MediSyn: uncertainty-aware visualization of multiple biomedical datasets to support drug treatment selection

Chen He, Luana Micallef, Ziaurrehman Tanoli, Samuel Kaski, Tero Aittokallio, Giulio Jacucci

Abstract

Background: Dispersed biomedical databases limit user exploration to generate structured knowledge. Linked Data unifies data structures and makes the dispersed data easy to search across resources, but it lacks supporting human cognition to achieve insights. In addition, potential errors in the data are difficult to detect in their free formats. Devising a visualization that synthesizes multiple sources in such a way that links between data sources are transparent, and uncertainties, such as data conflicts, are salient is challenging. Results: To investigate the requirements and challenges of uncertainty-aware visualizations of linked data, we developed MediSyn, a system that synthesizes medical datasets to support drug treatment selection. It uses a matrix-based layout to visually link drugs, targets (e.g., mutations), and tumor types. Data uncertainties are salient in MediSyn; for example, (i) missing data are exposed in the matrix view of drug-target relations; (ii) inconsistencies between datasets are shown via overlaid layers; and (iii) data credibility is conveyed through links to data provenance. Conclusions: Through the synthesis of two manually curated datasets, cancer treatment biomarkers and drug-target bioactivities, a use case shows how MediSyn effectively supports the discovery of drug-repurposing opportunities. A study with six domain experts indicated that MediSyn benefited the drug selection and data inconsistency discovery. Though linked publication sources supported user exploration for further information, the causes of inconsistencies were not easy to find. Additionally, MediSyn could embrace more patient data to increase its informativeness. We derive design implications from the findings.

Introduction

Biomedical research currently faces a significant "silo" problem. While vast amounts of curated data exist in databases like the Cancer Genome Interpreter (CGI) and Drug Target Commons (DTC), researchers must often manually correlate clinical biomarkers with pre-clinical bioactivity data. This process is cognitively exhausting and prone to overlooking critical discrepancies. MediSyn addresses this by providing a unified, matrix-based interface that explicitly highlights data uncertainty and inconsistencies, bridging the gap between the lab bench and the clinic.

System Design and Methodology

MediSyn integrates two primary datasets: CGI, which provides clinical evidence for cancer treatment biomarkers, and DTC, which contains pre-clinical drug-target bioactivity values. The system employs a matrix-based layout where rows represent drugs and columns represent genetic targets (mutations).

The MediSyn interface uses a matrix-based layout to visually link drugs, mutations, and tumor types, highlighting data provenance and specific bioactivity values like Kd.

A key feature is the use of "superimposed views." Instead of merging conflicting data into a single consensus value, MediSyn layers information within each cell. Clinical responsiveness is encoded by hue (green for responsive, red for resistant), while experimental potency is represented by overlaid black bars. This design makes "data inconsistency states" salient, prompting researchers to investigate the underlying evidence via direct links to source publications (DOIs). Furthermore, the inclusion of "wild-type" gene controls allows for the assessment of drug selectivity, helping to predict potential side effects.

Results

The effectiveness of MediSyn was evaluated through a user study with six domain experts. Participants performed drug selection tasks using both MediSyn and the baseline web interfaces of the source datasets.

Comparison of task completion times between the baseline interfaces and MediSyn across six participants, showing a substantial reduction in time for drug selection tasks.

The study revealed a significant performance boost: tasks that took an average of 6 minutes 22 seconds with baseline tools were completed in just 1 minute 57 seconds with MediSyn—a 3.3x improvement in efficiency. Beyond speed, participants identified all data conflicts correctly and reported higher confidence in their selections due to the transparent synthesis of evidence and provenance.

Discussion and Future Work

MediSyn demonstrates that making uncertainty a first-class citizen in visualization can improve scientific decision-making. While the tool successfully flags inconsistencies, future versions aim to provide more context on why these discrepancies occur, potentially by integrating patient-specific data such as treatment history and demographics. Additionally, the authors plan to explore disease-level linking to further enhance drug repurposing efforts.

Conclusion

By synthesizing disparate biomedical datasets into a single, uncertainty-aware interface, MediSyn empowers researchers to make more informed, evidence-based drug treatment selections. Its ability to highlight hidden opportunities for drug repurposing while maintaining scientific rigor through provenance links marks a significant step forward in biomedical data visualization.

Summary updated

Abstract

Abstract Background Dispersed biomedical databases limit user exploration to generate structured knowledge. Linked Data unifies data structures and makes the dispersed data easy to search across resources, but it lacks supporting human cognition to achieve insights. In addition, potential errors in the data are difficult to detect in their free formats. Devising a visualization that synthesizes multiple sources in such a way that links between data sources are transparent, and uncertainties, such as data conflicts, are salient is challenging. Results To investigate the requirements and challenges of uncertainty-aware visualizations of linked data, we developed MediSyn, a system that synthesizes medical datasets to support drug treatment selection. It uses a matrix-based layout to visually link drugs, targets (e.g., mutations), and tumor types. Data uncertainties are salient in MediSyn; for example, (i) missing data are exposed in the matrix view of drug-target relations; (ii) inconsistencies between datasets are shown via overlaid layers; and (iii) data credibility is conveyed through links to data provenance. Conclusions Through the synthesis of two manually curated datasets, cancer treatment biomarkers and drug-target bioactivities, a use case shows how MediSyn effectively supports the discovery of drug-repurposing opportunities. A study with six domain experts indicated that MediSyn benefited the drug selection and data inconsistency discovery. Though linked publication sources supported user exploration for further information, the causes of inconsistencies were not easy to find. Additionally, MediSyn could embrace more patient data to increase its informativeness. We derive design implications from the findings.
BibTeX
@article{he2017medisyn,
  author = {Chen He and Luana Micallef and Ziaurrehman Tanoli and Samuel Kaski and Tero Aittokallio and Giulio Jacucci},
  title = {{MediSyn: uncertainty-aware visualization of multiple biomedical datasets to support drug treatment selection}},
  journal = {BMC Bioinformatics},
  year = {2017},
  doi = {10.1186/s12859-017-1785-7},
}
Figure