Melodia

Melodia provides a Python library for describing, comparing and analysing the shape of protein structures using differential geometry of 3D curves and knot theory.

Key Features:

• Geometric Descriptor Generation: Melodia generates robust geometric descriptors for protein structures using differential geometry and knot theory.
• Shape Comparison: The descriptors enable advanced similarity analysis of protein structures that is more sensitive to structural feature variation than RMSD deviation.
• Sequence Structural Annotation: Melodia incorporates sequence structural annotation for analyzed proteins.
• 3D Visualization: The library includes 3D visualizations of protein structures.
• High-Throughput Processing: Melodia can generate geometric descriptors for thousands of shapes in just a few minutes.

Scientific Applications:

• Protein Engineering: Analysing protein structure similarities supports protein engineering workflows.
• Drug Discovery: Protein structure similarity analysis is an important step in drug discovery.

Methodology:

Melodia uses differential geometry of 3D curves and knot theory to generate geometric descriptors for protein structures.