Discovering the structure of proteins is useful. This information informs pathophysiology research, pharmacology research, material science, and so on. Prior to the development of cryogenic electron microscopy, x-ray diffraction of crystallised protein was the primary method for determining structure.
Different proteins will crystallise under different conditions, e.g. pH, temperature, the presence of specific salts and substituents, and the concentrations thereof. In the absence of a deterministic method for predicting crystallisation success, trial and error was undertaken en masse using trays with hundreds of experimental wells.
The wells on each tray varied in crystallisation condition(s) on each axis, and crystallisation data was obtained by serial photography of the wells to detect if any crystals had formed. In this way, thousands of crystallisation cocktails could be tested using automated systems.
The suite of tools presented on this page were developed as proof-of-concept for querying, visualising, and assessing the resulting crystallisation data. This work was undertaken in 2011-2012 in the context of a student internship with a hosting research facility.
CQuery is a web interface for generating a JSON request that can be used to query results from a crystallisation database. It allows for users to string together complicated AND-OR statements in order to isolate specific subsets of data.
The tool can be used to make broad selections, such as "all experiments using any carboxylic acid with a molecular mass between X and Y; AND solution pH between 4 and 7".
CGrid is a web interface for visualising a set of crystallisation experiments in a 2D grid, based on two dynamically selected factors.
Hovering the mouse over an experiment data point displays a popup with photo and information about the experiment.
Intended but not implemented was the ability to then select a region of results from the CGrid interface and compare them using the CVis interface (see below).
CVis is a web interface for visualising a set of crystallisation experiments in a table format.
Each column represents a single crystallisation experiment. Rows of relevant information can be dynamically added or removed depending on the salient features being investigated.
A crystallisation experiment is typically associated with 8-12 serial photographs, and these can be viewed by hovering the mouse over the thumbnail on given column.
ChemUtils was a programming library that provided a comprehensive set of unit conversion and chemistry calculation functions to support the development of CQuery, CGrid, and CVis.