He Group  
Home|People|Projects|Publications|FAQs|Links|News|Contact Us

He Group Projects

This web page introduces the research projects conducted in the Hegroup:

(1) Development and applications of community-based interoperable ontologies

Ontologies have emerged to become critical to biomedical data and knowledge integration, sharing and validation, as well as new knowledge discovery. A major field of AI is knowledge representation and reasoning (KR², KR&R), for which ontology is a part of the story and critical.

We have initiated and led the development of many community-based ontologies such as the Cell Line Ontology (CLO), Ontology of Biological and Clinical Statistics (OBCS), and Ontology of Host-Microbiome Interactions (OHMI). Dr. He is also an active developer of the Ontology for Biomedical Investigations (OBI), which is collaboratively developed by over 20 communities, and Dr. He is the representation of the vaccine community. To support the Kidney Precision Medicine Project, Dr. He is leading the development of the Kidney Tissue Atlas Ontology (KTAO). To face the challenge of the pandemic COVID-19, Dr. He has initiated the development of the community-driven Coronavirus Infectious Disease Ontology (CIDO).

These ontologies can be used in many applications, including standardized data and knowledge representation, sharing, integration, and advanced analysis. We can also apply ontology to different informatics methods such as literature mining and machine learning.

Representative publications:

The following is a list of ontologies that we have been leading in development. Note that these ontology development all follows the OBO Foundry principles and uses BFO as the upper level ontology:

AEO Ontology of Adverse Events (OAE): A community-based ontology targeting to represent adverse events for drugs, vaccines, and other biomedical reagents. Note: OAE was previously named AEO. (Reference: PMID: 25093068).
CLO Cell Line Ontology (CLO): A community-based cell line ontology. The CLO development originated from the research done by Sirarat (Sira) Sarntivijai before she joined He Lab for PhD thesis research. Sira got her PhD with ontology-focused research in 2012. Sira is still actively working on CLO update with international collaborations. (Reference: http://www.jbiomedsem.com/content/5/1/37/).
OGG Ontology of Genes and Genomes (OGG): A community-driven ontology of genes and genomes. (Reference: ICBO 2014 full length proceeding paper).
INO Interaction Network Ontology (INO): an ontology in the domain of interaction networks. INO develoment follows the OBO Foundry principles. INO can be used for automated reasoning such as literature mining and Bayesian network reasoning. (References: JBMS INO paper; arXiv:1311.3355).
Vaccine Ontology (VO): The VO project is a community-based project with an aim to develop the VO to ensure vaccine data standardization, exchange, and automated reasoning. (Reference: Nature Procedings 2009; PMID: 23256535; PMID: 21624163; PMID: 24209834) .

Other ontologies not listed above:

Note: Oliver prepared two tutorial websites. Although not routinely updated, they may still be helpful:

(2) Development of Ontoanimal tools to support interoperable ontology development & applications

We have developed many widely used ontology tools (e.g., Ontobee, Ontofox, and Ontorat), collectively names “Ontoanimal” tools. Each tool has specific functions; together, these tools are able to extract ontology subsets, provide ontology community views, generate and edit ontology terms, query and visualize ontology terms, provide statistics of ontologies, and compare ontologies.

Representative publications:

Below is a quick summary of these Ontoanimal tools.

Ontobee: A Linked Data Server that publishes RDF and HTML data simultaneously. The Ontobee web server aimed to facilitate ontology visualization, query, and development. OntoBee provides a user-friendly web interface for displaying the details and its hierarchy of a specific ontology term. Meanwhile, OntoBee provides a RDF source code for the particular web page, which supports remote query of the ontology term and the Semantic Web. (Reference: Xiang et al., 2011).
OntoFox: An Ontology tool that Fetches ontology terms and axioms. OntoFox is a web-based system to support ontology reuse. OntoFox facilitates ontology development by automatically fetching ontology terms and their annotations from existing ontologies and saving the results in an importable RDF/OWL format. (Reference: PMID: 20569493).
Ontodog: A web-based ontology community view generator. (Reference: PMID: 24413522).
Ontorat: An ontology tool that supports Ontology representation of axioms using templates and design patterns. Ontorat is a web server that automatically generates new ontology terms and axioms based on user-provided input file. (Reference: Xiang et al, 2012).

It is noted that most of our ontology tools use our Hegroup Resource Description Framework (RDF) triple store. The SPARQL endpoint of our triple store is: http://sparql.hegroup.org/sparql/.

We have also initiated the proposal of the eXtensible ontology development (XOD) principles and methods to support ontology interoperability. The above mentioned Ontoanimal tools can be used to support the implementation of the XOD principles:

With the XOD and our Ontoanimal tools, ontology development is no longer boring and has become more fun and useful!

(3) Vaccine informatics, host-pathogen interactions, and vaccine/drug safety

We study Vaccine Informatics, a branch of vaccinology that Dr. He and his colleagues pioneered. We have developed VIOLIN, the most comprehensive vaccine database and analysis system. The Vaccine Ontology (VO) has become a community effort for standard representation of vaccines, vaccine components, vaccination, and host responses to vaccines. We have also developed Vaxign, the first web-based publically available vaccine target design tool based on bioinformatics analysis of genome sequences using the strategy of reverse vaccinology.

We study host-pathogen interactions. We developed PHIDIAS, the most comprehesenive host-pathogen interaction database focusing on human and animal pathogens. As part of PHIDIAS, we have developed Victors, a comprehensive knowledge base of more than 3,000 virulence factors in >120 human and animal pathogens. We have also developed an Ontology of Host-Pathogen Interactions (OHPI) to logically represent the virulence factors and their interactions with the hosts.

We are interested in vaccine and drug safety study. We have led the development of the community-based Ontology of Adverse Events (OAE), and applied it to study vaccine and drug adverse events.

Vaccine Informatics. We have developed the VIOLIN vaccine database and analysis system. As a part of VIOLIN, we have developed Vaxign, the first web-based publically available vaccine target design tool based on bioinformatics analysis of genome sequences using the strategy of reverse vaccinology. logy. Comparative genomics. Various genome sequences can be compared to address many biomedical questions. Analysis of vaccine and drug adverse event mechanisms. We are interested in applying the Ontology of Adverse Events to study the mechanisms of vaccine and drug-induced adverse events.

Representative publications:

Related tools developed in He group:

VIOLIN: Vaccine Investigation and Online Information Network: A web-based database system for research and development of vaccines against various human pathogens (e.g., HIV, influenza, and tuberculosis) with high priority in public health and national biological defense.
PHIDIAS: Pathogen-Host Interaction Data Integration and Analysis System: A web-based data integration and analysis system for biomedical researchers to investigate genomic sequences, curated literature information, and gene expression data related to pathogen-host interactions for those pathogens with high priority in public health.

Other tools not listed above:

  • Vaxign, the first web-based reverse vaccinology tool for vaccine design
  • Victors, a comprehensive knowledge base of virulence factors in human and animal pathogens.
  • BBP (Brucella Bioinformatics Portal): A gateway for Brucella researchers to search, analyze, and curate Brucella genome data originated from public databases and literature.

(4) Systematic ontology and bioinformatics analysis of biological interaction networks

We hypothesized that ontology supports literature mining and analysis of biological interaction networks. As a result, we developed the Interaction Network Ontology (INO) and applied it to enhance literature mining performance. We are also collaborating with our collaborators, including Drs. Junguk Hur and Arzugan Özgür, to apply ontology for more advanced literature mining.

Bayesian network (BN) can model linear, nonlinear, combinatorial, and stochastic relationships among variables across multiple levels of biological organizations. We have developed new BN algorithms and tools for analysis of gene interaction networks using high throughput gene expression data.

Representative publications:

Related tools developed in He group:

miniTUBA: Medical Inference by Network Integration of Temporal Data Using Bayesian Analysis: A web-based dynamic Bayesian analysis engine for clinical and biomedical researchers to perform medical inference and predictions with temporal datasets.
GenoMesh: A genome-wide analysis of gene-to-gene relationships and pathways based on the association between individual genes and MeSH terms obtained from literature. Currently GenoMesh includes data on Escherichia coli and Brucella spp..

Other tools not listed above:

  • Ignet (Integrated gene network), a centrality- and ontology-based liteature discovery system for analyzing and visualizing biological gene interaction networks using all PubMed literature papers.
  • Dignet, a program inside Ignet that is able to dynamically search PubMed papers.
  • CRCView: A web-based microarray data analysis and visualization system, powered by a model-based clustering algorithm Chinese Restaurant Cluster (CRC) developed by Dr. Steve Qin. CRCView also incorporates several Bioconductor microarray analysis programs including GOStats, genefilter, and Heatplus.

In a 2014 paper (He, 2014) and then a 2016 paper (He, 2016), Dr. He proposed a new OneNet Theory of Life”. The OneNet theory states that the whole process of a life of an organism is a single complex and dynamic network (called “OneNet”). Dr. He also proposes OneNet tenets to characterize different aspects of the OneNet life. Ontologies and ontology-based bioinformatics tools, such as those introduced above, can be used to integratively represent and study such OneNet theory and OneNet knowledge of different organisms. Studying the theory has become Dr. He's hobby.

(5) Wet-lab microbiology and immunology research (not active now):

Our primary biomedical research interest is to understand pathogenesis and immunology of infectious diseases and develop vaccines against intracellular pathogens. In particular, we study Brucella, a facultative intracellular bacterium that causes zoonotic brucellosis in humans as well as domestic and wild life animals. Our research focuses on the role of Caspase-2, an enzyme that is critical in regulating cell death, DNA damage, stress, cancer, and microbial infections. Our research aims to continuously elucidate the caspase-2-mediated proinflammatory cell death pathway and its biological effect on microbial pathogenesis and protective immunity against brucellosis and other diseases.

Our wet-lab research is focused on the study of Brucella, a facultative intracellular bacterium that causes brucellosis, one of the most common zoonotic diseases in the world in humans and a variety of animal species. Our wet-lab Brucella research focuses on the analysis of caspase-2-mediated cell death and its role in Brucella pathogenesis and immunity. Our studies first identified a new caspase-2-mediated proinflammatory cell death, which exists in macrophages and dendritic cells infected with live attenuated rough Brucella strains (e.g., B. abortus cattle vaccine RB51 and B. suis vaccine candidate VTRS1) (see references: Bronner, O'Riordan, and He, 2013; Li & He, 2012; Chen et al, 2011; Chen & He, 2009). This type of cell death is different from non-proinflammatory apoptosis or caspase-1-mediated proinflammatory pyroptosis, and so we named the cell death "caspase-2-mediated pyroptosis". Interestingly, virulent Brucella inhibits such cell death in infected macrophages (He et al, 2006) but not in dendritic cells (Li & He, 2012). Based on the study of the RB51-macrophage interaction model, endoplasmic reticulum stress activates the inflammasome via NLRP3- and caspase-2-driven mitochondrial damage (Bronner et al, 2015).

Representative publications:

We further hypothesize that caspase-2-mediated immunity is important not only to Brucella, but also to other pathogens. We had experimentally tested how caspase-2 mediates the immune responses of macrophages infected by tuberculosis vaccine BCG, and generated some good prototype results. However, we have not continued this line of research. Instead, we have focused our research on computational medicine and bioinformatics.  


Your suggestions, comments, and collaborations are welcome and appreciated.

Ontobee|Ontofox|OAE|VO|Genomesh|Ignet|miniTUBA|PHIDIAS|VIOLIN
He Group
University of Michigan Medical School
Ann Arbor, MI 48109
U of M NIAIDNIAID