Highly Parallel Protein Discovery Workflows

We will discuss how high-resolution MS-based approaches, namely Multiple Attribute Method (MAM), together with mass photometry, capillary electrophoresis, and biophysical tools, provide powerful capabilities for the characterization of mAbs and other complex format biotherapeutics. Building on these tools, we expanded our analytical portfolio to support rAAVs characterization, enabling deeper assessment of viral protein ratios, key PTMs (e.g., glycosylation, oxidation, deamidation), and HCPs in both purified and in-process samples.

Delivering screening-ready protein for difficult targets requires moving beyond empirical construct design toward data-driven decision-making. Our integrated MS platform—HDX-MS, Native MS, and intact/peptide mapping—enables precisely this, compressing timelines and de-risking production of challenging targets. Using a transcription factor heterodimer case study, we demonstrate how HDX-MS identified critical construct boundaries missed by published designs, Native MS confirmed complex stoichiometry, and the resulting material directly enabled a covalent fragment screening campaign. Looking ahead, we preview our fast HDX platform engineered to detect transient druggable conformations within intrinsically disordered proteins, expanding the druggable proteome into historically intractable space.

Protein conformational stability influences the developability, pharmacokinetics, and efficacy of biological drugs. Conventional methods such as differential scanning fluorimetry and circular dichroism often miss transient local unfolding events linked to aggregation, proteolytic susceptibility, and rapid clearance. Here, we describe a high-throughput, automated thermolysin-based assay for kinetic stability profiling of proteins that captures dynamic conformational instability using minimal sample quantities, enabling rapid prioritisation of stable candidates during drug development.

We present a miniaturised expression platform that boosts screening throughput without compromising yield. By reducing culture volumes from 200 mL to 4.2 mL, we increased capacity from tens to hundreds of clones per batch while maintaining g/L production. The workflow integrates expression, purification (including tag removal), and conjugation into an automated, high-throughput process, enabling earlier developability assessment and scalable data generation to support advanced screening and predictive modelling.

Monoclonal antibody productivity in CHO cells depends on both sequence design and balanced chain expression. High-throughput screening combined with design-of-experiments enables rapid identification of optimal construct features such as codon usage, signal peptides, kappa/lambda choice, and LC:HC promoter ratios. These strategies accelerate selection of high-yielding candidates, improving titre, cell viability, and predictability before stable cell line development.

Integrated workflows for protein production and characterisation can generate high-quality datasets for machine learning. We outline practical frameworks for minimal yet sufficient data capture across the full pipeline, highlighting the value of negative data. Emphasis is placed on practical approaches to link and record design, expression, purification, and characterisation during protein production, and on implementing these approaches in both academia and industry while balancing data quality with resource constraints.

The Baculovirus-based BacMam system has recently regained attention through major therapeutic advancements. The production of complex bio-nanoparticles demands large transgenes and tunable expression, reviving this powerful platform. This presentation will highlight transient applications, recent progress, and introduce our newly developed REM-BAC system for rapid, efficient, and manifold baculovirus-mediated generation of stable mammalian cell lines via viral-footprint-free and site-specific genome integration.

The successful development of multispecific immune cell engaging-antibodies requires multifactorial screening of binding arm combinations, epitope pairing, and architectures. This approach is enabled by highly parallel protein and cell sciences workflow setups. In this presentation, we will cover four key areas: antigen toolboxes required for multi-epitope targeting of complex antigens; automation platforms enabling multiplexed screening of multispecific antibodies; an in-house large-scale production platform, and early-stage developability assessment. Together, these capabilities streamline multispecific antibody engineering and increase the likelihood of identifying potent, manufacturable immune cell engagers suitable for downstream development.

We are pioneering the use of IgE and IgA as therapeutic options in oncology. The engineering, production, and purification of IgE and IgA requires new, creative strategies compared to IgG. This talk will focus on our recent progress in producing candidate therapeutic IgA. Alongside natural and modified IgA molecules, we also engineer and produce more complex IgA fusion molecules, which will be covered in this talk.

A persistent challenge in biologics development is bridging early discovery with scalable manufacturing. Using a viral vector case study, this presentation illustrates how empirical process development evolved into a more predictive workflow integrating targeted experimentation, process modelling, and digital twins. By combining structured development approaches with data-driven tools, experimental burden can be reduced while improving process understanding, scalability, and decision-making, highlighting how predictive bioprocessing supports parallel and efficient discovery-to-manufacturing workflows.