dGPredictor

dGPredictor predicts standard Gibbs free energy changes (ΔrG'°) of enzymatic reactions using automated molecular fingerprints to incorporate stereochemistry and improve thermodynamic analysis for de novo metabolic pathway design.

Key Features:

• Automated molecular fingerprinting: Uses automated molecular fingerprints to capture detailed structural information, including stereochemistry, for metabolites.
• Stereochemistry-aware predictions: Incorporates stereochemical information to increase reaction coverage and resolve cases where traditional group contribution (GC) methods fail.
• Improved reaction coverage and accuracy: Extends prediction capability to reactions with no net group changes, including isomerases and transferases, yielding higher ΔrG'° prediction accuracy.
• Novel reaction prediction: Estimates ΔrG'° for novel, previously uncharacterized reactions to support synthetic biology applications.
• Integration with pathway design tools: Integrates with de novo metabolic pathway design tools such as novoStoic to enable thermodynamically informed pathway construction.
• Support for identifier and structure inputs: Accepts metabolite identifiers and structures such as KEGG IDs and InChI strings for thermodynamic estimation.
• Condition-specific estimates: Provides ΔrG'° predictions across specified pH values and ionic strengths.

Scientific Applications:

• Genome-scale metabolic network curation: Provides thermodynamic feasibility assessments for reactions within genome-scale metabolic models.
• De novo pathway design and pruning: Enables identification and exclusion of pathway steps with thermodynamically infeasible directionalities during pathway construction.
• Synthetic biology and metabolic engineering: Supports design and evaluation of novel biosynthetic routes by predicting ΔrG'° for uncharacterized reactions.
• Improved analysis of isomerases and transferases: Enhances prediction reliability for enzyme classes where traditional GC methods have low coverage.

Methodology:

Estimates standard Gibbs free energy changes (ΔrG'°) using an automated molecular fingerprint-based approach that captures metabolite structural detail, including stereochemistry, in contrast to manually curated group contribution methods.