Progressive Frame-Proposal Mining for Weakly Supervised Video Object Detection.

In this paper, we focus on the weakly supervised video object detection problem, where each training video is only tagged with object labels, without any bounding box annotations of objects. To effectively train object detectors from such weakly-annotated videos, we propose a Progressive Frame-Proposal Mining (PFPM) framework by exploiting discriminative proposals in a coarse-to-fine manner. First, we design a flexible Multi-Level Selection (MLS) scheme, with explicit guidance of video tags. By selecting object-relevant frames and mining important proposals from these frames, the proposed MLS can effectively reduce frame redundancy as well as improve proposal effectiveness to boost weakly-supervised detectors. Moreover, we develop a novel Holistic-View Refinement (HVR) scheme, which can globally evaluate importance of proposals among frames, and thus correctly refine pseudo ground truth boxes for training video detectors in a self-supervised manner. Finally, we evaluate the proposed PFPM on a large-scale benchmark for video object detection, on ImageNet VID, under the setting of weak annotations. The experimental results demonstrate that our PFPM significantly outperforms the state-of-the-art weakly-supervised detectors.

The metabolic slowdown caused by the deletion of pspA accelerates protein aggregation during stationary phase facilitating antibiotic persistence.

Entering a dormant state is a prevailing mechanism used by bacterial cells to transiently evade antibiotic attacks and become persisters. The dynamic progression of bacterial dormancy depths driven by protein aggregation has been found to be critical for antibiotic persistence in recent years. However, our current understanding of the endogenous genes that affects dormancy depth remains limited. Here, we discovered a novel role of phage shock protein A (pspA) gene in modulating bacterial dormancy depth. Deletion of pspA of Escherichia coli resulted in increased bacterial dormancy depths and prolonged lag times for resuscitation during the stationary phase. ∆pspA exhibited a higher persister ratio compared to the wild type when challenged with various antibiotics. Microscopic images revealed that ∆pspA showed accelerated formation of protein aggresomes, which were collections of endogenous protein aggregates. Time-lapse imaging established the positive correlation between protein aggregation and antibiotic persistence of ∆pspA at the single-cell level. To investigate the molecular mechanism underlying accelerated protein aggregation, we performed transcriptome profiling and found the increased abundance of chaperons and a general metabolic slowdown in the absence of pspA. Consistent with the transcriptomic results, the ∆pspA strain showed a decreased cellular ATP level, which could be rescued by glucose supplementation. Then, we verified that replenishment of cellular ATP levels by adding glucose could inhibit protein aggregation and reduce persister formation in ∆pspA. This study highlights the novel role of pspA in maintaining proteostasis, regulating dormancy depth, and affecting antibiotic persistence during stationary phase.

cfOmics: a cell-free multi-Omics database for diseases.

Liquid biopsy has emerged as a promising non-invasive approach for detecting, monitoring diseases, and predicting their recurrence. However, the effective utilization of liquid biopsy data to identify reliable biomarkers for various cancers and other diseases requires further exploration. Here, we present cfOmics, a web-accessible database (https://cfomics.ncRNAlab.org/) that integrates comprehensive multi-omics liquid biopsy data, including cfDNA, cfRNA based on next-generation sequencing, and proteome, metabolome based on mass-spectrometry data. As the first multi-omics database in the field, cfOmics encompasses a total of 17 distinct data types and 13 specimen variations across 69 disease conditions, with a collection of 11345 samples. Moreover, cfOmics includes reported potential biomarkers for reference. To facilitate effective analysis and visualization of multi-omics data, cfOmics offers powerful functionalities to its users. These functionalities include browsing, profile visualization, the Integrative Genomic Viewer, and correlation analysis, all centered around genes, microbes, or end-motifs. The primary objective of cfOmics is to assist researchers in the field of liquid biopsy by providing comprehensive multi-omics data. This enables them to explore cell-free data and extract profound insights that can significantly impact disease diagnosis, treatment monitoring, and management.

COMMO: a web server for the identification and analysis of consensus gene modules across multiple methods.

SUMMARY: A variety of computational methods have been developed to identify functionally related gene modules from genome-wide gene expression profiles. Integrating the results of these methods to identify consensus modules is a promising approach to produce more accurate and robust results. In this application note, we introduce COMMO, the first web server to identify and analyze consensus gene functionally related gene modules from different module detection methods. First, COMMO implements eight state-of-the-art module detection methods and two consensus clustering algorithms. Second, COMMO provides users with mRNA and protein expression data for 33 cancer types from three public databases. Users can also upload their own data for module detection. Third, users can perform functional enrichment and two types of survival analyses on the observed gene modules. Finally, COMMO provides interactive, customizable visualizations and exportable results. With its extensive analysis and interactive capabilities, COMMO offers a user-friendly solution for conducting module-based precision medicine research. AVAILABILITY AND IMPLEMENTATION: COMMO web is available at https://commo.ncpsb.org.cn/, with the source code available on GitHub: https://github.com/Song-xinyu/COMMO/tree/master.

Identification of functional gene modules by integrating multi-omics data and known molecular interactions.

Multi-omics data integration has emerged as a promising approach to identify patient subgroups. However, in terms of grouping genes (or gene products) into co-expression modules, data integration methods suffer from two main drawbacks. First, most existing methods only consider genes or samples measured in all different datasets. Second, known molecular interactions (e.g., transcriptional regulatory interactions, protein-protein interactions and biological pathways) cannot be utilized to assist in module detection. Herein, we present a novel data integration framework, Correlation-based Local Approximation of Membership (CLAM), which provides two methodological innovations to address these limitations: 1) constructing a trans-omics neighborhood matrix by integrating multi-omics datasets and known molecular interactions, and 2) using a local approximation procedure to define gene modules from the matrix. Applying Correlation-based Local Approximation of Membership to human colorectal cancer (CRC) and mouse B-cell differentiation multi-omics data obtained from The Cancer Genome Atlas (TCGA), Clinical Proteomics Tumor Analysis Consortium (CPTAC), Gene Expression Omnibus (GEO) and ProteomeXchange database, we demonstrated its superior ability to recover biologically relevant modules and gene ontology (GO) terms. Further investigation of the colorectal cancer modules revealed numerous transcription factors and KEGG pathways that played crucial roles in colorectal cancer progression. Module-based survival analysis constructed four survival-related networks in which pairwise gene correlations were significantly correlated with colorectal cancer patient survival. Overall, the series of evaluations demonstrated the great potential of Correlation-based Local Approximation of Membership for identifying modular biomarkers for complex diseases. We implemented Correlation-based Local Approximation of Membership as a user-friendly application available at https://github.com/free1234hm/CLAM.

Study on the Mental Health of the Elderly under Different Pension Models.

OBJECTIVE: To compare the mental health status of the elderly under different pension modes and to provide evidence for improving institutional services and the quality of life of the elderly. METHODS: A total of 118 elderly people in social welfare homes, nursing homes, and elderly welfare centers in D city and 165 elderly people from families in D city were assessed by self-made questionnaire, Geriatric Depression Scale (GDS), activities of daily living scale (ADL), and social support rating scale (SSRS). RESULTS: The total scores of mental health and self, emotion, and adaptation subscales in the social group were higher than those in the home group, with a significant difference (p > 0.05). The scores of cognitive and interpersonal subscales in the home group were higher than those in the social group, but the differences were not significant (p > 0.05). Under the mode of family pension and social institution pension, the health status of the elderly has certain differences. The elderly in different old-age care modes have good performance in diet and sleep, and there is no statistical difference between them (p > 0.05). CONCLUSION: The investigation shows that the mental health status of the elderly under the family pension model is obviously better than that under the social institution pension model.

MoGCN: A Multi-Omics Integration Method Based on Graph Convolutional Network for Cancer Subtype Analysis.

In light of the rapid accumulation of large-scale omics datasets, numerous studies have attempted to characterize the molecular and clinical features of cancers from a multi-omics perspective. However, there are great challenges in integrating multi-omics using machine learning methods for cancer subtype classification. In this study, MoGCN, a multi-omics integration model based on graph convolutional network (GCN) was developed for cancer subtype classification and analysis. Genomics, transcriptomics and proteomics datasets for 511 breast invasive carcinoma (BRCA) samples were downloaded from the Cancer Genome Atlas (TCGA). The autoencoder (AE) and the similarity network fusion (SNF) methods were used to reduce dimensionality and construct the patient similarity network (PSN), respectively. Then the vector features and the PSN were input into the GCN for training and testing. Feature extraction and network visualization were used for further biological knowledge discovery and subtype classification. In the analysis of multi-dimensional omics data of the BRCA samples in TCGA, MoGCN achieved the highest accuracy in cancer subtype classification compared with several popular algorithms. Moreover, MoGCN can extract the most significant features of each omics layer and provide candidate functional molecules for further analysis of their biological effects. And network visualization showed that MoGCN could make clinically intuitive diagnosis. The generality of MoGCN was proven on the TCGA pan-kidney cancer datasets. MoGCN and datasets are public available at https://github.com/Lifoof/MoGCN. Our study shows that MoGCN performs well for heterogeneous data integration and the interpretability of classification results, which confers great potential for applications in biomarker identification and clinical diagnosis.

[Integration-based co-expression network analysis to investigate tumor-associated modules across three cancer types].

In case/control gene expression data, differential expression (DE) represents changes in gene expression levels across various biological conditions, whereas differential co-expression (DC) represents an alteration of correlation coefficients between gene pairs. Both DC and DE genes have been studied extensively in human diseases. However, effective approaches for integrating DC-DE analyses are lacking. Here, we report a novel analytical framework named DC&DEmodule for integrating DC and DE analyses and combining information from multiple case/control expression datasets to identify disease-related gene co-expression modules. This includes activated modules (gaining co-expression and up-regulated in disease) and dysfunctional modules (losing co-expression and down-regulated in disease). By applying this framework to microarray data associated with liver, gastric and colon cancer, we identified two, five and two activated modules and five, five and one dysfunctional module(s), respectively. Compared with the other methods, pathway enrichment analysis demonstrated the superior sensitivity of our method in detecting both known cancer-related pathways and those not previously reported. Moreover, we identified 17, 69, and 11 module hub genes that were activated in three cancers, which included 53 known and three novel cancer prognostic markers. Random forest classifiers trained by the hub genes showed an average of 93% accuracy in differentiating tumor and adjacent normal samples in the TCGA and GEO database. Comparison of the three cancers provided new insights into common and tissue-specific cancer mechanisms. A series of evaluations demonstrated the framework is capable of integrating the rapidly accumulated expression data and facilitating the discovery of dysregulated processes.

TSMiner: a novel framework for generating time-specific gene regulatory networks from time-series expression profiles.

Time-series gene expression profiles are the primary source of information on complicated biological processes; however, capturing dynamic regulatory events from such data is challenging. Herein, we present a novel analytic tool, time-series miner (TSMiner), that can construct time-specific regulatory networks from time-series expression profiles using two groups of genes: (i) genes encoding transcription factors (TFs) that are activated or repressed at a specific time and (ii) genes associated with biological pathways showing significant mutual interactions with these TFs. Compared with existing methods, TSMiner demonstrated superior sensitivity and accuracy. Additionally, the application of TSMiner to a time-course RNA-seq dataset associated with mouse liver regeneration (LR) identified 389 transcriptional activators and 49 transcriptional repressors that were either activated or repressed across the LR process. TSMiner also predicted 109 and 47 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways significantly interacting with the transcriptional activators and repressors, respectively. These findings revealed the temporal dynamics of multiple critical LR-related biological processes, including cell proliferation, metabolism and the immune response. The series of evaluations and experiments demonstrated that TSMiner provides highly reliable predictions and increases the understanding of rapidly accumulating time-series omics data.

[Research progress of feature selection and machine learning methods for mass spectrometry-based protein biomarker discovery].

With the development of mass spectrometry technologies and bioinformatics analysis algorithms, disease research-driven human proteome project (HPP) is advancing rapidly. Protein biomarkers play critical roles in clinical applications and the biomarker discovery strategies and methods have become one of research hotspots. Feature selection and machine learning methods have good effects on solving the "dimensionality" and "sparsity" problems of proteomics data, which have been widely used in the discovery of protein biomarkers. Here, we systematically review the strategy of protein biomarker discovery and the frequently-used machine learning methods. Also, the review illustrates the prospects and limitations of deep learning in this field. It is aimed at providing a valuable reference for corresponding researchers.

PANDA: A comprehensive and flexible tool for quantitative proteomics data analysis.

SUMMARY: As the experiment techniques and strategies in quantitative proteomics are improving rapidly, the corresponding algorithms and tools for protein quantification with high accuracy and precision are continuously required to be proposed. Here, we present a comprehensive and flexible tool named PANDA for proteomics data quantification. PANDA, which supports both label-free and labeled quantifications, is compatible with existing peptide identification tools and pipelines with considerable flexibility. Compared with MaxQuant on several complex datasets, PANDA was proved to be more accurate and precise with less computation time. Additionally, PANDA is an easy-to-use desktop application tool with user-friendly interfaces. AVAILABILITY AND IMPLEMENTATION: PANDA is freely available for download at https://sourceforge.net/projects/panda-tools/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

KeggExp: a web server for visual integration of KEGG pathways and expression profile data.

SUMMARY: Effective visualization is important for knowledge discovery when analysing expression profile data. However, existing tools for visually integrating expression profile data with KEGG pathway maps lack extensive interactive visualization operations. KeggExp simultaneously presents the pathway map of one pathway, dendrogram and heatmap of the genes in the pathway and scatter map of one gene; and also provides interactive operations for highlighting specific genes on the pathway map, including differentially-expressed genes, co-expressed genes selected from the heatmap and user-input genes. With KeggExp, researchers, including those without programming skills, can take advantage of its interactive operations to determine key genes or pathways when analysing expression profile data. AVAILABILITY AND IMPLEMENTATION: Freely available at http://www.fgvis.com/expressvis/KeggExp/. Language: JavaScript, python; Libraries: D3.js, Rxjs, Angular, Django, Django rest frame work, Scipy. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

PPIExp: A Web-Based Platform for Integration and Visualization of Protein-Protein Interaction Data and Spatiotemporal Proteomics Data.

Integrating spatiotemporal proteomics data with protein-protein interaction (PPI) data can help researchers make an in-depth exploration of their proteins of interest in a dynamic manner. However, there is still a lack of proper tools for the biologists who usually have few programming skills to construct a PPI network for a protein list, visualize active PPI subnetworks, and then select key nodes for further study. We propose a web-based platform named PPIExp that can automatically construct a PPI network, perform clustering analysis according to protein abundances, and perform functional enrichment analysis. More importantly, it provides multiple effective visualization interfaces, such as the interface to display the PPI network map, the interface to display a dendrogram and heatmap for the clustering result, and the interface to display the expression pattern of a selected protein. To visualize the active PPI subnetworks in specific space or time, it provides buttons to highlight the differentially expressed proteins under each condition on the network map. Additionally, to help researchers determine which proteins are worth further attention, PPIExp provides extensive one-click interactive operations to map node centrality measures to node size on the network and highlight three types of proteins, that is, the proteins in an enriched functional term, the coexpressed proteins selected from the dendgrogram and heatmap, and the proteins input by users. PPIExp is available at http://www.fgvis.com/expressvis/PPIExp .

Deterministic entanglement generation from driving through quantum phase transitions.

Many-body entanglement is often created through the system evolution, aided by nonlinear interactions between the constituting particles. These very dynamics, however, can also lead to fluctuations and degradation of the entanglement if the interactions cannot be controlled. Here, we demonstrate near-deterministic generation of an entangled twin-Fock condensate of ~11,000 atoms by driving a arubidium-87 Bose-Einstein condensate undergoing spin mixing through two consecutive quantum phase transitions (QPTs). We directly observe number squeezing of 10.7 ± 0.6 decibels and normalized collective spin length of 0.99 ± 0.01. Together, these observations allow us to infer an entanglement-enhanced phase sensitivity of ~6 decibels beyond the standard quantum limit and an entanglement breadth of ~910 atoms. Our work highlights the power of generating large-scale useful entanglement by taking advantage of the different entanglement landscapes separated by QPTs.

[Applications of meta-analysis in multi-omics].

As a statistical method integrating multi-features and multi-data, meta-analysis was introduced to the field of life science in the 1990s. With the rapid advances in high-throughput technologies, life omics, the core of which are genomics, transcriptomics and proteomics, is becoming the new hot spot of life science. Although the fast output of massive data has promoted the development of omics study, it results in excessive data that are difficult to integrate systematically. In this case, meta-analysis is frequently applied to analyze different types of data and is improved continuously. Here, we first summarize the representative meta-analysis methods systematically, and then study the current applications of meta-analysis in various omics fields, finally we discuss the still-existing problems and the future development of meta-analysis.

Systematic analyses of the transcriptome, translatome, and proteome provide a global view and potential strategy for the C-HPP.

To estimate the potential of the state-of-the-art proteomics technologies on full coverage of the encoding gene products, the Chinese Human Chromosome Proteome Consortium (CCPC) applied a multiomics strategy to systematically analyze the transciptome, translatome, and proteome of the same cultured hepatoma cells with varied metastatic potential qualitatively and quantitatively. The results provide a global view of gene expression profiles. The 9064 identified high confident proteins covered 50.2% of all gene products in the translatome. Those proteins with function of adhesion, development, reproduction, and so on are low abundant in transcriptome and translatome but absent in proteome. Taking the translatome as the background of protein expression, we found that the protein abundance plays a decisive role and hydrophobicity has a greater influence than molecular weight and isoelectric point on protein detectability. Thus, the enrichment strategy used for low-abundant transcription factors helped to identify missing proteins. In addition, those peptides with single amino acid polymorphisms played a significant role for the disease research, although they might negligibly contribute to new protein identification. The proteome raw and metadata of proteome were collected using the iProX submission system and submitted to ProteomeXchange (PXD000529, PXD000533, and PXD000535). All detailed information in this study can be accessed from the Chinese Chromosome-Centric Human Proteome Database.

Systematic analysis of missing proteins provides clues to help define all of the protein-coding genes on human chromosome 1.

Our first proteomic exploration of human chromosome 1 began in 2012 (CCPD 1.0), and the genome-wide characterization of the human proteome through public resources revealed that 32-39% of proteins on chromosome 1 remain unidentified. To characterize all of the missing proteins, we applied an OMICS-integrated analysis of three human liver cell lines (Hep3B, MHCC97H, and HCCLM3) using mRNA and ribosome nascent-chain complex-bound mRNA deep sequencing and proteome profiling, contributing mass spectrometric evidence of 60 additional chromosome 1 gene products. Integration of the annotation information from public databases revealed that 84.6% of genes on chromosome 1 had high-confidence protein evidence. Hierarchical analysis demonstrated that the remaining 320 missing genes were either experimentally or biologically explainable; 128 genes were found to be tissue-specific or rarely expressed in some tissues, whereas 91 proteins were uncharacterized mainly due to database annotation diversity, 89 were genes with low mRNA abundance or unsuitable protein properties, and 12 genes were identifiable theoretically because of a high abundance of mRNAs/RNC-mRNAs and the existence of proteotypic peptides. The relatively large contribution made by the identification of enriched transcription factors suggested specific enrichment of low-abundance protein classes, and SRM/MRM could capture high-priority missing proteins. Detailed analyses of the differentially expressed genes indicated that several gene families located on chromosome 1 may play critical roles in mediating hepatocellular carcinoma invasion and metastasis. All mass spectrometry proteomics data corresponding to our study were deposited in the ProteomeXchange under the identifiers PXD000529, PXD000533, and PXD000535.

SILVER: an efficient tool for stable isotope labeling LC-MS data quantitative analysis with quality control methods.

SUMMARY: With the advance of experimental technologies, different stable isotope labeling methods have been widely applied to quantitative proteomics. Here, we present an efficient tool named SILVER for processing the stable isotope labeling mass spectrometry data. SILVER implements novel methods for quality control of quantification at spectrum, peptide and protein levels, respectively. Several new quantification confidence filters and indices are used to improve the accuracy of quantification results. The performance of SILVER was verified and compared with MaxQuant and Proteome Discoverer using a large-scale dataset and two standard datasets. The results suggest that SILVER shows high accuracy and robustness while consuming much less processing time. Additionally, SILVER provides user-friendly interfaces for parameter setting, result visualization, manual validation and some useful statistics analyses. AVAILABILITY AND IMPLEMENTATION: SILVER and its source codes are freely available under the GNU General Public License v3.0 at http://bioinfo.hupo.org.cn/silver.

[A comparative study on behaviors of two depression models in rats induced by chronic forced swimming stress].

OBJECTIVE: To compare the behaviors of rats with depressions induced by chronic forced swimming stress under two different conditions. METHODS: Eighteen male rats were randomly divided into 3 groups, with 6 rats in each group. The rats in the control group (C group) were not forced into swimming, while the rats in the stress groups (S1 and S2) were forced to swim for 14 consecutive days. The rats in S1 group and S2 group swam for five minutes every morning, in water with (23 +/- 1) degree C, and (10 +/- 0.5) degree C in temperature, respectively. The weight gain, food intake, open-field test and saccharin solution test were observed on the seventh day and fourteenth day. RESULTS: On the seventh day following chronic swim stress, the rats in the S2 group had significant lower ratio in weight gain and food intake than the controls (P < 0.05). On the fourteenth day, the rats in the S2 group had significant lower ratio in weight gain (12.26 +/- 4.04)%, food intake (9.49 +/- 0.96)%, sucrose intake (28.63 +/- 3.51) g, and preference for saccharin solution (76.25 +/- 2.51)%, and less number of crossing (12.17 +/- 9.00) and times of rearing (3.17 +/- 3.60) than the controls (P < 0.05). The rats in the S1 group had significant lower ratio in weight gain and food intake than the controls on the seventh day following forced swimming. On the fourteenth day, the rats in the S1 group still had lower ratio in weight gain, but had higher ratio in food intake and preference for saccharin solution, and greater number of crossing than the controls. CONCLUSION: Chronic forced swimming at a lower temperature could induce depression better than at a higher temperature.

Findings of P300-like and CNV-like potentials in rat model of depression following repeatedly forced swim stress.

The aim of this study was to explore whether there were abnormalities of CNV-like and P(300)-like potentials in stressed rats following repeatedly forced swim stress. Forty male rats were randomly divided into 4 groups: the control groups (Control-1 and Control-2) and the stressed groups (Stress-1 and Stress-2). Rats in stressed groups were administered to repeatedly forced swim 7 or 14 days respectively. Body weight gain, saccharin preference test and open field test were performed. After being anesthetized with urethane, P(300)-like potentials were evoked by the oddball auditory stimulation and CNV-like potentials were elicited by condition-test stimulus. Results of behavioral tests displayed less body weights and less saccharine solution intake in stressed rats and significant effects of stress on the number of crossing squares, the duration of rearing and the number of grooming in open field test. Prolonged P3 latencies and decreased P3 amplitudes of P(300)-like potentials were found in the stressed rats. CNV amplitudes were lower in the stressed rats than those in control. Moreover, there were significant correlations between parameters of ERPs (including P3 latency, amplitude and CNV amplitude) and a serial of behavioral traits. This study provides an important evidence of changes of CNV-like and P(300)-like potentials in depressed rats following repeatedly forced swim stress. Based on this study, ERPs should be taken into consideration and applied as the useful tools in the research work of depressed animal models.