Retrieval of crystallographically-derived molecular geometry information

Area of Science:

• Crystallography
• Computational Chemistry
• Structural Biology

Background:

• The Cambridge Structural Database (CSD) contains valuable crystallographically determined molecular geometry data.
• Conventional methods for accessing CSD data require significant user intervention, leading to underutilization.
• Existing access methods limit direct integration with client applications.

Purpose of the Study:

• To develop an automated program for retrieving molecular geometry data from the CSD.
• To overcome the limitations of conventional substructure searching for accessing CSD information.
• To make CSD geometrical data readily accessible for computational applications.

Main Methods:

• Development of the Mogul program for automated data retrieval.
• Implementation of a key system to encode chemical environments of molecular fragments (bonds, angles, torsions).
• Utilizing a search tree and similarity calculation for efficient matching of query fragments to stored geometrical distributions.

Main Results:

• Mogul successfully automates the retrieval of molecular geometry data from the CSD.
• The program efficiently groups chemically identical fragments and computes geometrical parameter distributions.
• Search results provide precise and unbiased estimates of molecular geometrical preferences.
• Validation experiments confirm the accuracy and utility of the retrieved data.

Conclusions:

• Mogul significantly enhances the accessibility and utility of CSD molecular geometry data.
• Automated retrieval facilitates validation of modeled and experimental molecular structures.
• The program aids in structure solution from low-resolution X-ray diffraction data.
• Mogul represents a significant advancement in accessing and utilizing crystallographic data.