How well does sequence similarity predict similarity in binding specificity?

• Title: "How well does sequence similarity predict similarity in binding specificity?"

• Paper to be examined: “A Specificity Map for the PDZ Domain Family”, PLoS Biol 6(9): e239 [1]

• Key claim of the paper: "Similarity in sequence of binding motifs of PDZ domains can be used to identify 16 distinct specificity classes."

• Data and Instructions

• • Download all the raw data needed for this module at: Media:Module3.zip
  • Try opening the LOLA software (see instruction in README.txt).
  • Make sure R is installed in your computer
  • Make sure you have Java

• Schedule:
  • H1: General intro to the paper/motivation
  • H2-3: Write code to import the data and compute the Position Weight Matrices (PWM).
  • H4-6: Write code to compute and compare different types of similarity values and do the clustering.
  • H7: Use the LOLA visualization software to look at the data.
  • H8: Start writing the report
  • H9: Broader view on the use and challenges with similarity metrics in biology, and other applications of the mathematical tools developed in this paper.

• Key bioinformatics concept of this module:
  • Similarity measures between biological objects (here protein sequence and binding specificity)
  • Clustering

• back to UNIL MSc course: "Case studies in bioinformatics 2017"