-CoV-specific IgG and neutralizing antibody were induced by SARS-CoV in the interferon group was weaker than in control group. Hematology items showed no apparent changes after virus attack in treated group. Through pathological examination, the morphology of the lung tissues of two Macaques in the treated group was normal, while the other three displayed the interstitial pneumonia with the thickened septum and infiltration with mononuclear cells. Among which, one monkey showed part of thickened septum fused with each other. These lesions in the interferon treated animals were similar to those seen in the animals in control group, but with smaller scope of pathological changes. No significant abnormity was detected in other organs. CONCLUSION: Recombinant IFN-alpha2b could effectively interdict or weaken SARS-CoV injury in monkeys.

[Potent neutralization antibody elicited in mice by SARS-associated coronavirus spike protein S1 domain].

OBJECTIVE: To study the antigenicity of SARS associated coronavirus (CoV) spike S1 (12-672Aa) domain. METHODS: BALB/c mice were immunized with a plasmid bearing codon-optimized SARS-CoV (Tor2 strain) S1 domain and then boosted with purified S1 protein; the SARS-CoV specific IgG antibody was tested by ELISA and neutralization antibody was determined by in vitro microneutralization assay. RESULTS: S1 domain of SARS-CoV spike, which has been demonstrated harboring the receptor binding domain, successfully elicited SARS-CoV specific IgG antibody in mouse after combined immunization with DNA and purified S1 protein; the antibody elicited solely by S1 could potently neutralize SARS-CoV (HKU-39849) in vitro, 50% of 1 000 TCID50 SARS-CoV challenged cells were protected from viral infection by a 1:1499.68 dilution of mice sera immunized with S1 protein, but negative control sera showed no protection. CONCLUSION: S1 domain of SARS-CoV spike protein, which is responsible for receptor binding, can efficiently and sufficiently induce highly potent neutralizing antibody in mice. This result suggested that S1 domain could be an effective subunit vaccines against SARS-CoV.