Delivering Bioinformatics Training: Bridging the Gaps Between Computer Science and Biomedicine

1
Delivering Bioinformatics Training Bridging the
Gaps Between Computer Science and Biomedicine

• Christopher Dubay Ph.D.
• James M. Brundege Ph.D.
• William Hersh M.D.
• Kent Spackman M.D., Ph.D.
• Division of Medical Informatics Outcomes
  Research, OHSU

2
Overview

• A View of Bioinformatics
• Background Significance
• An Integrative Information Science
• The Gap
• Between Computer Science Biomedicine
• OHSU Bioinformatics Education
• Questions Next Steps

3
What is bioinformatics?

• Two perspectives
• A set of tools techniques to support biological
  science
• Equivalent in scope to new assay methodology or
  new investigative techniques
• A science that supports the systematic
  development and analysis of such tools
• Investigation of the set of scientific principles
  forming the foundation for successful
  bioinformatics applications

4
Background of Students

• Computer information science
• Need to understand existing tools, scientific
  approach, and needs of biological research
• Biomedicine
• Need to learn a set of tools and skills
• May also need to understand the deeper scientific
  issues

5
The Gap I

• Biological scientists and investigators cant
  build their own tools
• Computer scientists dont know what tools to build

6
The Gap II

• Putting a biological investigator and a system
  implementer together in a room doesnt solve the
  problem
• Barriers include
• Language
• Methodology
• Conceptualization

7
The Gap III

• Computer science is a science of the artificial
• Mainly concerned with human artifacts i.e.
  creations limited mainly by conceptualization and
  imagination
• Biomedicine is a science of discovery
• Mainly concerned with how organisms function, the
  limiting factors are often a result of limits of
  investigative methods and tools

8
Bioinformaticsdefined in terms of tools

• General Tools
• WP, Spreadsheets, Robotics, Instrumentation
• Communications
• E-Mail, Networks, Internet World Wide Web
• Databases
• Storage, Organization
• Analysis Tools
• Examination Discovery
• Informatics Has Changed How Science is Done

9
Bioinformatics Significance
10
Bioinformatics defined in terms of a skill set

• Know-how with Practical Tools
• Cross Cultural Exchange
• Language of Biomedical Research
• Language of Informatics
• Solving Scientific Problems using Computers
• Database Interoperation
• Process Modeling Data Visualization
• Bioinformatics is an Information Science

11
Informatics SkillsSystem Design Implementation
Data Production Data Gathering
Problem Determination User Requirements
DESIGN
User Interface
System Configuration Maintenance
DEVELOP
Data Distribution
Usage Outcomes
Results Interpretation
EVALUATE
12
Where Should We Put the Emphasis?

• From the perspective of students entering the
  field and wondering about their future careers
• Will bioinformatics turn out to be mainly a means
  to an end (a tool set)? Or will it turn out to be
  a viable science in its own right?

13
Bioinformatics as scienceOntologies

• Ontology a formalization of a
  conceptualization
• Mathematics of ontologies description logics
• Getting useful discovery answers from large
  databases probably will depend on a concept model
  i.e. an ontology
• A potential goal for students Understanding how
  to build and use an ontology for bioinformatics
  applications

14
Bridging the Gap Education
15
A little history

• Post-doctoral fellowship program at OHSU began in
  1992
• Both MD and PhD post-docs
• One PhD (genetics) informatics post-doc stayed on
  as faculty responsible for bioinformatics
• NLM funded bioinformatics curriculum development
  as a supplement

16
Current Educational Activities

• OHSU Three Term Curriculum in Bioinformatics
• OHSU, PSU, OGI
• Distance Local
• Work with Biomedical Research Groups
• Research Information Systems Steering Committee
• Bioinformatics Sub-Committee

17
OHSU Curriculum

• Fall
• MINF 571 Computers in Bioscience
• MINF 572 Bioinformatics Laboratory
• Winter
• MINF 573 Topics in Bioinformatics
• Spring
• MINF 575 Bioinformatics Systems Development
• Every student does a project every term

18
MINF 571 Computers in Bioscience. Course
Objectives

• Survey Course in Bioinformatics
• Understand basic computing and networking
  concepts.
• Introduce basic concepts of molecular biology and
  genetics.
• Focus on biomolecular databases to retrieve and
  publish, genetic analysis, gene expression
  analysis, proteomics (structure / function),
  systems biology.

19
Course Description

• This course surveys the applications of
  informatics to biological problems, specifically
  those problems encountered in studies of genomes
  employing molecular biology and genetic
  techniques.
• The course follows a paradigm of how
  bioinformatics applications have been developed
  to aid in genome research in each step of
  biologic expression from the DNA template,
  through transcription, translation, protein
  structure and function, as well as in analyzing
  meiotic events, and genetic epidemiology.
• The course is designed for both users and
  developers of bioinformatics applications, and
  thus addresses both the algorithms underlying the
  applications and their implementation. To
  equilibrate the backgrounds of biologists and
  computer scientists introductory lectures are
  provided during the first two weeks of class.

20
MINF 572 Bioinformatics Laboratory. Course
Objectives

• Internet Navigation.
• Introduction to the UNIX Operating System.
• Learn to use the GCG Program Suite
• UNIX Interface
• SeqWeb Interface
• Use tools to visualize datasets (e.g. expression
  analysis) and biomolecules.

21
MINF 573 Topics in Bioinformatics. Course
Objectives

• Drill down into topics of choice from MINF 571
• Focus on Databases
• Topics are presented in terms of their historical
  development, current literature, and future
  directions
• Journal Club for bioinformatics
• Lectures from those using the tools

22
MINF 573 Topics in Bioinformatics. Course
Objectives

• Examples of topic areas include
• DNA micro-array technology
• bio-sequence analysis
• functional genomics
• data warehousing/data mining
• genetic linkage analysis
• Web and Internet based software development, etc.

23
MINF 575 Bioinformatics Systems Dev. Course
Objectives

• Learn bioinformatics software development
  best-practices and methodologies
• Emphasis on functionality prevalent in
  bioinformatics tools
• database interoperability, client/server and
  distributed computing designs, visual user
  interfaces, etc
• Paradigm for the course is that of a software
  development project

24
Course Participant Survey

• To gauge the educational audience
• Skills
• Interests
• Directions
• Tailor course emphasis to participants
• Create a database of Skills and Interests
• Useful for Course Projects

25
Course Participant Survey
26
Course Participant Survey
27
Course Participant Survey
28
Course Participant Survey
29
Bioinformatics Curriculum Elements Survey

• Four Schools
• OHSU (3 Courses)
• UCSC (3 Courses)
• UCLA (3 Courses)
• Stanford (4 Courses)
• Created Matrix of Topics
• Created a Taxonomy for Topics Applications

30
Bioinformatics Curriculum Elements

• Data Storage and Retrieval
• Information Retrieval
• Molecular Biology Genomics Sequence
  Analysis Linkage Analysis Gene
  Expression Proteomics

31
Bioinformatics Curriculum Elements

• Software Engineering Laboratory Information
  Mgmt Sys Data acquisition software Data
  analysis software Statistical
  software Databases Internet

32
Bioinformatics Curriculum Elements

• Algorithms Applications Gene
  identification Sequence Alignment Molecula
  r Models Techniques Dynamic
  programming Neural networks Hidden Markov
  Models Bayesian statistics

33
Bioinformatics Curriculum Elements

• Biological Models Molecular Models
  Structure-Function Biological
  Pathways Structural Models Anatomy V
  isible Human Project Human Brain
  Project Evolutionary Models Phylogeny

34
Bioinformatics Curriculum Elements

• Data Acquisition/Analysis Laboratory
  Information Mgmt Sys Automated data acquisition
  (high-throughput) DNA microarray Biostatis
  tics Signal Processing Data Visualization

35
Bioinformatics Curriculum Elements

• Biostatistics Data analysis Statistical
  models Stochastic models Hidden Markov
  Models Bayesian statistics Biometry
• Clinical Applications
• Ethics in Bioinformatics

36
Taxonomy for Bioinformatics
Academics
PrivateEnterprise
Setting
Research
Training
Systems Development
Activity
Focus
Tools
Methods
Theory
Formalization (Model)
Design
Implementation
Product
Example node AcademicResearchToolsImplementat
ion Required bioinformatics skill set
programming, databases, system analysis Example
application Implementing a data warehouse for
the storage of DNA microarray data used for gene
expression profiling of tumor subtypes.
37
To Bridge The Gap

• Give the broad picture of Bioinformatics to all
  disciplines
• Computer Scientists
• Live in the Lab
• Follow Biology Literature
• Biologists
• Learn Software Development Principles
• Exposure to new information technologies
• Do actual work Course Projects

38
Next Steps

• Deploy Database of Bioinformatics Projects and
  Interests to
• Link projects with students
• Continued Development and Documentation of
  Bioinformatics Educational Elements Paths
• Expanding Audience for Bioinformatics Education
• Relevance Modules

39
Questions?

• Syllabi for Courses
• http//medir.ohsu.edu/bioinf/