Opportunities and Limitations of Using Melting Curve Dissimilarity to Determine Protein Interactions

• Author / Creator

  Teitz, Joshua

• Proteins are large biomolecules that perform many functions within an organism. Proteins frequently form interactions with other biomolecules. A group of proteins held together by interactions is known as a protein complex.
  Protein interactions are essential to many biological processes. Thus, determining interactions is a key objective in biology. Since experimental techniques that can directly infer interactions are currently expensive and time consuming, there is considerable interest in computational techniques that can determine protein interactions from biological data.
  When a protein mixture is heated, the proportion of insoluble protein molecules forming precipitant increases for each kind of protein. A protein's melting curve records the fraction of the protein that is soluble over a series of increasing temperatures, relative to the amount of the protein that is soluble at the initial temperature. Thermal Proximity Co-Aggregation (TPCA) is the observation that interacting proteins tend to have similar melting curves. Although a previous work has provided empirical evidence of TPCA, no work has applied computational tools to melting curve datasets in an attempt to determine protein interactions. In this thesis, we apply computational tools to melting curve datasets, and describe the opportunities and limitations of
  such tools for determining interactions.
  Clustering is the task of finding groups of related objects in a dataset. We first explore whether clusters found in melting curve datasets correspond to protein complexes. We then turn our attention to pull-down assays, a commonly used biochemical technique that generates a list of potential interaction partners for a protein-of-interest. To determine interactions from the results of a pull-down assay, subsequent validation experiments must be performed, each involving the protein-of-interest and a potential interaction partner. Since
  validation experiments are expensive and time-consuming, we explore whether performing validation experiments in order of increasing melting curve dissimilarity from the protein-of-interest can limit the number of experiments necessary to find interactions. Lastly, we describe how melting curve dissimilarity can be used to detect proteins that differentially interact in two conditions.

• Subjects / Keywords

  • melting
  • curve
  • dissimilarity
  • protein
  • interaction
  • complex
  • clustering
  • information
  • retrieval
  • metric
  • learning

• Graduation date

  Fall 2022

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/r3-hp8g-7g28

• License

  This thesis is made available by the University of Alberta Library with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.