We use cookies to enhance your browsing experience and analyze site traffic. By continuing to use this site, you agree to our cookie policy. Privacy Policy

CDR Diversity Engineering

Our Approach

An antibody's specificity and affinity are largely determined by its complementarity-determining regions (CDRs). Through structure-guided, precise CDR diversification, we systematically explore the sequence space while preserving the antibody's structural stability and affinity—capturing a broader sequence diversity of same-function antibodies with smaller, higher-quality libraries. The system offers sequence- and structure-based dual-branch workflows, and uses energy scoring plus position-level protection strategies to deliver a library-ready, high-quality candidate set.

Key Features

Structure-Guided
Design Strategy
Dual-Branch
Workflow
4+1
Numbering Schemes
2h
Computation Time

Why Is This Needed?

"Traditional CDR diversification relies on random mutagenesis or semi-rational design: random mutations generate many ineffective variants, burying truly valuable sequences among countless candidates; mutations without structural constraints easily disrupt antibody folding, causing poor expression or aggregation; and boosting hit rates often requires screening libraries of millions or even billions, making cost and timeline the bottleneck. What teams need is a smaller, higher-quality, structurally sound candidate library."

Key Advantages

Key Advantages

Traditional Limitations

Inefficient Sequence-Space Exploration

Random mutations generate large numbers of ineffective variants, burying the truly valuable sequences among countless candidates.

Hard to Ensure Structural Stability

Mutations without structural constraints easily disrupt antibody folding, leading to poor expression or aggregation.

Library Scale vs. Quality Trade-off

Boosting hit rates requires screening libraries of millions or even billions, making cost and timeline the bottleneck.

Our Advantages

Structure-Guided Mutation Design

Evaluates the impact of each mutation at the 3D protein-structure level, automatically identifying key antigen-contact residues and structural support sites, and quantifying mutation effects to narrow the search space.

Intelligent Sequence-Space Sampling

Directed sampling based on physicochemical properties and evolutionary information; supports position-level protection (avoid/include) and configurable N/C-terminal extensions to precisely control mutation positions and CDR boundaries.

Multi-Scheme Numbering Support

Supports the four mainstream numbering schemes—Kabat, Chothia, IMGT, AHO—plus the ALL mode (union of all four), ensuring compatibility with diverse databases and literature.

Dual-Branch Workflow

The sequence branch runs structure prediction and mutation design directly from sequences with no manual modeling; the structure branch leverages an existing complex structure for more precise energy calculations and site-specific analysis.

How It Works

How It Works

Two branches—sequence and structure—auto-identify CDR regions, predict/refine structures, then drive mutation design and ranking with energy scoring to deliver a library-ready, high-quality candidate set.

Sequence Branch

Input antibody and antigen sequences → structure prediction → energy scoring → mutation design. Ideal for early-stage projects with sequence information only—fully automated, no manual modeling required.

Structure Branch

Input an antibody–antigen complex structure → structure refinement and analysis → saturation mutagenesis scan → precise scoring. Suited to projects with a crystal structure or high-quality model, providing more accurate energy calculations and site-specific analysis.

Input Sequence / Structure
CDR Region Identification
Structure Prediction / Refinement
Mutation Design & Ranking

Input sequence / structure → multi-scheme CDR auto-annotation (with protection strategy) → structure prediction / refinement → energy-scoring-driven mutation design → Top candidate sequences and ranking per CDR

Report

Report Example

The following screenshots are excerpts from a CDR Diversity Design report, demonstrating the full computational workflow from parental sequence characterization to CDR mutation design and candidate sequence evaluation.

Use Cases

Use Cases and Deliverables

Best suited when you have an affinity-qualified lead antibody and want to explore sequence diversity through CDR optimization ahead of library screening.

2 hours12,888 credits / antibody / run

Best-Fit Scenarios

Explore the sequence-diversity space while preserving function
Discover candidate molecules with differentiated sequence features
Optimize developability metrics (expression, aggregation tendency, thermostability, etc.)
Support early-stage molecule design and screening for me-better and biosimilar projects

Inputs to Prepare

Sequence mode: heavy (VH), light (VL), and antigen sequences (single or multiple antigen chains)
Structure mode: a PDB antibody–antigen complex structure and chain IDs
The chosen numbering scheme and position-level protection strategy

What You Receive

Theoretical library scale for each CDR region
Structural-perturbation scores and preferred mutation strategies
A basis for downstream antibody optimization and experimental library construction

How is CDR diversification different from affinity maturation?

Affinity maturation focuses on strengthening binding to a specific antigen, typically via saturation mutagenesis and ML-based mutation-effect prediction. CDR diversification focuses on exploring sequence diversity, generating a diverse variant library while preserving structural stability. The two are complementary: use CDR diversification to build the initial library, then affinity maturation to optimize the hits.

Why is a structure-based approach better than random mutagenesis?

Antibody folding follows strict physicochemical rules, and random mutations are likely to break them, causing misfolding or aggregation. Structure-guided design uses energy calculations to pre-exclude these "dead-end" sequences, ensuring every variant in the library has a reasonable chance of folding correctly—substantially improving screening efficiency.

What library scale is supported?

A single run can generate hundreds to thousands of candidate CDR sequences (Top 5–100 candidates per CDR; the combinatorial theoretical diversity across 6 CDRs can reach millions or more). The actual library size depends on the sequence-identity parameters and protection strategy. For wet-lab screening, we recommend combining the Top candidates of each CDR to build an appropriately sized library on demand.

What is the runtime and pricing?

~2 hours per run, priced at 12,888 credits / antibody / run.

Build Smaller, Better, High-Quality Libraries with Structure-Guided CDR Diversification

Submit sequences or a complex structure, choose the numbering scheme and protection strategy, and receive a structurally sound, library-ready high-quality CDR candidate set for screening.