Skip to content

Surface Properties

This interactive dashboard combines a surface properties grid with a 3D structural viewer.

This tool is useful for identifying and addressing surface properties liabilities in antibodies. AbLead calculates surface hydrophobicity, positive and negative charge patches, and overall charge dipole distributions using structural models. This tool allows for evaluation of the properties at the sequence, residue, and structure levels, and can be used to design for desired surface properties, if possible to do so while maintaining binding.

Accessing the Tool

Select exactly one antibody in the Project View which has an associated structure model. Go to the Analysis menu and select Surface Properties. This will open the Surface Properties workspace in a new tab.

This analysis gives residue-level details of SPH (Surface Hydrophobicity), SPP (Positive Patch Charge), SPN (Negative Patch Charge), and SPCD (Charge Dipole) scores. Evaluating these properties is critical for predicting developability liabilities such as aggregation, non-specific binding, increased clearance rates, and viscosity.

Select chain modes (e.g. VL Only or VH Only) to see how the surface properties change in an isolated chain format. Toggle select residue rows in the grids to highlight the residue on the 3D structure.

Surface Properties Dashboard

Engineering of anomalous surface regions has been shown to improve the developability of antibodies.

Highly exposed hydrophobic patches and dense surface charge patches are key drivers of poor antibody developability. By identifying these undesired regions, developers can design targeted mutations to optimize the surface profile:

  • Hydrophobic Patches (High SPH): Exposed hydrophobic clusters on the variable domain surface can drive self-association, aggregation, and accelerated clearance. Mutating exposed hydrophobic residues (e.g., Ile, Leu, Val, Phe) to hydrophilic residues (e.g., Ser, Thr) can dramatically improve solubility and biophysical stability without compromising binding affinity.
  • Charge Patches (High SPP/SPN): Large, concentrated positive (SPP) or negative (SPN) surface charge patches promote non-specific interactions with off-target tissues or serum components, leading to high clearance rates. Electronegative surfaces can lead to high viscosity at high concentrations. Neutralizing these patches by mutating charged residues to neutral amino acids or introducing opposing charges helps restore a balanced electrostatic surface. See Liu et al. concerning clearance and Agrawal et al. concerning viscosity.
  • Isolated Domain Liabilities: In single-chain or monomeric formats (such as scFvs or nanobodies), the variable domain interface is exposed to solvent. Interface residues are naturally hydrophobic and optimized for inter-chain packing, representing severe liabilities when exposed. Evaluating and engineering these isolated regions is critical to prevent aggregation of non-Fv formats.

Surface Property Metrics Table

The following table summarizes the surface property metrics computed by the AbLead internal engine, based on relative solvent accessibility and spatial neighborhoods.

Metric Name Severity Cutoff Calculation Basis Description
SPH Surface Hydrophobicity \(\ge 4.0\) (High/Red) Eisenberg Hydrophobicity Scale Measures the exposure of hydrophobic residues. High SPH indicates large hydrophobic clusters that may drive aggregation, self-association, or rapid clearance.
SPP Positive Charge Patch \(\ge 2.0\) (High/Red) Lys, Arg, His positive charges Evaluates the density of positively charged patches on the surface. Excessive positive charge density drives fast clearance and non-specific binding.
SPN Negative Charge Patch \(\ge 2.0\) (High/Red) Asp, Glu negative charges Evaluates the density of negatively charged patches on the surface. High negative patches can also influence solubility, pI, and off-target interaction profiles.
SPCD Charge Dipole N/A (Distribution metric) VL/VH exposed charge product Measures Fv charge distribution asymmetry. The overall SPCD is the product of net exposed VL and VH charges, representing global polarization rather than a local hotspot.

Stability and Property Cutoffs

Due to differences in biophysical impact and scale calibration, the recommended thresholds for identifying surface developability risks are:

  1. SPH (Surface Hydrophobicity): Recommended cutoff is 4.0 (Severe/Red). Residues exceeding this value are major hydrophobic hotspots and primary targets for solubilizing mutations.
  2. SPP (Positive Patch): Recommended cutoff is 2.0 (Severe/Red). Residues exceeding this value represent highly positive local environments that promote non-specific tissue binding.
  3. SPN (Negative Patch): Recommended cutoff is 2.0 (Severe/Red). Residues exceeding this value represent highly negative local environments.
  4. SPCD (Charge Dipole): Because SPCD is a global distribution metric reflecting inter-chain polarization, it does not have a discrete per-residue severity cutoff. It is used to analyze overall antibody dipole asymmetry.

The Calculation Method

Properties are calculated on the fly using structural coordinates:

  1. Structure Cleaning: Hydrogen atoms and heteroatoms (solvent water) are detached from the parsed PDB model.
  2. SASA Calculation: Solvent accessible surface area (SASA) is computed for all atoms using the Shrake-Rupley algorithm (probe radius 1.40 Å, Richards atomic radii).
  3. Exposure Cutoff: Residues are classified as exposed if their relative solvent accessibility (RSA) is at least \(7.5\%\) (\(\text{RSA} \ge 0.075\)).
  4. Salt Bridge Neutralization: If two exposed residues form a salt bridge (distance between opposite charge centers \(\le 4.0\) Å), their hydrophobic and charge contributions are neutralized (set to \(0.0\)).
  5. Score Accumulation: Per-residue scores sum the contributions of all other exposed neighbor residues within a spatial cutoff distance of 7.5 Å (heavy-atom closest approach) using a distance-dependent decay:
\[\text{Score}_{\text{res}} = \sum_{j \neq \text{res}} \frac{\text{Scale}(\text{res}) \times \text{Scale}(j)}{d_{\text{res}, j}^2}\]

where \(d_{\text{res}, j}\) is the closest heavy-atom distance, and \(\text{Scale}\) is the normalized hydrophobicity or charge property scale. Overall scores represent the sum of all individual residue-pair interactions. Note that SPCD uses a specific charge scale calibrated at pH 5.5 (setting Histidine to \(+0.9\)) to match standard FvCSP/SFvCSP definitions.

See Raybould et al. for details.

Recalculating for Single Chains

In the Fv complex, the variable light (VL) and variable heavy (VH) domains interface tightly, burying contact residues at the interface.

To evaluate developability properties in formats where a chain is isolated (such as single-chain Fvs or variable domain monomers), users can change the Chains Display setting to VL Only (Interface exposed) or VH Only (Interface exposed).

  • Isolated Calculations: The unselected chain is detached from a structure copy in the backend prior to executing the Shrake-Rupley SASA calculations. This exposes the interface residues, enabling their properties to be evaluated in their isolated state.
  • Main Structure Preservation: The unselected chain ribbon cartoon remains visible in the Molstar viewer (unless "Show Both Chains" is unchecked) to provide structural context, but its sidechains are not colored by property.
  • Dual Grid Presentation: Both VL and VH grids remain visible. The active chain displays its isolated recalculated properties, while the inactive chain continues to display its standard Fv baseline properties.

Using the Tool

  • Settings
    • Display Property: Select SPH, SPP, SPN, or SPCD to display.
    • Chains Display: Select Fv (VL & VH), VL Only (Interface exposed), or VH Only (Interface exposed) calculations.
    • Show Both Chains: Toggle whether to show the non-selected chain ribbon cartoon in the 3D viewer.
    • Select Severe: Selects residues exceeding developability hotspot risk thresholds (\(\text{SPH} \ge 4.0\), \(\text{SPP} \ge 2.0\), \(\text{SPN} \ge 2.0\)). Disabled for SPCD since SPCD is a distribution metric with no discrete residue-level severity scale.
    • Spacefill Selected: Toggles whether selected residues are shown as spacefill spheres in the 3D viewer.
    • Color Legends: Details the color scales and severity ranges.
  • Sequence Aligner Track: Visualizes variable domain sequences colored by the selected property scale. Rotated system numbering is shown above, and hovering over a residue displays a tooltip listing its SPH, SPP, SPN, and SPCD values alongside its 3D spatial neighbors.
  • Residue Data Grids: Detailed residue-level property tables for both the VL and VH domains are kept side-by-side at all times. Clicking a row highlights the residue in the 3D viewer, and clicking a residue letter in the aligner track scrolls the grid to focus on that position.
  • Indicator Dots: Dynamic mutation (squares) and observation (circles) indicators are rendered directly on the residue cells. Clicking a residue letter in the grids opens the Engineering / Observations modal to save or edit mutation designs, instantly refreshing the layout.
  • 3D Viewer & Neighborhood Highlighting: Displays the Fv structure ribbon backbone colored by region. Sidechain atoms for exposed or selected residues are shown colored by their property scores. Selecting any residue in the sequence or grid highlights the residue and all of its 3D spatial neighbors (heavy-atom distance \(\le 7.5\) Å) in the viewer as sidechain ball-and-stick representations. If Spacefill Selected is checked, selected residues are shown as spacefill spheres, and static labels formatted as [Neighbor <imgt>] <AA> are added to the neighbors.
  • Export Excel: Generates a single-worksheet spreadsheet containing mature linear numbering, residue regions (colored by region scheme), amino acid classes (colored by ABHAND), and property values. Zero-value cells (0.0) are left uncolored to reduce clutter, and the "3D Neighbors" list is left-aligned with text wrapping disabled.
  • Export PyMOL (.py): Generates a self-contained PyMOL Python script (.py) matching the currently selected Display Property (SPH, SPP, SPN, or SPCD) and Chains Display parameters. Executing the script inside PyMOL:
    • Creates and groups region sets (selections for CDRs and Frameworks) matching Liabilities.
    • Colors all cartoon and surface representations to the metric's baseline color.
    • Individually colors residues with active metric values by their property scores and displays their sidechains as sticks.
    • Renders a solid, opaque surface (transparency 0.0) over the active chain(s).
    • Hides all stick/line representations of hydrogen atoms while preserving their physical volume contribution to the calculated surface shape.

References

  • The PDB structure viewer is provided by Mol*.
  • The background inspiration for the Surface Properties (Tap-like method) is described in Raybould et al., PNAS (2019).
  • Liu, Shufang, Ashwni Verma, Hubert Kettenberger, Wolfgang F. Richter, and Dhaval K. Shah. “Effect of Variable Domain Charge on in Vitro and in Vivo Disposition of Monoclonal Antibodies.” mAbs 13, no. 1 (2021): 1993769.
  • Agrawal, Neeraj J., Bernhard Helk, Sandeep Kumar, et al. “Computational Tool for the Early Screening of Monoclonal Antibodies for Their Viscosities.” mAbs 0862, no. December (2015): 1–6.