This course is being offered in Fall 2005:
Lectures: Tuesdays 20:30 to 21:00 in H-400
Office Hours: Fridays 15:30 to 17:30 in EV003.219
It assumes that students have a general knowledge of bioinformatics in terms of how it assists life sciences research. Any of the 600-level courses in bioinformatics will provide this background: COMP 691R Bioinformatics Algorithms; COMP 691S Bioinformatics Databases and Systems; COMP 691E Bioinformatics Applications of Machine Learning;
It assumes that students have a comprehensive knowledge of undergraduate computer science including software, databases, and artificial intelligence.
However, any deficiencies in background can be remedied through
additional reading and class discussion.
So talk to the instructor Dr Butler
The principal objectives of the course are to keep students abreast of developments in bioinformatics for the acquisition, representation, management and reasoning with biological knowledge. Example systems and applications will be studied. Background material from artificial intelligence, knowledge-based systems, and genomics will be introduced; with in-depth class discussion of technology and its application to bioinformatics.
Bioinformatics is a relatively new discipline dealing with the computational needs of genomics. Biology has become a data-intensive activity. Genomics databases must deal with this variety and scale, as well they must integrate disparate databases that are their information sources; must provide flexible, friendly user interfaces for querying and data mining; and cope with incomplete and uncertain data.
Information technology must also address the workflow within the laboratory, including the automated analysis of data. Some analysis techniques require significant computational resources, and the management of large-scale distributed computation is an issue.
Knowledge management aims to turn data and information into knowledge that can be applied to the problems facing the scientist.
See Internal Pages.
InforSense presentations
see
(1) Open Discovery Workflows Integrating bioinformatics and cheminformatics
(2) Transforming Drug Discovery: Text Mining Use Cases.
J. Van Bemmel, Data, Information and Knowledge bridging bio-, neuro- and medical informatics, at Synergy between Research in Medical Informatics, Bio-Informatics and Neuro-Informatics , 4 December 2001, Pyramids, Place Rogier, Brussels.
Knowledge: management, representation, engineering, reasoning.
Bioinformatics knowledge: exsiting conceptual models and ontologies.
Case studies.
Selected topics: ontologies and the semantic web, multi-agent systems, data integration, knowledge integration using bayesian networks, expert systems, natural language processing, data mining and text mining.
Students are required to write two essays, each worth 20%. Students are required to perform a mini-project and to make a class presentation (20%) and write a term paper (40%) on it.
Each student will select a case study for the mini-project: that is, they will select an article(s) describing a life sciences study. The mini-project is to analyse the knowledge used in the study and design (on paper) a knowledge-based system to support similar studies.
The mini-project will involve selecting knowledge representations (KRs) suitable to the study and writing down some knowledge of the study in those KRs. It will also decide what reasoning is required, and document examples of the reasoning.
Possible KRs will include ontologies, first-order logic, production rules, state machines, and process algebras depending on the study.
Assignment 1 (20%) Due week 3 at 20:30 Tuesday September 20th, email pdf to gregb@cs
Write a 10-page essay explaining the difference between data, information,
and knowledge.
Use two examples, one from software development and one from life sciences,
to illustrate your explanation by giving concrete instances of data, information,
and knowledge from each of the two examples.
You will be evaluated on the clarity of your explanation of the differences.
Assignment 2 (20%) Due week 6 at 18:00 on Friday October 14th, email pdf to gregb@cs
You must attend the
Workshop on Knowledge-Based
Bioinformatics and select one of the talks for deeper study.
Write a 20-30 page paper clearly and fully explaining the topic of the talk
and the state-of-the-art in that area of knowledge-based bioinformatics.
You will be evaluated by the depth of your survey of the topic, your critical
assessment of the state-of-the-art, and the clarity of your writing.
Class Presentation (20%) In class during weeks 10 to 13.
Present the current state of the analysis and design for your mini-project.
Each student will do a 30-40 minute presentation to the class.
Each presentation will be followed by discussion.
You will be evaluated based on your presentation (clarity, information content,
focus)
and you ability in responding to questions during the discussion.
Term Paper (40%) Due December 13, 2005.
Write a 30-page report on the mini-project.
It will cover both the analysis and the design.
It will emphasise the analysis of the knowledge used in the study
and the selection of the knowledge representations for the knowledge.
You be evaluated on the clarity of your analysis and design and how it clearly
shows the role and impact of the knowledge when using the chosen knowledge
representations.
Multi-Sources Information Fusion
Blackboard Pattern
Blackboard - Agent Architectures
Applications of Blackboard Architectures
An Introduction to Bayesian and Dempster-Schafer Data Fusion
GERTIS: a Dempster-Shafer approach to diagnosing hierarchical hypotheses. An expert system applying D-S theory.
Modelling intracellular signalling networks using behaviour-based systems and the blackboard architecture. Early work still.