Catalyst design in asymmetric reaction development has traditionally been driven by empiricism, wherein experimentalists attempt to qualitatively recognize structural patterns to improve selectivity. Machine learning algorithms and chemoinformatics can potentially accelerate this process by recognizing otherwise inscrutable patterns in large datasets. Herein we report a computationally guided workflow for chiral catalyst selection using chemoinformatics at every stage of development. Robust molecular descriptors that are agnostic to the catalyst scaffold allow for selection of a universal training set on the basis of steric and electronic properties. This set can be used to train machine learning methods to make highly accurate predictive models over a broad range of selectivity space. Using support vector machines and deep feed-forward neural networks, we demonstrate accurate predictive modeling in the chiral phosphoric acid-catalyzed thiol addition to N-acylimines.

A drug-screening system includes an encoding module, a candidate-drug generating module and a drug-ranking module. The encoding module is configured to encode a drug expression and at least one drug-ranking indicator to generate a first encoding variable. The candidate-drug generating module is configured to train a generative adversarial network according to the first encoding variable to generate a plurality of candidate drugs, wherein each of the candidate drugs has a generative drug expression and at least one generative drug-ranking indicator. The drug-ranking module is configured to rank strengths of the candidate drugs according to the generative drug-ranking indicator of each of the candidate drugs.

A system, device, and method for predicting a docked position of a target ligand in a binding site of a biomolecule is disclosed. The prediction makes use of a template ligand-biomolecule complex structure in order to predict a target ligand-biomolecule complex structure. The system and device contain modules allowing for the prediction of a target-ligand biomolecule complex structure. A preparation module can receive information identifying a target ligand and a template ligand-biomolecule structure. A pharmacophore matcher module can identify common pharmacophores between the template ligand and the target ligand. A docking module can predict a docked ligand position of the target ligand by overlapping the pharmacophore models of the target ligand and template ligand while the template ligand is in the binding site of the biomolecule. A biomolecule modification module can modify the biomolecule to reduce clashes between the docked target ligand and the biomolecule.

A system, device, and method for predicting a docked position of a target ligand in a binding site of a biomolecule is disclosed. The prediction makes use of a template ligand-biomolecule complex structure in order to predict a target ligand-biomolecule complex structure. The system and device contain modules allowing for the prediction of a target-ligand biomolecule complex structure. A preparation module can receive information identifying a target ligand and a template ligand-biomolecule structure. A pharmacophore matcher module can identify common pharmacophores between the template ligand and the target ligand. A docking module can predict a docked ligand position of the target ligand by overlapping the pharmacophore models of the target ligand and template ligand while the template ligand is in the binding site of the biomolecule. A biomolecule modification module can modify the biomolecule to reduce clashes between the docked target ligand and the biomolecule.

Lack of safety and efficacy are the two major unwanted biological responses that play as critical bottlenecks for the success of drug candidates in drug discovery and development. Conventional systems and methods involve ineffective exploration and use of chemical information space and thereby, may fail to address safety and efficacy issues. Embodiments of the present disclosure provides an effective solution to the above bottle-necks with the effective exploration/search of chemical information space using effective statistical techniques that yield meaningful chemical information comprising relevant descriptors, fingerprints, fragments, optimized set of structural images, and the like. Further, it provides robust predictive models for the biological response, example renal toxicity using the selected chemical information in an automated manner for a given experimental data and alerts/rules that can be successfully employed to address failures of drug candidates during discovery and development.

Lack of safety and efficacy are the two major unwanted biological responses that play as critical bottlenecks for the success of drug candidates in drug discovery and development. Conventional systems and methods involve ineffective exploration and use of chemical information space and thereby, may fail to address safety and efficacy issues. Embodiments of the present disclosure provides an effective solution to the above bottle-necks with the effective exploration/search of chemical information space using effective statistical techniques that yield meaningful chemical information comprising relevant descriptors, fingerprints, fragments, optimized set of structural images, and the like. Further, it provides robust predictive models for the biological response, example renal toxicity using the selected chemical information in an automated manner for a given experimental data and alerts/rules that can be successfully employed to address failures of drug candidates during discovery and development.

The present application is directed to methods and systems for identifying small molecule compounds in mixtures using a library comprising calculated structures and corresponding calculated mass spectral fragmentation patterns of known and/or hypothetical small molecule compounds that may be in the mixture and screening of a mass spectrum of the mixture using the library to identify matching fragmentation patterns. If a mass spectral fragmentation pattern present in the mass spectrum of the mixture matches a calculated fragmentation pattern of one of the known or hypothetical compounds this confirms the identity of a compound in the mixture as the known or hypothetical compound. The method represents a platform method that can be used for a multitude of purposes related to the screening and identification of compounds in mixtures. Therefore the methods and systems of the present application represent an approach that is uniquely capable of navigating chemical space and providing a understanding of desired families and pharmacophores.

A library of known intrinsic spectra is provided to identify at least one known material in a sample of interest. The library includes individual intrinsic spectra channels defined by the assignment of intrinsic spectra of at least one known material, and combinations thereof, so that the assigned intrinsic spectra of each intrinsic spectra channel is correlated to at least one known material. The at least one known material is identified in the sample of interest when intrinsic spectra obtained from the sample of interest is matched to an assigned intrinsic spectra of an intrinsic spectra channel of the library of known intrinsic spectra.

A library of known intrinsic spectra is provided to identify at least one known material in a sample of interest. The library includes individual intrinsic spectra channels defined by the assignment of intrinsic spectra of at least one known material, and combinations thereof, so that the assigned intrinsic spectra of each intrinsic spectra channel is correlated to at least one known material. The at least one known material is identified in the sample of interest when intrinsic spectra obtained from the sample of interest is matched to an assigned intrinsic spectra of an intrinsic spectra channel of the library of known intrinsic spectra.

A method includes receiving one input feature and one output feature of importance of polymers used to stabilize a protein. A set of polymers from a library are identified based on the input feature and the output feature of importance that are applied to a machine learning model. Data for each polymer in the library includes features for each polymer and reagents for stabilizing the protein. Each polymer in the identified set is used to stabilize samples of the protein in well plates in a well plate array based on the reagents from the library data in the identified set. A score for each sample of the protein is assigned by comparing the measured output feature from the well plates corresponding to the identified set to the output feature of importance. The samples of the protein are identified having scores higher than a predefined threshold.