AbTrimmer

1 Introduction

Antibody drug developability risk assessment and druggability analysis are critical steps in the drug discovery pipeline, aiming to identify promising clinical candidates early in the development process while mitigating potential risks. Building upon previous work (TAP tool)^1,2, we developed AbTrimmer, a computational tool that evaluates antibody drug development risks based on multiple biophysical parameters, including Patches of Surface Hydrophobicity (PSH), Patches of Surface Positive Charge (PPC), Patches of Surface Negative Charge (PNC), Structural Fv charge symmetry parameter (SFvCSP), and aggregation scores⁶. By precisely quantifying antibody features such as hydrophobicity and charge distribution, and comparing against clinically validated or marketed therapeutic antibodies, AbTrimmer enables comprehensive risk assessment of antibody molecules.

Prior to calculating these parameters, we established the following definitions:

• Solvent-accessible surface area (SASA) of each residue was calculated using the Shrake and Rupley algorithm with a spherical probe radius of 1.4 Å. Residues with SASA ≥ 7.5% of their theoretical maximum were considered surface-exposed^3,4.
• CDR-proximal regions were defined as: surface-exposed residues within 4.5 Å distance from CDR regions under IMGT numbering scheme, plus two residues immediately preceding and following each CDR region.
• Charge assignments for amino acids were as follows: D=-1, E=-1, K=+1, R=+1, H=+0.1, while all other amino acids were assigned 0. Residues participating in salt bridges also received a charge value of 0.
• Amino acid hydrophobicity was determined using the Kyte&Doolittle hydrophobicity scale⁵, with values normalized to the range of 1 to 2.

After using ImmuneBuilder2⁷ to predict the structure of the antibody, calculate the PSH, PPC, PNC and SFvCSP as follows:

\begin{aligned} \text{PPC} &= \sum_{i,j \in R} \frac{Q_i + Q_j}{d_{ij}} (Q_i > 0, Q_j > 0, d_{ij}<8\AA) \\ \text{PNC} &= \sum_{i,j \in R} \frac{Q_i + Q_j}{d_{ij}} (Q_i < 0, Q_j < 0, d_{ij}<8\AA) \\ \text{PSH} &= \sum_{i,j \in R} \frac{H_i \cdot H_j}{d_{ij}} (d_{ij}<8\AA)\\ \text{SFvCSP} &= \sum_{i \in \mathcal{H}_{\text{exposed}}} Q_i \cdot \sum_{j \in \mathcal{L}_{\text{exposed}}} Q_j \end{aligned}

Where:

• R represents CDR-proximal residues.
• Q_i denotes the charge index of the i^{th} amino acid.
• H_i indicates the hydrophobicity index of the i^{th} amino acid.
• d_{ij} is the Euclidean distance between Cα atoms of residue i and residue j.
• \mathcal{H}_{\text{exposed}} and \mathcal{L}_{\text{exposed}} refer to heavy chain surface-exposed residues and light chain surface-exposed residues, respectively.
• SFvCSP (Structural Fv charge symmetry parameter) is not calculated for nanobodies

Additionally, we employ CANYA⁶, the current state-of-the-art AI-based aggregation prediction model for antibody developability assessment. The performance of CANYA is detailed in Figure 1.

Figure 1. the performance of CANYA Model.

We collected 1050 therapeutic antibodies and 41 therapeutic nanobodies from the SAbDab database, statistically analyzed the numerical values of various metrics, and visualized their distributions using histograms. Taking the PSH score of therapeutic antibodies as an example, the numerical distribution is shown in the following figure:

Figture 2. Therapeutic antibody PSH score histogram.

2 Parameters

Input: Heavy chain and light chain sequences of antibodies, or nanobody sequences.

3 Results Explanation

The result includes a .csv file and an image. The .csv file contains the input antibody sequences along with the values of various metrics predicted by AbTrimmer. The image on the results page marks the 5th and 95th percentiles. Please refer to the table below for the specific risk ranges of each metric.

Antibody:

Nanobody:

4 Reference