Abstract
BACKGROUND: Laboratory Information Management Systems (LIMS) are an increasingly important part of modern laboratory infrastructure. As typically very sophisticated software products, LIMS often require considerable resources to select, deploy and maintain. Larger organisations may have access to specialist IT support to assist with requirements elicitation and software customisation, however smaller groups will often have limited IT support to perform the kind of iterative development that can resolve the difficulties that biologists often have when specifying requirements. Translational medicine aims to accelerate the process of treatment discovery by bringing together multiple disciplines to discover new approaches to treating disease, or novel applications of existing treatments. The diverse set of disciplines and complexity of processing procedures involved, especially with the use of high throughput technologies, bring difficulties in customizing a generic LIMS to provide a single system for managing sample related data within a translational medicine research setting, especially where limited IT support is available. RESULTS: We have designed and developed a LIMS, BonsaiLIMS, around a very simple data model that can be easily implemented using a variety of technologies, and can be easily extended as specific requirements dictate. A reference implementation using Oracle 11 g database and the Python framework, Django is presented. CONCLUSIONS: By focusing on a minimal feature set and a modular design we have been able to deploy the BonsaiLIMS system very quickly. The benefits to our institute have been the avoidance of the prolonged implementation timescales, budget overruns, scope creep, off-specifications and user fatigue issues that typify many enterprise software implementations. The transition away from using local, uncontrolled records in spreadsheet and paper formats to a centrally held, secured and backed-up database brings the immediate benefits of improved data visibility, audit and overall data quality. The open-source availability of this software allows others to rapidly implement a LIMS which in itself might sufficiently address user requirements. In situations where this software does not meet requirements, it can serve to elicit more accurate specifications from end-users for a more heavyweight LIMS by acting as a demonstrable prototype.