How Solugen 10Xed their Sample Throughput using Ganymede’s platform

The chemicals sector is a $5 trillion market that affects all aspects of the modern economy, from agriculture, to building materials, to oil & gas. The industry relies heavily on fossil fuels for manufacturing, but thanks to rising biomanufacturing companies like Solugen, the chemicals industry is on the path to rapidly become more sustainable.

Being a pioneer in an emerging sector means generating huge amounts of R&D and Process Engineering data, and with that comes data inefficiencies that become a bottleneck for Solugen in developing and scaling their manufacturing capabilities. Enter Ganymede.

Among various groups, the Analytical Chemistry team has been leveraging Ganymede’s Scientific Data Cloud to automate their instrument data capture and analysis, saving significant time, improving throughput, and giving new insights into how to accelerate their process development.

Solugen’s Challenges?

Manual movement from instrument to ELN?

• Before bringing on Ganymede, data was entered into Solugen’s ELN manually, and would sometimes take days or weeks to get entered, leading to missed data.?
• Teams had to enter dozens of request results into their ELN per day, creating a large burden of data entry as well - scientists could spend hours filling out experiments.?

Limited ability to compare historical data?

• There was no easy way to compile historical data to identify trends. This was done by manually searching through ELN entries and copy pasting data into a spreadsheet.?
• This also prevented Solugen from building up a centralized data store for future AI or machine learning analytics on their historical data?

Data integrity?

• Each scientist was manually performing repetitive statistical analysis on their optical spectrometry data and photometric data?
• Analysis traceability was lost - whose Excel sheet calculated what??

Solution?

• Ganymede installed its agents on Solugen’s instrument PCs to capture data automatically - no human input required?
• Ganymede parses the instrument data into a standardized, instrument-agnostic form?
• Ganymede automatically calculates some of Solugen’s scientific analyses
• Finally, Ganymede automates data entry into Benchling

Using Solugen’s ICP-OES spectroscopy system as an example:?

• When a request comes in and an entity is registered in Benchling, Ganymede logs this?
• Now, when the user runs analysis, Ganymede associates the request with analysis results?
• Ganymede also automates the Area Under the Curve analysis that was previously manual?
• Additionally, Ganymede greatly reduces time spent curating the data by automating data cleaning and automatically creating data views the scientists were manually compiling?

Outcomes?

• Automated analysis and entry?

Ganymede’s agents and flows to automated spectroscopy and discrete analyzer data. Scientists automatically had data analyzed and uploaded into their ELN system, saving several hours per instrument run.?

• Harmonized data set?

The Analytical Chemistry team can now compile a harmonized historical data set in minutes to generate dashboards and derive insights from Ganymede.?

• Improved Data Integrity?

Non data experts are able to record and report on data instead of having to a data engineer to process it for them?

• Data Breadth?

By ensuring that data is automatically captured and structured uniformly, Solugen’s deployment of Ganymede ensured 100% data quality and compliance for every instrument run. Solugen was also able to capture more fields of structured data without burdening scientists with manually recording it.

Key Value Results

To ask questions directly, join myself and Mark Stuczynski, lead automation and data engineer at Solugen, in a webinar happening tomorrow delving into this case study.

What: Webinar with Nathan Clark and Mark Stuczynski discussing the Solugen/Ganymede case study?

When: Thursday August 22nd at 2 - 2:30 PM EST / 11 - 11:30 AM PST.

Registration Link: https://app.livestorm.co/ganymede-bio/solugen-and-ganymede-10x-sample-throughput?s=432e777d-5de4-470a-b857-2daa21174e3b