Biological macro-molecules come in a large variety of complexes. Representation and modelling their structures are essential to provide a computational analogue of experimentally determined structures. In this work, the authors propose a representational scheme for biomolecules with the aid of approximate cell decomposition and a nano-robotic schema. The kino-chemical information of a molecule is programmatically retrieved from reliable repositories (RLooM, Protein data bank etc.)of biochemical structures and a Nano-manipulator schema is obtained from the D-H representation of the molecular chain under consideration. The molecule is enclosed in a suitable bounding box, and using approximate cell decomposition in conjunction with bonding information, a volumetric steric clash description of the chain is proposed which serves as a model for infringement function estimation and subsequent structural prediction for proteins and nucleic acids. From the robotic representation, the variable kinematic description due to rotating atoms is always at hand to generate the new corresponding infringement description within the bounding cuboid. Volumes of appropriately indexed cells give a reasonably accurate estimate of infringement in a particular configuration and the estimates are found to be comparable to experimental results. The proposed rendering have been extensively compared with a number of professional software renderings from NMR and crystallographic data.
Paper
Full text
Volumetric Representation of Biomolecules Using Cell Decomposition and Robotics
Semantic Scholar · Engineering · 2019
Abstract
Biological macro-molecules come in a large variety of complexes. Representation and modelling their structures are essential to provide a computational analogue of experimentally determined structures. In this work, the authors propose a representational scheme for biomolecules with the aid of approximate cell decomposition and a nano-robotic schema. The kino-chemical information of a molecule is programmatically retrieved from reliable repositories (RLooM, Protein data bank etc.)of biochemical structures and a Nano-manipulator schema is obtained from the D-H representation of the molecular chain under consideration. The molecule is enclosed in a suitable bounding box, and using approximate cell decomposition in conjunction with bonding information, a volumetric steric clash description of the chain is proposed which serves as a model for infringement function estimation and subsequent structural prediction for proteins and nucleic acids. From the robotic representation, the variable kinematic description due to rotating atoms is always at hand to generate the new corresponding infringement description within the bounding cuboid. Volumes of appropriately indexed cells give a reasonably accurate estimate of infringement in a particular configuration and the estimates are found to be comparable to experimental results. The proposed rendering have been extensively compared with a number of professional software renderings from NMR and crystallographic data.