Case Studies
User-friendly database querying
Value
Seamless connectivity between design and decision making tools and the data management system ensures that project teams have access to the latest information without needing to compile the data manually from many sources. This leads to better decisions and shorter project timelines.
The search interface has been designed with flexibility in mind. It can be easily adapted to different database architectures and APIs.
A key challenge in drug discovery is ensuring that project leaders and decision makers have access to the latest and most relevant data for their projects. In this case study we describe the results of a collaborative effort to develop a user-friendly graphical interface for creating, sharing and executing structured database queries and presenting the results in StarDrop, enabling the user to visualise and analyse their data to identify optimisation strategies and guide the design of new, high quality compounds.
Search Interface Requirements
Before embarking on development, it was essential to understand the requirements for a successful outcome. Working closely with end-users, the following points were identified:
- User-friendly definition of search criteria and fields
- Save, share, edit and execute pre-defined queries
- Support for criteria based on chemical structure, numerical, date, textual and categorical fields
- Support for multiple data aggregation levels
- On demand drill down to data underlying aggregated values
- Refresh query to update results and analyses with new data
- Provide access to multiple data sources
Solution
A search interface was developed within StarDrop, enabling users to create, share and execute database queries across multiple data sources. An intuitive drag-and-drop graphical user interface makes it easy for end users to define the data they need without the need to understand the underlying database architecture.
The query tool translates the user input into an SQL query that is run against a data mart via an ODBC connection. Additional data sources may be easily configured and the tool may be adapted to other database APIs.
The data are returned directly in StarDrop, without the need for further formatting. The results displayed may represent aggregated values from multiple measurements and the individual, underlying data points and associated meta-data may be easily accessed via a drill-down feature. This enables users to easily investigate outliers and identify potential errors.
Architecture
This project was completed in collaboration with The Edge Software Consultancy and the underlying architecture is based on their BioRails™ database and data mart.
- Biological data is extracted and aggregated by compound, salt and lot nightly
- Results are stored in a data warehouse with associated data from the compound registration database
- These data are pivoted and stored in a data mart which facilitates easy access to results and associated metadata
- StarDrop users can use the query tool to design and run queries against the data mart.
Data Visualisation and Analysis
Value comes from data through the selection and design of high quality compounds, using capabilities such as:
- Data visualisation, including StarDrop’s Card View
- Analysis of structure-activity relationships
- Multi-parameter optimisation
- In silico modelling and de novo design
