A computational approach to identify efficient RNA cleaving 10-23 DNAzymes.

Area of Science:

• Biochemistry
• Molecular Biology
• Computational Biology

Background:

• DNAzymes are catalytic DNA molecules with therapeutic potential due to high specificity and low toxicity.
• Current methods for identifying efficient DNAzymes are time-consuming and costly, hindering development.
• Lack of specific tools limits the discovery of potent DNAzyme therapeutics.

Purpose of the Study:

• To develop a computational methodology for predicting the efficiency of 10-23 DNAzymes targeting specific RNA molecules.
• To create a tool that can triage large libraries of potential DNAzymes, identifying promising candidates for further investigation.
• To accelerate the discovery and development of novel DNAzyme-based therapeutics.

Main Methods:

• A logistic regression model was developed to predict DNAzyme efficiency.
• The model was initially trained on published DNAzyme activity data.
• The methodology was validated and refined through laboratory testing of newly synthesized DNAzymes.

Main Results:

• The computational method successfully predicts DNAzyme efficiency, enabling rapid screening of potential candidates.
• Binding free energy between the DNAzyme and target RNA is a key factor in efficiency.
• Internal DNAzyme structure also significantly influences catalytic activity.

Conclusions:

• The proposed computational methodology offers an efficient approach to identify effective DNAzymes for therapeutic applications.
• This tool can significantly reduce the time and cost associated with DNAzyme drug discovery.
• A publicly available program implements this predictive method, facilitating broader research and development.