Bioinformatics Related Research Brad Hemminger bmh@ils. unc. edu School of Information and Library Science University of North Carolina at Chapel Hill
SILS and CCGS • The School of Information and Library Science (SILS) is at the periphery of CCGS. It is essentially different from basic science disciplines (genetics, molecular biology, biochemistry, etc). We’re also different from Computer Science or Bio. Statistics in that we’re not actively developing programs or algorithms for computational biology. • The focus of SILS is investigating and designing systems to better integrate people, information and technology. Our closest CCGS cousins in this sense are ethics, philosophy, social medicine and public health. But we’re also different from them in that we’re interested not just in the informatics issues, but the technology that goes with them.
How does SILS fit? • We’re similar and we’re different. This diversity within CCGS is a good thing! • CCGS has supported this diversity, helped SILS formulate a vision for how to best integrate into the CCGS, and then supported these efforts. The result was the formation last summer of the CCGS Biomedical Informatics Research, Service, and Training Program (BIRSTP)
SILS Research Interactions • Challenges – Too much information (explosion of journal articles) – Many kinds of information (articles, dataset, lab result, statistical analysis, image, video) – Collaborations between many scholars, and across disciplines (most big science requires interdisciplinary research) – Transformative change—digital representation of information • Our research interactions with CCGS scientists tends to fall into two categories – Studying CCGS scientists information use to develop better information tools and infrastructures – Assisting CCGS scientists through the development of tools and resources that help them do their research better or more efficiently
My Primary Research Areas • Information Representation – What are the best representations for exchanging/sharing information (both people and machines)? – What novel representations allow us to more effectively solve scholarly communications problems? • Interactive Information Visualization – Highly interactive visualizations allow us to better comprehend complicated information – Novel visualizations can solve difficult problems • Example: Seethru for surgery planning
Research Roadmap • Information Representations – Understanding Information Seeking Behavior/Use • • – Building a Common Data Model for Bioinformatics Research Labs • • – (ISB study) KT Vaughan, other librarians DETAIL Individual Researcher Interview Studies P 20 Grant, Interviewing labs Kirk, other collaborators from RENCI slides DETAIL Small scale; develop standard representations for new research efforts (Kevin Weeks in Chemistry) Development of shared electronic scholarly communications model (Neo. Ref) • • All digital content deposited in open repositories (papers, databases, etc) Representation of content as objects and relationships (facets), i. e. not standard (hierarchical taxonomies) Studying annotations, work by John Macmullen Studying use of collaborative tools, in particular wikis (conferences, research projects, classes) – Novel Information Representations – Infrastructure developed to facilitate research • • • Ultrastructure representing bioinformatics, with Morgan Giddings, Chris Maier, Jeff Long SILS masters papers digital library, testing automatic submission, metadata extraction UNC campus ETDs SCC (Copyright agreement addendums, Journal author rights tool). Institutional Repository, Biomedical Data Repository Interactive Visualizations – Searching in Billions and Billions of content items (or how I learned to love Google) (Sullivan, Vision) • • – User interfaces for exploring large spatial spaces • • – Evaluation of Neo. Ref article review interface tool Comparison of searching articles via full-text vs metadata DETAIL Visualizing the body for surgery/therapy interventions (See. Thru) Searching/Navigating in large images (Pan. Zoom) Integrating Multiple Resources • Bioinformatics Information Resources (TAMAL, Sullivan) – Sharing Information via dynamic interfaces – Structured visualizations for dynamic exploration and sense making • • • Personal Health Records (PHR) Reviewing and making sense of survey comments (ICIS tool for ISB study, KT Vaughan, HSL librarians) Journal author rights tool for helping faculty make “good” choices in journals to publish in.
Information Representation (1/3) • Understanding Information Seeking Behavior/Use – UNC ISB study, results – National ISB Study – Individual Researcher Interview Studies • Building a Common Data Model for Bioinformatics Research Labs – Survey UNC labs, and develop generalized common data model, that can be applied at UNC. Involves many labs at UNC, with significant input from Kirk Wilhelmsen and others of CCGS. – Small scale: develop standard representations for new research paradigms. An example is our formulation of an XML format for h. SHAPE, for Kevin Weeks in Chemistry, of their lab’s new quantitative RNA structure analysis output. This standard data structure allows experimental data to be exchanged as well as validated using universally available tools (XML and browsers).
Information Representation (2/3) • Development of shared electronic scholarly communications model (Neo. Ref) – All digital content will be deposited in open repositories (papers, databases, etc) to address serials crisis and build on success of pubmed (shared, open, integrated databases) – Representation of content will be as objects and relationships (facets), i. e. not as previously done by libraries when materials were all physical (standard hierarchical taxonomies) because that is too limiting. Reviews, comments, annotations are full fledged content items, all searchable in full text. (Connotea, Faculty of 1000) – Example project is work by Phd Student John Mac. Mullen, studying contextual analysis of variation and quality in human curated Gene Ontology annotations. He is collaborating with two curation efforts, the Saccharomyces Genome Database (model organism database for yeast and the Gene Ontology Consortium Meeting and Annotation Camp, and studying their curation records as well as observing the curation process first hand – Studying use of collaborative tools, in particular wikis (conferences including ASIST 2005, ASIST 2006, JCDL 2005, research projects, classes)
Information Representation (3/3) • Novel Information Representations – Ultrastructure representation for biological knowledge, with Morgan Giddings, Chris Maier, Jeff Long • Infrastructure developed to facilitate research – SILS masters papers digital library, testing automatic submission, metadata extraction – UNC campus ETDs – SCC (Copyright agreement addendum, Journal Author Rights Tool, web portal for campus, outreach/education program for the faculty). – Institutional Repository – Biomedical Data Repository (to come from CTSA & CCEGA work)
Interactive Information Visualization • Searching in Billions and Billions of content items (or how I learned to love Google) (Sullivan, Vision) – Evaluation of Neo. Ref article review interface tool – Comparison of searching articles via full-text vs metadata DETAIL • User interfaces for exploring large spatial spaces – Visualizing the body for surgery/therapy interventions (See. Thru) – Searching/Navigating in large images (Pan. Zoom) • Integrating Multiple Resources – Bioinformatics Information Resources (TAMAL, Sullivan) • Sharing Information via dynamic interfaces – Personal Health Records (PHR) • Structured visualizations for dynamic exploration and sense making – Reviewing and making sense of survey comments (ICIS tool for ISB study, KT Vaughan, HSL librarians) – Journal author rights tool for helping faculty make “good” choices in journals to publish in.
Beyond Research: Professional School Perspective As a professional school, SILS is particularly concerned with providing excellent training and educational experiences to students, as well as providing service to our community.
Programs Supporting Engagement • SILS MS Bioinformatics Certificate Program • Campus Ph. D BCB Bioinformatics Certificate Program (to be degree granting program) • EPA training Program (CEBRC, Fred Wright PI) • CCGS-BIRSTP Training Program (KT Vaughn, Barrie Hayes, Nani Vaidhyanathan)—we’re here to help. • Integrated Biomedical Informatics Support on campus for the first time through CTSA (Paul Watkins PI)
The End
–Virseum—not bioinformatics, but truly novel
Скачать презентацию Bioinformatics Related Research Brad Hemminger bmh ils unc edu
1c1ba085eaf290becb334dadd623815d.ppt