Methods and apparatuses for a unified artificial intelligence platform to synthesize diverse sets of peptides and peptidomimetics
Inventors
Lee, Francis • STECKBECK, Jonathan D. • Holste, Hannes
Assignees
Publication Number
US-11587643-B2
Publication Date
2023-02-21
Expiration Date
2041-08-13
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
In one aspect, a method is disclosed wherein an artificial intelligence (AI) enabled automated flow synthesis platform is configured to generate optimized synthesizing recipes which enable a sequence to be synthesized using an automated flow process. The method includes receiving a synthesizing recipe including parameters used during the automated flow process to synthesize the sequence, receiving spectral data from detectors monitoring the automated flow process in a reaction chamber, where the spectral data corresponds to a reaction point in the automated flow process, and determining, based on indicators associated with the spectral data, characteristics of a chemical reaction at the reaction point in the automated flow process. An artificial intelligence engine determines the chemical reaction. The method includes associating, based on the spectral data, the synthesizing recipe with the chemical reaction.
Core Innovation
The invention relates to an artificial intelligence (AI)-enabled automated flow synthesis platform designed to generate optimized synthesizing recipes that enable protein sequences, such as peptides and peptidomimetics, to be synthesized using an automated flow process. The system comprises an AI engine with machine learning models, including a creator module, trained to receive vector representations of heterogeneous biological context networks. Using compressed vector representations and causal inference with counterfactuals, the AI engine efficiently generates protein sequences based on desired drug activity levels.
The disclosed methods enable monitoring of the synthesis process in a reaction chamber by receiving spectral data from detectors that correspond to reaction points where amide coupling occurs. The AI engine determines the characteristics of these chemical reactions and associates synthesizing recipes with the observed chemical reactions based on the spectral data. The platform can adapt synthesizing parameters in response to real-time data, generate spectral profiles, and filter synthesizing recipes using statistical, probabilistic, or arithmetical differences to achieve desired reaction outcomes.
The invention aims to address the inefficiency, limited application, and complexity associated with conventional drug discovery and synthesis techniques, which often rely on narrow design spaces, require substantial resources, and are unable to efficiently optimize for a broad range of desired therapeutic effects. By leveraging machine learning and AI-driven optimization in flow chemistry, the platform enables scalable, accurate, and resource-efficient protein sequence synthesis with real-time feedback and adaptability.
Claims Coverage
The independent claims encompass multiple inventive features related to an AI-driven automated flow synthesis platform for generating and optimizing synthesizing recipes for protein sequence production.
AI-enabled flow synthesis platform for generating synthesizing recipes
The platform integrates an AI engine with trained machine learning models that receive vector representations of heterogeneous biological context networks and generate, using causal inference and counterfactuals, protein sequences based on desired drug activity levels. The AI engine compresses vector representations from higher-order to lower-order dimensions to reduce processing complexity.
Automated real-time monitoring and association of synthesis parameters
The system monitors the automated flow synthesis process in a reaction chamber using detectors that capture spectral data corresponding to specific reaction points. The AI engine determines characteristics of the chemical reactions at these points and associates the synthesizing recipe with the chemical reaction based on the spectral data.
Adaptive optimization and filtering of synthesizing recipes
The platform enables the determination and implementation of subsequent synthesizing recipes based on correlations between recipes and observed chemical reactions. The AI engine further filters and selects synthesizing recipes using statistical, probabilistic, percentage, or arithmetical differences to optimize chemical outcomes for effective protein sequence synthesis.
The claims establish a comprehensive system that integrates machine learning, real-time spectroscopic monitoring, causal inference, and adaptive optimization to automate and improve the synthesis of protein sequences via flow chemistry.
Stated Advantages
The platform enables scalable, efficient, and accurate generation of protein sequences over a broad design space, overcoming limitations of conventional drug discovery techniques.
Real-time monitoring and adaptive control reduce resource consumption, processing time, and material waste during synthesis.
Optimized synthesizing recipes minimize undesired side reactions and improve yield and quality of synthesized protein sequences.
The AI-driven approach reduces computational complexity and processing burden, enabling the discovery and synthesis of therapeutics with desired activity profiles.
Documented Applications
Therapeutics development, including anti-infective, anti-cancer, antimicrobial, anti-viral, anti-fungal, anti-inflammatory, anti-cholinergic, anti-dopaminergic, anti-serotonergic, anti-noradrenergic, and anti-prionic effectiveness.
Peptide and peptidomimetic synthesis for therapeutic and biomedical applications.
Synthesis of protein sequences for drug discovery based on biological activity requirements.
Interested in licensing this patent?