The Challenge
The client needed a complete software suite for automated analysis and visualisation of metagenomic and metabolomic data. Existing workflows involved multiple tools and manual data transfers, limiting efficiency.
Our Approach
We standardised and automated omics analysis, combined machine learning and interactive tools, and integrated biomarker discovery into a single pipeline.
- Standardisation & automation: Implemented a standardised statistical methodology and built a modular pipeline with analytical and visualisation tools.
- Machine learning & tools: Developed ML models for metabolite identification and a RShiny web application for data analysis with ongoing support.
- Integration & biomarker discovery: Used the integrated pipeline to discover microbiome‑derived biomarkers and streamline the client’s gastrointestinal simulation technology.
The Outcome
- Delivered a custom, end‑to‑end software suite that automates analysis and reduces manual errors.
- Combined physical rules and ML to achieve accuracy metrics over 90 % for precision and recall.
- Reduced LC‑MS sample processing time to around 30 seconds.
- Provided interactive data exploration that empowers scientists.
Why It Matters
Bespoke analytics platforms enable CROs to deliver high‑quality insights to their clients while improving operational efficiency. Automating complex analyses reduces costs and accelerates discovery.