How Matchpoint Therapeutics accelerated discovery by fusing siloed data into a unified pipeline
Highlights
Unified data pipeline
Integrating own assay results, public data and off-the-shelf tools in a flexible cloud
Speed and flexibility
Experiment-to-insight in realtime, tailored to Matchpoint’s unique approach
Data quality by design
Maximum automation with scientist input guided by intelligent assistants
Client
Matchpoint is a biotechnology company harnessing the power of covalency to discover precision covalent medicines to transform the treatment of immune diseases and other serious illnesses. The company's proprietary Advanced Covalent Exploration (ACE) platform integrates advanced chemoproteomics, machine learning and covalent chemistry library evolution. Matchpoint has an emerging pipeline of novel covalent medicines initially focused on immunology.
The company was founded in 2021 by Atlas Venture and Access Biotechnology together with leading scientists from the Dana Farber Cancer Institute and Stanford and has raised $100M to date.
Challenges
Speed, flexibility and data quality need to be balanced out
High volume of non-standard experiments
Cross-functional teams need to work independently of data support
Solution
Pipeline architecture
Success factor
Custom dashboards for all standardized analysis
Moving quickly from experiment to insight to decision is key to accelerating discovery.
Therefore, we created custom dashboards automating the most common types of analysis, reflecting Matchpoint’s unique way of looking at the data.
Moreover, experimental results are contextualized with further annotations integrated from external databases.
Success factor
Convenient user interfaces supporting manual data ingestion steps along the pipeline
When ingesting new data, expert input is crucial e.g. for quality checks and organizing it into meaningful logical units.
To make this as efficient as possible, we developed a web-based user interface which guides through the entire process.
Apart from speed and convenience, this also reduces the scope for errors and enforces Matchpoint’s internal standards by design.
Success factor
Close collaboration with the science and platform technology teams on all key design decisions
To optimally support discovery, data structure and pipeline setup need to be closely aligned with the R&D strategy.
Therefore, we co-designed the system to match current and anticipated future needs.
This created a high level of transparency and alignment, reflected in a visual map of the pipeline on a shared canvas, which is continuously updated with changes and feature requests.
What our client says
“It is a pleasure working with the idalab team on our data and machine learning pipeline. They are an outstanding strategic partner, collaborating seamlessly with our science team. Fast, clear communication, structured — yet always happy to adapt ad hoc, if necessary. We are looking forward to continuing the collaboration.“
Suresh Singh, PhD
Senior Vice President,
Computational Sciences
Get in touch
Let's talk data
If you are not getting what you want out of your data: Let’s talk.
Fully confidential (NDA), with no strings attached.
Frequently asked questions
-
We supported Matchpoint along the entire process. Working closely with the science team, we conceptualised the pipeline in a series of workshops. Afterwards, we implemented it piece-by-piece in Matchpoint’s cloud environment while keeping the science team in the loop to ensure data quality and consistency.
-
Twelve weeks from the kick-off workshop to the delivery of the initial pipeline.
-
Anything goes, even custom tools written in Fortran can be brought into a 21st century cloud pipeline.
-
Definitely! We would love to team up with your data engineering team, and actively support phasing us out when the time is right.
-
Programming language: Python
Web-app development: Streamlit (a Python package)
Data lake: Google Cloud Storage
Data warehouse: Google BigQuery
Web-app deployment: Google App Engine
Securing web-app access: Google Identity-Aware Proxy
Integration of external tools: Google Cloud Functions
For this project, the requirement was to implement everything in Matchpoint’s Google Cloud environment, but any other cloud provider (e.g., AWS, Azure) works as well.
Learn more
