FakeRotLib: Expedient Noncanonical Amino Acid Parametrization in Rosetta

Area of Science:

• Biochemistry
• Computational Biology
• Structural Biology

Background:

• Noncanonical amino acids (NCAAs) are crucial in natural biology and synthetic applications.
• Current deep learning methods face challenges in modeling NCAAs due to sparse training data.
• Biophysical methods like Rosetta are effective for modeling NCAAs.

Purpose of the Study:

• To explore the parametrization of noncanonical amino acids (NCAAs) for Rosetta.
• To identify key factors influencing NCAA parametrization in Rosetta.
• To introduce a novel method for generating rotamer distributions for NCAAs.

Main Methods:

• Discussed various aspects of parametrizing noncanonical amino acids (NCAAs) for Rosetta.
• Identified rotamer distribution modeling as a critical factor for NCAA parametrization.
• Developed FakeRotLib, a method using statistical fitting of small-molecule conformers to generate rotamer distributions.

Main Results:

• FakeRotLib significantly improves the parametrization of noncanonical amino acids (NCAAs) in Rosetta.
• The method outperforms existing approaches in terms of speed and efficiency.
• FakeRotLib successfully parametrizes NCAA types previously unmodeled by Rosetta.

Conclusions:

• Rotamer distribution modeling is a key determinant of success in parametrizing noncanonical amino acids (NCAAs) for Rosetta.
• FakeRotLib offers a faster and more effective solution for generating NCAA rotamer libraries.
• This advancement expands the capabilities of Rosetta for modeling diverse amino acid structures.