Related Experiment Videos
Is there a code for protein-DNA recognition? Probab(ilistical)ly. . .
Panayiotis V Benos1, Alan S Lapedes, Gary D Stormo
1Department of Genetics, Washington University, School of Medicine, St. Louis, MO 63110, USA.
Summary
This study reviews models of transcription factor binding to DNA. It proposes a unifying "probabilistic code" for understanding protein-DNA recognition, supported by additive interaction assumptions.
Area of Science:
- Molecular Biology
- Genetics
- Bioinformatics
Background:
- Gene transcription is initiated by transcription factors binding to DNA promoter regions.
- Transcription factors utilize diverse mechanisms for recognizing specific DNA target sites.
- Previous decades saw attempts to define rules and models for these recognition mechanisms.
Purpose of the Study:
- To provide an overview of existing models for transcription factor-DNA recognition.
- To critically evaluate these models within a probabilistic framework.
- To propose a unifying concept for protein-DNA recognition.
Main Methods:
- Historical review of transcription factor-DNA recognition models, including the 'recognition code' concept.
- Development and application of a probabilistic framework for model comparison.
- Analysis of simplifying assumptions, particularly additivity of interactions.
Main Results:
- Existing models for transcription factor-DNA interaction are presented and discussed.
- A probabilistic framework allows for comparative analysis of model advantages and disadvantages.
- Additive interaction assumptions are shown to be sufficiently justified in many scenarios.
Conclusions:
- The study supports the validity of additive interaction assumptions in protein-DNA recognition.
- These assumptions can be extended to other situations, allowing for broader applicability.
- A unifying concept of a 'probabilistic code' for protein-DNA recognition is defined.