Inferring binding energies from selected binding sites

Area of Science:

• Biophysics
• Computational Biology
• Genomics

Background:

• Transcription factors (TFs) regulate gene expression by binding to specific DNA sequences.
• Accurate modeling of TF-DNA interactions is crucial for understanding gene regulation.
• Existing methods for TF binding site prediction often struggle with complex binding landscapes.

Purpose of the Study:

• To develop a novel biophysical model for TF-DNA binding that accounts for non-linear relationships.
• To improve the accuracy of estimating TF binding parameters.
• To enhance the prediction of TF binding site distributions.

Main Methods:

• Employed a biophysical model incorporating TF chemical potential, non-specific DNA binding affinity, and sequence-specific parameters.
• Utilized maximum likelihood estimation for parameter fitting.
• Introduced a high-throughput SELEX (HT-SELEX) procedure coupled with next-generation sequencing.

Main Results:

• The proposed model yielded significantly more accurate parameter estimates compared to standard probabilistic methods on simulated data.
• The model demonstrated superior prediction of selected binding site distributions.
• HT-SELEX enabled accurate binding parameter estimation after a single selection round, outperforming standard motif identification algorithms.

Conclusions:

• The developed biophysical model offers a more accurate and robust approach to characterizing TF-DNA interactions.
• The HT-SELEX procedure provides an efficient and powerful tool for determining TF binding specificity.
• This work advances the understanding and prediction of gene regulation by transcription factors.