Auto-CHO

Auto-CHO generates potential synthetic pathways for hierarchical, programmable one-pot oligosaccharide synthesis by arranging thioglycoside building blocks (BBLs) according to relative reactivity values (RRVs) predicted with support vector regression to support planning of complex glycan assembly.

Key Features:

• Pathway generation: Generates candidate synthetic pathways for oligosaccharide assembly using a programmable one-pot approach.
• Thioglycoside building blocks (BBLs): Uses thioglycoside BBLs as the modular units for synthesis planning.
• Relative Reactivity Values (RRVs): Orders BBLs by RRVs to guide reaction sequence selection.
• RRV prediction (support vector regression): Predicts RRVs using support vector regression models.
• BBL library: Leverages a library comprising over 150 validated and more than 50,000 virtual BBL entries.
• Hierarchical one-pot algorithm: Implements a hierarchical one-pot synthesis algorithm that structures multi-step assembly.
• Iterative fragment repurposing: Reuses fragments generated in initial reactions as new BBLs for subsequent assembly steps.
• Feedback mechanism for virtual BBL evaluation: Incorporates feedback to evaluate and retain valuable virtual BBLs for future use.

Scientific Applications:

• Oligosaccharide synthesis planning: Plans hierarchical one-pot syntheses for complex oligosaccharides.
• Scale-up synthesis: Supports strategies aimed at synthesizing large quantities of oligosaccharides.
• Synthesis of SSEA-4: Applied to the one-pot synthesis of stage-specific embryonic antigen 4 (SSEA-4).
• Complex glycan research: Enables studies in developmental biology and regenerative medicine involving complex glycans.

Methodology:

Generates synthetic pathways using a hierarchical one-pot algorithm that arranges thioglycoside BBLs by RRVs predicted via support vector regression, utilizes a library of >150 validated and >50,000 virtual BBLs, repurposes reaction fragments as new BBLs, and applies feedback evaluation to virtual BBLs.