A variant on the CREST iMTD algorithm for noncovalent clusters of flexible molecules

Area of Science:

• Computational Chemistry
• Molecular Modeling
• Quantum Chemistry

Background:

• Conformational ensemble generation is vital for understanding molecular behavior.
• Identifying the global minimum conformation remains a significant challenge.
• Existing methods may struggle with flexible molecules and noncovalent clusters.

Purpose of the Study:

• To introduce a novel algorithm, LEDE-CREST, for enhanced conformer-rotamer ensemble sampling.
• To improve the generation of structural ensembles and energetics for noncovalent clusters.
• To provide a more efficient tool for computational chemistry studies.

Main Methods:

• Development of the low-energy diversity-enhanced variant on CREST (LEDE-CREST) algorithm.
• Utilizing the semiempirical GFN2-xTB approach for energy evaluations.
• Applying both CREST and LEDE-CREST to generate ensembles for various molecular clusters.

Main Results:

• LEDE-CREST demonstrates utility in generating diverse conformational ensembles.
• The algorithm effectively handles noncovalent clusters of both flexible and rigid monomers.
• Comparative analysis shows the strengths of the new LEDE-CREST approach.

Conclusions:

• LEDE-CREST offers an effective enhancement to existing ensemble generation techniques.
• The method is valuable for studying the structural and energetic properties of molecular clusters.
• This work contributes a powerful new tool for computational chemists.