Optimal codon randomization via mathematical programming.

Area of Science:

• Molecular Biology
• Bioinformatics
• Computational Biology

Background:

• Protein library generation is crucial for protein engineering and drug discovery.
• Degenerate oligonucleotides are commonly used for codon randomization, but achieving specific amino acid probabilities is challenging.
• Current methods often rely on heuristics, lacking a systematic approach to minimize oligonucleotide mixtures.

Purpose of the Study:

• To develop a computational method for determining the minimal oligonucleotide mixture required for targeted amino acid probabilities in protein libraries.
• To address the limitations of existing heuristic approaches in codon randomization.

Main Methods:

• Utilized integer programming to model the problem of oligonucleotide mixture computation.
• Developed a method to calculate the minimal mixture for arbitrary target amino acid probabilities.
• Considered both conventional and spiked oligonucleotide randomization strategies.

Main Results:

• The proposed integer programming methodology provides an exact solution for computing minimal oligonucleotide mixtures.
• The method allows for precise control over amino acid probabilities, accommodating both equal and arbitrary nucleotide distributions.
• Demonstrated a systematic approach to optimize oligonucleotide mixtures, outperforming heuristic methods.

Conclusions:

• Integer programming offers a robust and precise solution for designing degenerate oligonucleotides in protein library construction.
• This approach enhances the efficiency and accuracy of generating protein libraries with desired amino acid compositions.
• The developed methodology has significant implications for protein engineering and synthetic biology applications.