ConsoneAI / DioScor

Context

The six-week research internship was set around a dose-toxicity platform. The value of the work was learning how messy biological and toxicological evidence has to be structured before it can support analysis and have its chemical and biological context propagated.

My role

I supported toxicology data mapping and structured data-preparation work, keeping the description here high level because the platform context was proprietary.

Approach

• Identify relevant toxicology and chemical data sources for extraction or comparison.
• Review dose-toxicity source fields and terminology differences.
• Organise compound, organism, route, dose-unit, test-type and effect-category information where available.
• Clean missing values, split combined categories and document limitations.
• Map fields to high-level information needs.

Outputs

• Internal mapping notes.
• Organised non-public data outputs.
• Dose-toxicity summaries.
• Poster presentation at the Anglia Ruskin PGR Conference 2024

What this shows

This was an early step from foundational biomedical science into data-enabled research: Python programming, careful extraction, cleaning, mapping and communication.