AAV affinity purification depends on interacting parameters such as load volume, linear velocity, buffer pH and conductivity that create a design space too large to screen exhaustively. INSUS learns from runs you have already done, predicts recovery and purity for candidates and proposes which experiment to run next. Process development teams keep change control.
01
The goal is a strong condition in fewer runs.
02
Every completed experiment sharpens the next recommendation.
03
Scientists decide what gets run under existing change control.
The problem
AAV affinity chromatography depends on parameters that interact: load volume, linear velocity during wash and elution, buffer pH and conductivity. With serotype-specific behaviour, the space cannot be explored exhaustively. Each run consumes material, column time and analytical capacity, and recovery and purity are known only when the run is finished. The constraint is how many experiments it takes to find a better operating region.
The shift
Every purification experiment you have already run is information. What changes is whether that history is used to choose the next one.
Today
With INSUS
How it works
This fits OptiStudio as process development decision support beside existing chromatography workflows and LIMS. Access is read-only. The process development team decides which experiments are run.
Connect
Connect chromatography parameters such as load volume, linear velocity, buffer pH and conductivity with recovery, capsid titer and purity results, plus product and serotype context from existing workflows and LIMS or QC systems.
Understand
Learn how purification parameters have related to recovery, purity and titer across your experimental history, with product and serotype differences represented rather than averaged away.
Predict
Estimate likely recovery, purity and titer for candidate parameter combinations, together with uncertainty around each estimate, so the next experiment can balance promise and learning.
Advise
Propose the next experiment using optimization that weighs promising conditions against those that would teach the model the most. The team reviews each proposal and decides whether to run it.
Verify
Compare each result with what the model predicted and fold new data back in so recommendations become more specific to this product and process.
Getting started
For gene therapy process development teams with affinity purification and QC history. Delivered through OptiStudio as downstream decision support. Expert Services can help with data assessment and validation. Advisory and read-only. No new purification hardware.
Related for biotech
AAV affinity purification optimization sits beside cell culture titer optimization and Annex 1 contamination control strategy on the same biologics operation.
AAV affinity purification has a large, interacting parameter space where each run is costly. This use case learns from prior runs and proposes which condition to test next for process-development review.
Purification parameters with matching recovery, capsid titer and purity results, plus product or serotype information. No new sensors when chromatography and QC workflows already capture those records.
It weighs predicted outcomes and uncertainty, favoring experiments that either move toward a better operating region or reduce uncertainty where the model knows least.
Yes, where enough process data exists. Each product or serotype generally needs its own calibration rather than assuming one model transfers unchanged.
No. It proposes experiments for the process development team. Nothing is written back to any system.
Through OptiStudio as process development decision support. Expert Services can support data assessment, integration and validation with your team.
Performance depends on experimental history and coverage. Published Bayesian optimization studies for AAV purification are not INSUS results for your product. Recovery and purity targets must still be confirmed by your own runs and QC.
Biotech · Downstream purification
Start with one product or serotype and the purification history behind it. The useful first answer is what your existing runs already say about where the next one should be.
