Software Engineering Challenges in Bioinformatics

1
Software Engineering Challenges in Bioinformatics

• Jonathan Barker and Janet Thornton
• 26th ICSE (2004)
• pp. 12-15
• Poster by Kai Chi

2
Summary

• Bioinformatics a cross-disciplinary field
  involves computational techniques to gather and
  analyze biological data.
• Challenges Automated data processing in terms of
  integration and presentation to be addressed with
  Software Engineering.

3
Background Central DogmaFrom Genes to Function
4
Background Central Dogma (contd)

• DNA 4 kinds of nucleotides, form base pairs (A,
  T, C, G)
• Amino Acid 20 kinds, building blocks of proteins
• Proteins 3-D structure, varies with heat, pH,
  M
• Why Protein Matters?
• How we walk Myocin Actin
• How we dine Pepsin (Enzyme)
• How we grow Somatropin (HGH)
• How we feel Epinephrine (NT, H)

5
Core Activities of Bioinformatics

• Activities with Biological Data
• Storage
• Organization
• Retrieval
• Analysis
• What kind of Biological Data?
• Linear sequences of genes are not enough
• Need 3D structure of protein to understand their
  function
• Presently, not possible to determine the
  structure based on the sequence

6
SE Challenge Data Integration

• Domain Specific
• Genetics
• Evolutional Biology
• Pharmacy
• Large Volume of New and Legacy Data
• 4.7 Mbp in E. Coli
• 6 Bbp in Human
• Current Solutions
• Perl, Python Scripts
• SOAP and WS
• Open Source

7
SE Challenge Data Presentation
8
Critique Strength and Weakness

• Strength Clearly described some of the data
  requirements of bioinformatics.
• Weakness Failed to provide specifics about what
  needs to be improved with current solutions.

9
Questions

• What are the 4 core activities of bioinformatics
  with biological data?
• What are the 2 main software engineering
  challenges in bioinformatics described in this
  paper?
• Are the software engineering challenges mentioned
  in the paper unique to bioinformatics? If not,
  give an example.