Summary
- The problem
- A culture comparison often survives as an overlay in a presentation. The cell source, media and feed versions, actual additions and samples that explain the result sit in separate files, so a difference between runs is hard to attribute.
- Seal’s approach
- Inoculation, feeding, sampling and harvest protocols are configured and executed in Seal, and each run keeps its cell source, recipe version, actual additions, conditions, samples and harvest. Neil can help compare cultures and draft the next study; the team reviews the conclusions.
- What changes
- Two runs are compared with the feed strategy and source records behind them, and a pattern stays a question until a replicated study tests it. The harvest carries its culture history into downstream purification.
- Where to start
- One culture protocol and a representative study. Book a demo.
Your culture protocol. Executed in Seal.
Upstream development usually records runs in notebooks and bioreactor software, with feeds, additions and samples tracked in spreadsheets and the comparison assembled in a presentation. Inoculation, feeding and sampling become configured workflows in Seal, and each run records its actual additions, observations and sample links as the study runs, not only in the final comparison. A recipe that works can become the controlled process for scale-up and transfer, with its development evidence attached.
Fed-batch culture
- Identify the cell source and media
- Record feeds and culture conditions
- Collect samples and record the harvest
USP-029
- Feed strategy
- Split feed
- Harvest day
- 14
- Harvest viability
- 89%
Harvest recorded
Configure with Neil. Test with your team. Reuse the protocol.
Each run keeps its own values, evidence and history.
Compare what changed in the bioreactor.
Compare culture profiles alongside the feed strategy and harvest results.
A culture comparison often survives as an overlay in a presentation, separated from the cell source, media and feed versions, actual additions and samples that explain it. In Seal each run keeps those records, so a difference between runs can be examined against what actually happened rather than what the recipe intended.
The fictional study below compares two 2 L fed-batch runs. USP-028 used a daily feed under recipe UP-08 v02; USP-029 used a split feed under v03. Each point is one reported observation, with no replicates. Choose a measure to see the profiles and Neil’s reading of them.
Two feed strategies. Same cell source.
Inspect the study records
Cell source CL-04 / bank WCB-02, starting working volume 2 L, basal medium M-06 v01 and feed formulation F-03 v02. Recipes UP-08 v02 and v03 differ in the planned feed schedule. Compare actual additions and control histories before treating this as a matched experiment.
| Day | USP-028 | USP-029 |
|---|---|---|
| 0 | 0.5 | 0.5 |
| 2 | 2.2 | 2.3 |
| 4 | 6.5 | 6.8 |
| 6 | 12 | 12.5 |
| 8 | 17 | 17.5 |
| 10 | 18 | 19 |
| 12 | 16 | 18 |
| 14 | 12 | 16 |
Viable cell density and viability: CELL-02 v03. Glucose: MET-04 v02. Harvest titre: TTR-02 v03; USP-028 3.6 g/L, USP-029 4.1 g/L. One observation at each time, not replicate means. No complete product-quality panel is supplied.
The cultures separate late in the run.
USP-029 retains more viable cells after day 10. Compare the glucose and viability profiles before attributing the difference to the feed schedule.
Two runs suggest a question—not an optimum.
Read the draft process comparison
Test whether the pattern repeats.
The split-feed run has a higher observed day-14 titre (4.1 versus 3.6 g/L) and viability (89% versus 78%). The sampled glucose profile also differs. These observations do not isolate a feed effect.
Proposed follow-up
- Reconcile delivered feed, seed history, pH, dissolved oxygen and temperature records for both runs.
- Design a replicated feed comparison with matched cell source, media, sampling and harvest criteria.
- Assess product quality and harvest suitability alongside titre and viable cell density.
Draft based on fictional USP-028 and USP-029, recipe versions and sample results. No feed strategy or scale-up is approved.
USP-029 keeps more viable cells after day 10 and finishes at 89% viability against 78%. That is a pattern worth testing, not evidence that split feeding is better. Glucose was sampled every two days, so the samples do not show the minimum between measurements, and titre and quality results still need review before a harvest strategy is chosen. Neil drafts the follow-up study; the team decides its design and which process to advance.
From the seed train to the harvest.
Each study should inform the next rather than become another disconnected spreadsheet.
Follow the starting vial through seed expansion and into the bioreactor, with media and feed versions, actual additions, control profiles, interventions and samples recorded against process age. Fed-batch and perfusion runs keep the time windows and calculation bases their analysis needs. When the process moves to a larger scale, carry the rationale with it: geometry, mixing and transfer assumptions, operating ranges and confirmation studies. Matching setpoints do not by themselves establish equivalent performance.
Culture conditions and control
Compare the process that actually ran: pH, dissolved oxygen, temperature, gases and interventions.
What stays connected
Align setpoints, measured signals, alarms and manual changes with process age and sample times. Retain vessel and sensor context rather than treating an overlay as proof of comparable conditions.
Media and feed development
See the formulation and actual additions behind the performance—not just the recipe name.
What stays connected
Retain material lots, formulation versions, feed timing and amounts, sample results and deviations. Separate intended changes from what happened in the run.
Seed train and culture
Follow the starting vial through expansion and into the bioreactor. Compare the run in context.
What stays connected
Keep inoculum age, viability, vessel, working volume, setpoints, actual signals and interventions with the samples they affect. Preserve time windows for fed-batch and perfusion analysis.
Scale-up and harvest
Carry the process rationale to the next scale, then hand downstream a traceable harvest.
What stays connected
Record geometry, mixing and transfer assumptions, operating ranges, confirmation studies and harvest conditions. Connect the resulting material to purification results without treating scale-up as a copy of setpoints.
The harvest is a handoff, not the end of the record.
Let downstream see which culture, conditions and samples produced its starting material. Bring purification findings back into the development decision.
Explore downstream process developmentKeep the instruments. Connect the evidence.
Work from Ambr, bench-scale bioreactor and historian exports, analytical results and existing study records. Scope the connection around the question your team needs to answer.
Explore connectionsAQuestions and answers
Can we configure and execute protocols in Seal?
Yes. Configure reusable protocols with instructions, fields, calculations, linked samples and review steps. Each execution retains the version used, recorded values, observations and results. Neil can help prepare the configuration and analyse the resulting evidence; your team tests the workflow and defines the review and approval requirements for its intended use.
How is this different from cell line development?
Cell line development evaluates and selects the biological candidate. Upstream process development establishes how to culture that cell line or strain: media, feeds, seed train, control strategy, process mode, harvest and scale-up. The records connect, but they answer different questions.
Can we compare Ambr and bench-scale studies?
Start with permitted exports or an agreed connection. Keep vessel geometry, working volume, cell source, media and feed versions, process age, actual control profiles and sampling methods in view. An overlay alone does not establish that scales or conditions are equivalent.
Does the example prove split feeding is better?
No. It shows two fictional 2 L fed-batch runs with different feed schedules and one reported observation at each sample time. There are no replicate estimates or complete quality results. The pattern supports a follow-up comparison, not a causal claim or an approved feed strategy.
Do you cover perfusion as well as fed-batch?
Yes. Connect exchange and bleed rates, cell retention, filter state, viable cell density, sample times and harvest intervals. Use calculation bases appropriate to the mode; a perfusion harvest interval is not interchangeable with a final fed-batch harvest.
What does the team receive?
A source-linked run comparison, differences in recipe and actual execution, open analytical questions and a draft follow-up study. Scientists review the interpretation, choose the design and decide which process to advance.
What is carried into scale-up and transfer?
The selected recipe, materials, operating ranges, scale criteria, equipment assumptions, analytical methods and supporting studies. Keep receiving-site differences and verification work explicit. Copying setpoints does not demonstrate equivalent process performance.