Artificial neural networks for guest chirality classification through supramolecular interactions

Area of Science:

• Supramolecular Chemistry
• Analytical Chemistry
• Computational Chemistry

Background:

• Chirality recognition is crucial in pharmaceutical and biological applications.
• Developing efficient methods for determining enantiomeric purity remains a challenge.
• Anion-receptor chemistry offers a platform for molecular recognition and sensing.

Purpose of the Study:

• To report a novel strategy for classifying guest chirality using artificial neural networks (ANNs) combined with anion-receptor chemistry.
• To demonstrate the capability of the developed system in recognizing the stereochemistry of biologically relevant carboxylate guests.
• To explore the potential of NMR spectroscopy and chemometrics in chiral analysis.

Main Methods:

• Design and synthesis of a novel anion receptor capable of forming supramolecular complexes with carboxylates.
• Utilizing proton nuclear magnetic resonance ((1)H NMR) spectroscopy to monitor chemical shift changes upon anion binding.
• Training an artificial neural network on a dataset of known anion-guest complexes to recognize stereochemistry-dependent patterns.
• Applying principal component analysis (PCA) for data discrimination and identification of key receptor protons involved in chirality transfer.

Main Results:

• The anion receptor formed distinct supramolecular complexes with various carboxylate guests.
• (1)H NMR spectroscopy revealed complex guest-stereochemistry-dependent chemical shift patterns.
• The trained neural network successfully identified the unknown chirality of 14 out of 14 guests in the test set.
• Principal component analysis discriminated between 26 studied guests and identified specific receptor protons crucial for chirality information transfer.

Conclusions:

• The combination of artificial neural networks and anion-receptor chemistry provides a powerful and novel strategy for guest chirality classification.
• This method offers a sensitive and accurate approach for determining the stereochemistry of biologically important carboxylates.
• The study highlights the potential of integrating advanced computational methods with supramolecular chemistry for chiral sensing applications.