How to find the right candidate faster in early biologics development

When failure happens in biologics development, it tends to happen at a late stage – and at significant cost. A molecule that fails in an advanced stage due to poor pharmacokinetics or manufacturability issues has likely already consumed years and millions of dollars.

The majority of biologics failures can be traced back to decisions made in the first 12–18 months of a program, such as poor assumptions about the target affinity, undetected stability issues that emerge in longer studies, or manufacturability problems that are discovered when reformulation is no longer an option. Many traditional biologics workflows delay rigorous biophysical characterization during this stage to maintain early throughput levels. But when the risk of late detection is high, this can be a false economy.

A successful biologics workflow is not about screening the most candidates, but about screening more thoroughly: asking the right biophysical questions at the right stage, and being ready to act on the answers. By rapidly identifying molecules with appropriate developability characteristics, you can reduce the risk of costly failures further down the line. A well-filtered funnel takes 3–5 candidates with a higher individual probability of success, lowering the total program cost and enabling a cleaner signal from in vivo studies.

Discover Malvern Panalytical’s complete biologics workflow

This practice requires knowing what to measure, when, and how to interpret it in a forward-looking way. In this blog, we outline the criteria for success – and the analytical instruments to help you get there.

Three key metrics for screening success

In the early development stage, many candidates are screened for hits with a cascade of biochemical and biophysical assays, aiming to validate hits and identify lead candidates. The goal is to identify which molecules bind the target meaningfully, behave in a way that suggests they can survive manufacturing and clinical use, and warrant further investment.

The early screens with the highest forward-predictive value include:

Binding affinity and kinetics

Affinity (K_D) is the most commonly reported early metric. However, it can be misinterpreted, since two molecules with identical K_D values can have different biological effects. For a fuller picture of future efficacy, combine affinity with the full thermodynamic and kinetic profile:

• ΔH (binding enthalpy): reports on the heat released or absorbed when one molecule binds another, reflecting the strength and nature of non‑covalent interactions such as hydrogen bonds and van der Waals contacts.

• −TΔS (entropic contribution): to the binding free energy and captures changes in molecular order and solvent release upon complex conformational change.

• k_off (dissociation rate): A slow k_off means the drug stays bound longer, which can translate to better in vivo efficacy, even at lower doses.

• k_on (association rate) matters in contexts with low target concentrations, such as in rare receptors.

Isothermal titration calorimetry (ITC), a label-free technology where both interaction partners are in solution, measures the heat released or absorbed during binding and can reveal whether binding is driven predominantly by enthalpy or entropy. It provides direct insight into the thermodynamics of the interaction; even when two interactions exhibit the same K_D.

Grating-coupled interferometry (GCI), a label-free optical biosensor technology, can measure mass accumulation on a sensor surface in real time, thereby providing k_on, k_off, and K_D.

Particle size analysis

Aggregation is a primary barrier to developability in biologics. Dynamic light scattering (DLS) enables you to screen for this early by measuring the size distribution of molecules in solution, flagging candidates that already show signs of aggregation or polydispersity under standard solution conditions. DLS offers insights into the following:

• Polydispersity index (PDI): High PDI early in development is a predictor of formulation and manufacturing difficulty – a clear criterion for reprioritization.

• Colloidal stability under stress: Candidates that maintain tight, monodisperse distributions across stress conditions – such as temperature and pH variations – are prioritized.

• Formulation space mapping: Running DLS across gradients of pH, ionic strength, and excipient concentration identifies the colloidal stability envelope for each candidate.

Thermal stability analysis

Thermal and colloidal stability (T_m, T_agg) are the fastest readouts of how a molecule will behave in a manufacturing context. A molecule with a lower Tm generally presents greater formulation risk. Screening this before committing to large-scale expression can save weeks of wasted production effort.

Differential scanning calorimetry (DSC) is the gold standard for characterizing the thermodynamic stability of a biologic’s folded structure, functioning as a high-information surrogate for long-term storage stability. Critical parameters include:

• Melting temperature (T_m) and onset temperature (T_onset): DSC measures the heat absorbed as a protein unfolds with increasing temperature. Higher T_m generally correlates with better real-world stability at 2–8 °C or room temperature.

• Conformational vs. colloidal stability: Combined with DLS data, DSC reveals whether aggregation is coupled to unfolding (a two-state failure) or occurs independently at sub-denaturing temperatures (a colloidal stability problem).

• Lot-to-lot comparability: DSC thermograms are sensitive fingerprints of higher-order structure. In early development, this makes DSC useful for confirming that different expression batches or purification conditions are producing structurally equivalent material.

 Your essential analytical toolkit – only with us

The GCI Wavesystem and MicroCal PEAQ-ITC from Malvern Panalytical support target selection and validation by working together to build a complete picture of a binding interaction before significant downstream commitment. GCI can be used to screen the potential of hundreds of compounds, establishing whether the target has the expected kinetic characteristics and whether it is prone to surface artefacts such as rebinding or heterogeneous ligand activity. ITC measures heat evolved or absorbed when two molecules interact – the only technique that directly measures the full thermodynamic signature of a binding event. A candidate that clears both stages – with high-quality kinetics from GCI and a clean thermodynamic profile from ITC – enters lead optimization with the data quality and confidence needed to support faster, better-informed decisions, helping reduce the risk of costly failures further down the line.

The Zetasizer Advance range combines industry-leading particle characterization, intelligent automation, and compliant data integrity in one powerful platform. It supports target selection and validation by confirming that a target protein is monodisperse, homogeneous, and behaves as expected in solution. Across multiple expression batches, the Zetasizer serves as a rapid, low-material comparability tool, confirming structural consistency before committing to more resource-intensive studies.

The MicroCal PEAQ-DSC is ideally suited to biosimilarity and batch-to-batch comparability assessment, as well as for the optimization of purification and manufacturing conditions, requiring little assay development and no labeling or immobilization.

These technologies offer the greatest power when combined: a candidate that looks excellent on GCI affinity but shows high PDI during DLS and a low T_m from DSC reveals developability liabilities that binding data alone would never surface. Catching these failures at microgram scale and early timelines, rather than at scale-up or manufacturing, offers a strong economic and strategic rationale for integrated biophysical screening in early biologics development.

Find out more about Malvern Panalytical’s complete biologics characterization workflow

Further reading:

• A week in the life of a structural biologist