An antibacterial peptides recognition method based on BERT and Text-CNN / 生物工程学报

Antimicrobial peptides (AMPs) are small molecule peptides that are widely found in living organisms with broad-spectrum antibacterial activity and immunomodulatory effect. Due to slower emergence of resistance, excellent clinical potential and wide range of application, AMP is a strong alternative to conventional antibiotics. AMP recognition is a significant direction in the field of AMP research. The high cost, low efficiency and long period shortcomings of the wet experiment methods prevent it from meeting the need for the large-scale AMP recognition. Therefore, computer-aided identification methods are important supplements to AMP recognition approaches, and one of the key issues is how to improve the accuracy. Protein sequences could be approximated as a language composed of amino acids. Consequently, rich features may be extracted using natural language processing (NLP) techniques. In this paper, we combine the pre-trained model BERT and the fine-tuned structure Text-CNN in the field of NLP to model protein languages, develop an open-source available antimicrobial peptide recognition tool and conduct a comparison with other five published tools. The experimental results show that the optimization of the two-phase training approach brings an overall improvement in accuracy, sensitivity, specificity, and Matthew correlation coefficient, offering a novel approach for further research on AMP recognition.

Advances of chromatogram retention time alignment algorithms in proteomics / 生物工程学报

Chromatography is a basic process in the current proteomics workflow, and the retention time alignment of the chromatogram is one of the important steps to effectively improve the identification and quantification accuracy. After years of development, a series of algorithms for retention time alignment have been developed. This review summarizes the advances of chromatographic retention time alignment algorithms and tools for proteomics analysis from the perspective of proteomics users, and discusses the development and future application directions.

Application of neural network autoencoder algorithm in the cancer informatics research / 生物工程学报

Cancers have been widely recognized as highly heterogeneous diseases, and early diagnosis and prognosis of cancer types have become the focus of cancer research. In the era of big data, efficient mining of massive biomedical data has become a grand challenge for bioinformatics research. As a typical neural network model, the autoencoder is able to efficiently learn the features of input data by unsupervised training method and further help integrate and mine the biological data. In this article, the primary structure and workflow of the autoencoder model are introduced, followed by summarizing the advances of the autoencoder model in cancer informatics using various types of biomedical data. Finally, the challenges and perspectives of the autoencoder model are discussed.

Integration-based co-expression network analysis to investigate tumor-associated modules across three cancer types / 生物工程学报

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.

Potential microenvironment of SARS-CoV-2 infection in airway epithelial cells revealed by Human Protein Atlas database analysis

The outbreak of COVID-19 has caused serious epidemic events in China and other countries. With the rapid spread of COVID-19, it is urgent to explore the pathogenesis of this novel coronavirus. However, the foundational research of COVID-19 is very weak. Although angiotensin converting enzyme 2 (ACE2) is the reported receptor of SARS-CoV-2, information about SARS-CoV-2 invading airway epithelial cells is very limited. Based on the analysis of the Human Protein Atlas database, we compared the virus-related receptors of epithelial-derived cells from different organs and found potential key molecules in the local microenvironment for SARS-CoV-2 entering airway epithelial cells. In addition, we found that these proteins were associated with virus reactive proteins in host airway epithelial cells, which may promote the activation of the immune system and the release of inflammatory factors. Our findings provide a new research direction for understanding the potential microenvironment required by SARS-CoV-2 infection in airway epithelial, which may assist in the discovery of potential drug targets against SARS-CoV-2 infection.

A new bioinformatics approach for prediction of potential tumor neoantigens based on the cancer genome atlas dataset / 生物工程学报

Tumor-specific gene mutations might generate suitable neoepitopes for cancer immunotherapy that are highly immunogenic and absent in normal tissues. The high heterogeneity of the tumor genome poses a big challenge for precision cancer immunotherapy. Mutations characteristic of each tumor can help to distinguish it from other tumors. Based on these mutations' characteristic, it is possible to develop immunotherapeutic strategies for specific tumors. In this study, a tumor neoantigen prediction scheme was proposed, in which both the intracellular antigen presentation process and the ability to bind with extracellular MHC molecule were taken into consideration. The overall design is meritorious and may help reduce the cost for validation experiments compared with conventional methods. This strategy was tested with several cancer genome datasets in the TCGA database, and a number of potential tumor neoantigens were predicted for each dataset. These predicted neoantigens showed tumor type specificity and were found in 20% to 70% of cancer patients. This scheme might prove useful clinically in future.

Advances in the research of extracellular matrix protein prediction tools / 生物工程学报

Extracellular matrix (ECM) proteins play an important role in a series of biological processes in the cell, and their abnormal regulation can lead to many diseases. The theoretical ECM reference dataset is the basis for efficient identification of extracellular matrix proteins. Researchers have developed various ECM protein prediction tools based on machine learning methods. In this review, the main strategy of development of ECM protein prediction tools that based on machine learning methods has been introduced. Then, advances and specific characters of the existing ECM protein prediction tools have been summarized. Finally, the challenges and possible improvements of ECM protein prediction tools are discussed.

Research progress of feature selection and machine learning methods for mass spectrometry-based protein biomarker discovery / 生物工程学报

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.

Progress in the spectral library based protein identification strategy / 生物工程学报

Exponential growth of the mass spectrometry (MS) data is exhibited when the mass spectrometry-based proteomics has been developing rapidly. It is a great challenge to develop some quick, accurate and repeatable methods to identify peptides and proteins. Nowadays, the spectral library searching has become a mature strategy for tandem mass spectra based proteins identification in proteomics, which searches the experiment spectra against a collection of confidently identified MS/MS spectra that have been observed previously, and fully utilizes the abundance in the spectrum, peaks from non-canonical fragment ions, and other features. This review provides an overview of the implement of spectral library search strategy, and two key steps, spectral library construction and spectral library searching comprehensively, and discusses the progress and challenge of the library search strategy.

Construction and application of Human Liver Disease Ontology / 军事医学

Objective China has the biggest population of commonly encountered liver diseases , so hepatology resear-ches are of great importance to human health .In order to comprehensively collect/organize and effectively share/display liv-er physiopathology knowledge scattered around databases and literature , a Human Liver Disease Ontology ( HuLDO ) is built and applied in many ways .Methods HuLDO systematically classified and annotated various human liver diseases based on extensive cross mapping and integration of several authoritative nomenclatures of diseases and two monographs of hepatology .Results In the present version , HuLDO encompassed knowledge of 227 human liver diseases and was enriched by plenty of synonyms,definitions, descriptions, references, and logical or pathological relations between different disea-ses.Compared with other existing nomenclatures , HuLDO extended far beyond them in the scope of liver diseases .Several applications of HuLDO in database construction , knowledge collection and text mining were also described .Conclusion HuLDO offers a sound basis for knowledge mining , data integration and data analysis in the field of hepatology .HuLDO is publicly available at ftp://liveratlas.hupo.org.cn/5.liver 5.liver%20Disease%20Ontology/.

Progress in proteogenomics of prokaryotes / 生物工程学报

With the rapid development of genome sequencing technologies, a large amount of prokaryote genomes have been sequenced in recent years. To further investigate the models and functions of genomes, the algorithms for genome annotations based on the sequence and homology features have been widely implemented to newly sequenced genomes. However, gene annotations only using the genomic information are prone to errors, such as the incorrect N-terminals and pseudogenes. It is even harder to provide reasonable annotating results in the case of the poor genome sequencing results. The transcriptomics based on the technologies such as microarray and RNA-seq and the proteomics based on the MS/MS have been used widely to identify the gene products with high throughput and high sensitivity, providing the powerful tools for the verification and correction of annotated genome. Compared with transcriptomics, proteomics can generate the protein list for the expressed genes in the samples or cells without any confusion of the non-coding RNA, leading the proteogenomics an important basis for the genome annotations in prokaryotes. In this paper, we first described the traditional genome annotation algorithms and pointed out the shortcomings. Then we summarized the advantages of proteomics in the genome annotations and reviewed the progress of proteogenomics in prokaryotes. Finally we discussed the challenges and strategies in the data analyses and potential solutions for the developments of proteogenomics.

Applications of meta-analysis in multi-omics / 生物工程学报

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.

Advance of Peptide Detectability Prediction on Mass Spectrometry Platform in Proteomics / 分析化学

As the complexity of samples and experimental processes, the repeatability of mass spectrometry experiments is still not satisfactory, the results of peptide identification and quantification show high randomicity), the probability of peptide being detected by mass spectrometry in proteome research, especially in quantitative proteomic study, has received much attention. Therefore, a lot of experimental researches have been done, as well as a number of computational prediction methods have been developed. In this article, we summarized the important factors impacting the peptide detectability, investigated the existing prediction methods) and reviewed their applications in experimental study.

Effect of continuous aspiration of subglotlic secretions on the prevention of ventilator-associated pneumonia in mechanically ventilated patients:a prospective, randomized, controlled clinical trial / 中华内科杂志

Objective To evaluate the effect of continuous aspiration of subglottic secretions (CASS) on the prevention of ventilator-associated pneumonia (VAP) in mechanically ventilated patients. Methods Patients ventilated mechanically at the ICU from October, 2004 to April,2006 were randomly divided into 2 groups: one group received CASS and the other did not (NASS group). CASS was performed immediately after admission for patients in the CASS group. The diagnosis d VAP was made based on clinical presentations, and the evaluation of YAP was done using simplified version of the clinical pulmonary infection score (CPIS). The general status of the patients, days of ventilated treatment, the volume of daily aspirated aubglottic secretions, the morbidity and timing of VAP, days of stay in ICU and mortality within 28 days of hospitalization were recorded. Results One hundred and one patients were included in the study. There were 48 patients in the CASS group who were treated with mechanical ventilation more than 48 hours,and 43 patients in the NASS group. There was no significant difference in the general status of the patients and days of ventilation between 2 groups with the averaged score of APACHE Ⅱ being 20.8± 6.1. The average of CPIS was of 5.6±1.0 when VAP was diagnosed. The mean volume of aspirated subglottic secretions within the first 24 hours in the CASS group (n=48) was (27.2±21.2)ml. The morbidity of VAP in the CASS and the NASS groups was 25.0% and 46. 5% respectively (P=0.032), and the length of time before the onset of VAP in these 2 groups was (7.3±4.2) days and (5.1±3.0) days respectively (P=0.100). There was a significant increase in the percentage of gram-positive cocci from the lower respiratory tracts in the NASS group compared with that in the CASS group (P=0.004). In the CASS group, the volume of the first daily aspirated subglottic secretions in patients with VAP was significantly less than that in patients without VAP(P =0.006). The morbidity of VAP in patients with failed early aspiration (the volume of first daily aspirated secretions≤20 ml) was significantly higher than that in patients in whom the aspiration was effective (P<0.01). The length of mechanical ventilation in patients with VAP was significantly longer than that in patients without VAP(P=0.000). The in-hospital mortality in patients with VAP was significantly higher than that in patients without VAP(P=0.009), and the mortality in 28 days after admission in patients with VAP was significantly higher than that in patients without VAP(P=0.035).Conclusion Effective continuous aspiration of subglottic secretions could significantly reduce the morbidity of early-onset VAP.

Methods and Strategies of Novel Proteins Identification in Proteomics / 生物化学与生物物理进展

The combination of tandem spectrometry and database searching is one of the most popular technologies for protein identification.However,only those proteins in the searching database could be identified,and current database is far from completeness.So it is necessary to mining the MS/MS data comprehensively,in which novel protein identification is the most important one.The definition of novel protein could be divided into three levels according to their annotations of sequences and functions.As a part of protein identification,the main approaches used to identify novel protein are basing on the following two different ways:de novo sequencing combined with similarity search and searching against nucleotide acid databases such as EST or genome databases.Several mature or newly developed methods and techniques were summarized,and the problems and strategies discussed here would be helpful for the related researches.

An Integrated Strategy for Functional Analysis in Large-scale Proteomic Research by Gene Ontology / 生物化学与生物物理进展

Data analysis poses a significant challenge to the large-scale proteomics studies. Based on the structured and controlled vocabularies-Gene Ontology (GO), and the GO annotation from related databases, a strategy composed of several programs and local databases is developed to identify the functional distribution and the significantly enriched functional categories of the proteomic expression profile. It would be helpful for understanding the overall functions of these identified proteins and supply the fundamental information for further bioinformatics exploration. This strategy has been successfully used in the Human Fetal Liver (HFL) proteomic research, which is available online at http://www.hupo.org.cn/GOfact/.

Basic Dose Response of Fluorescent Screen-based Potal Imaging Device / 대한방사선종양학회지

PURPOSE: The purpose of this study is to investigate fundamental aspects of the dose response of fluorescent screen-based electronic portal imaging devices (EPIDs). MATERIALS AND METHODS: We acquired scanned signal across portal planes as we varied the radiation that entered the EPID by changing the thickness and anatomy of the phantom as well as the air gap between the phantom and the EPID. In addition, we simulated the relative contribution of the scintillation light signal in the EPID system RESULTS: We have shown that the dose profile across portal planes is a function of the air gap and phantom thickness. We have also found that depending on the density change within the phantom geometry, errors associated with dose response based on the EPID scan can be as high as 7%. We also found that scintillation light scattering within the EPID system is an important source of error. CONCLUSION: This study revealed and demonstrated fundamental characteristics of dose response of EPID, as relative to that of ion chambers. This study showed that EPID based on fluorescent screen cannot be an accurate dosimetry system