Protein Surface Geometrical Features for Tubulin Classification

A research paper presents a method for classifying protein surfaces using geometrical descriptors adapted from 3D face recognition. The study, authored by Luca Di Grazia from Politecnico di Torino, tested the method on tubulin isotypes and compared tubulin with its bacterial analogue FtsZ. Different classification methodologies were used, with the best result obtained using a Support Vector Machine with a radial basis function kernel.

Use Cases

• Classifying tubulin protein isotypes based on extracted surface geometrical features.
• Comparing structural morphology between tubulin and its bacterial analogue FtsZ.
• Testing protein classification methods using Support Vector Machine and k-means clustering.
• Applying computer vision feature extraction principles to protein surface analysis.

Strengths

• Methodology is tested on specific protein examples, including tubulin isotypes and FtsZ.
• Results are described as significant and competitive with state-of-the-art protein classification methods.
• The paper describes a novel application of biometric concepts to protein biophysics.

Limitations

• The dataset's size, row count, and specific file formats are unknown.
• Column-level documentation is absent; field semantics must be inferred from the paper.
• Last update date is unknown; freshness unverified.

Provenance

Source

Politecnico di Torino

Collection Method

Dataset built using a simulation tool, with feature extraction using novel geometrical descriptors.