Currently, the platform uses random integration. However, copy numbers in the cells are typically quite low, so only a few copies are generally needed. 

The development team is also working on semi-targeted and fully targeted integration approaches. Early results have been very promising, and the long-term goal is to advance toward fully targeted integration as the future direction of the technology.

Monoclonality is ensured through two complementary approaches. 

First, for every customer project, a monoclonality assessment is performed using high-resolution microscopy. Images are captured on Day 0 to demonstrate that only a single cell was present in the nanowell. The nanowell structure, combined with high-resolution imaging, allows the number of cells to be assessed accurately. A second image is captured on Day 1 to confirm early cell growth from the original single cell. 

Second, an extensive monoclonality validation program has been conducted. Multiple single-cell cloning experiments were performed to evaluate factors that could negatively affect monoclonality, such as the presence of "ghost cells" that may not be visible on Day 0 but become apparent on Day 1. Based on this validation work, an overall monoclonality probability of 99.6% was achieved. 

Customers receive a monoclonality validation report documenting this high monoclonality probability, along with microscopy images for each top clone that provide evidence of monoclonality for the specific clone. 

The assay can be performed using different Octet^® platforms, but the Octet^® RH96 offers the advantage of enabling the entire workflow on a single plate. This includes ligand loading, equilibration, and sample measurement within one integrated process. 

Earlier-generation Octet^® instruments that do not support 96-well plate measurements can also be used. In these cases, the ligand can be coated onto the biosensors offline before the measurement step. While this approach does not allow real-time monitoring of ligand loading, it still enables the actual sample measurement to be performed successfully on the instrument. 

As a result, the assay can be adapted for laboratories that do not have access to the latest Octet^® platform, providing flexibility while maintaining assay functionality. 

The optional 96-well run is performed in a deep-well plate format rather than a classical fed-batch system. The process uses a 96-well deep-well plate with a working volume of approximately 300 µL per well, enabling small-scale evaluation of clone performance under fed-batch-like conditions.

For the Ambr^® 15 perfusion mimic, an intensified N-1 seed train is not required. The necessary cell density can be achieved through standard cell expansion prior to inoculation. Intensified N-1 perfusion is typically only needed for larger-scale perfusion systems. For example, before inoculating a production-scale perfusion bioreactor, an N-1 perfusion step may be performed in a 10 L bag to reach the required cell numbers. However, for the Ambr^® 15 perfusion mimic, standard expansion is sufficient. 

Prior to implementation in customer projects, the Ambr^® 15 perfusion system was extensively optimized by the development team. This included optimization of feeding strategies, perfusion media, cell bleed control, inoculation density, and target cell concentration. The optimized cell bleed concentration is typically maintained at 20 million cells/mL, which is lower than the values commonly used in larger perfusion systems. 

In terms of reproducibility, the system demonstrates relatively low variability when the same clone is run under identical conditions. Cell growth and product titers are generally comparable, with typical variation ranging from 10–20%. Internal studies in which individual clones were inoculated in duplicate across different Ambr^® 15 culture stations produced highly similar results, demonstrating good reproducibility across stations. 

Octet^® assay customization and method development typically require approximately six weeks. To minimize overall project timelines, method development is initiated as early as possible, often in parallel with gene synthesis. This allows assay implementation to progress while other project activities are underway, helping to avoid delays before single-cell cloning can begin. 

The timeline can vary slightly depending on ligand availability. While ligands do not need to be supplied in a biotinylated form—internal biotinylation can be performed if required—the ligand itself must be available. Early stages of the project may also involve literature review and ligand sourcing to identify appropriate binding reagents. 

When the ligand is readily available, Octet^® method development is typically completed within six weeks. 

In general, no. Experience has shown that extending the pool phase does not typically reduce the number of high-expressing clones. However, the outcome depends on the pool strategy being used. 

For pools that are intended for long-term cultivation and subsequent single-cell cloning, mini-pools are generally preferred over large, highly heterogeneous pools. Large pools can carry a higher risk that low-producing clones may outgrow higher-producing clones during extended cultivation, freeze-thaw cycles, or other handling steps. 

In contrast, mini-pools have consistently shown strong performance, with no significant overgrowth of low-producing clones observed. As a result, mini-pool strategies help maintain clone quality while providing greater confidence in downstream clone selection. 

Yes, in principle, predictive assays can be developed for non-Protein A-binding molecules. In these cases, alternative bead types and specialized coating chemistries would be required instead of Protein A-based capture methods. 

However, developing such an assay is generally only economically viable if it can be applied across multiple molecules. For example, it is most practical when a customer has a molecule platform where the same assay can be used for several related products. 

A key consideration is the development of a robust training dataset. Predictive models rely on sufficient training data, ideally generated from multiple products, to achieve optimal performance and predictive accuracy. As a result, while developing predictive assays for non-Protein A-binding molecules is certainly feasible, additional assay development and data generation are required.

Integrating perfusion during cell line development allows researchers to identify and select the clones that perform best under perfusion conditions from the very beginning of the development process. Not all clones behave equally well in perfusion, and meaningful differences in performance can exist between candidates. 

By evaluating and selecting clones under the same conditions that will ultimately be used for manufacturing, organizations can increase confidence that the chosen clone will perform effectively in the final production process. Early clone selection under perfusion conditions also reduces the amount of process optimization required during later development and scale-up stages. 

As a result, incorporating perfusion into CLD can streamline downstream development efforts, reduce risk, and help ensure that the most suitable clone is advanced into manufacturing.