Related Experiment Videos
Atomic contact vectors in protein-protein recognition
Julian Mintseris1, Zhiping Weng
1Bioinformatics Program, Boston University, Boston, Massachusetts 02215, USA.
Proteins
|October 28, 2003
Summary
We developed atomic contact vectors (ACVs) to analyze protein-protein interfaces structurally. This method effectively distinguishes complex types and identifies unique recognition complexes, advancing molecular recognition studies.
Area of Science:
- Structural biology
- Biochemistry
- Bioinformatics
Background:
- Understanding protein-protein interactions (PPIs) structurally is key to molecular recognition and biochemical network function.
- Comparing PPI interfaces requires robust representation and analysis methods.
Purpose of the Study:
- Introduce atomic contact vectors (ACVs) for representing and comparing protein-protein interfaces.
- Utilize ACVs to classify interfaces and mine protein data bank (PDB) for complex types.
- Establish a non-redundant dataset of protein recognition complexes.
Main Methods:
- Developed atomic contact vectors (ACVs) to capture physico-chemical interface characteristics.
- Applied ACVs for classification tasks, distinguishing homodimers from crystal contacts.
- Mined the PDB to differentiate transient recognition complexes from permanent oligomeric ones.
- Derived a rule for assessing interface redundancy using ACVs.
Main Results:
- ACVs demonstrated effectiveness in classifying protein-protein interfaces.
- Achieved a 91% success rate in distinguishing transient recognition from permanent oligomeric complexes.
- Identified distinct contact preferences beyond accessible surface area.
- Generated the largest reported non-redundant set of 209 recognition complexes.
Conclusions:
- ACVs provide an intuitive and powerful measure for comparing protein-protein interfaces.
- The ACV method surpasses sequence-based approaches for interface analysis.
- This work advances the understanding of molecular recognition and protein complex classification.