IP Library › Granted Patent US 12,462,902
Granted Patent B2
US 12,462,902 · App. 17/339,520 · Granted Nov 4, 2025

Artificial intelligence engine architecture for generating candidate drugs

Inventors: Francis Lee (Cambridge, MA); Jonathan D. Steckbeck (Cranberry Township, PA); Hannes Holste (Los Angeles, CA)
Assignee: Peptilogics, Inc.
G16C20/70G06N3/042G06N3/044G06N3/045G06N3/08G06N20/00G16C20/50G16C60/00
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,462,902
App. No.
17/339,520
Filed
Jun 4, 2021
Granted
Nov 4, 2025
Kind
B2
Art Unit
1686
USPC
703/11
Abstract

A method is disclosed for using an artificial intelligence engine to generate candidate drug compounds, wherein the method comprises: generating candidate drug compounds comprising sequences via a creator module of the artificial intelligence engine. The method includes generating, via a descriptor module, a respective description for each of the candidate drug compounds at nodes in a knowledge graph, wherein the knowledge graph comprises a multi-dimensional representation of the candidate drug compounds and the respective description comprises drug compound structural information, drug compound activity information, and drug compound semantic information. The method includes determining a shape of the multi-dimensional representation of the candidate drug compounds; determining, based on the shape, a slice configured to be obtained from the representation; determining, using a decoder, which dimensions are included in the slice; and based on the dimensions, determining an effectiveness of a biomedical feature of the slice.

Claims (94)

1 . A computer-implemented method for using an artificial intelligence engine to generate candidate drug compounds, wherein the computer-implemented method comprises:

generating a multi-dimensional representation of a plurality of protein drug compounds, wherein for each protein drug compound of the plurality of protein drug compounds, the multi-dimensional representation includes one or more relationships between protein drug compound structural information, protein drug compound activity information, and protein drug compound semantic information;

translating, by the artificial intelligence engine, the multi-dimensional representation to a plurality of encodings;

concatenating, by the artificial intelligence engine, a plurality of encodings to form a concatenated vector, wherein:

the encodings are each respective protein sequences represented in a vector,

a first encoding of the plurality of encodings pertains to the protein drug compound structural information,

a second encoding of the plurality of encodings pertains to the protein drug compound activity information, and

a third encoding of the plurality of encodings pertains to the protein drug compound semantic information;

using an autoencoder of the artificial intelligence engine to compress the concatenated vector from a higher-dimensional vector to a lower-dimensional vector, wherein compressing the concatenated vector from the higher-dimensional vector to the lower-dimensional vector reduces processing complexity;

generating, using the lower-dimensional vector, a candidate drug compound comprising a protein sequence via a creator module of the artificial intelligence engine, wherein the artificial intelligence engine is executed by one or more processing devices;

determining, using a decoder of the artificial intelligence engine, which dimensions are included in the candidate drug compound by converting the candidate drug compound to the higher-dimensional vector and obtaining a set of coordinates from the higher-dimensional vector, wherein the coordinates represent the protein drug compound structural information, the protein drug compound activity information, the protein drug compound semantic information, or some combination thereof; and

based on the dimensions, determining, by the artificial intelligence engine, an effectiveness of a biomedical feature of the candidate drug compound, wherein the creator module comprises a generator machine learning model and a discriminator machine learning model, wherein:

the generator machine learning model is trained to receive the lower-dimensional vector and to generate, based on a counterfactual comprising a specification for modifying a different candidate drug compound, the candidate drug compound, and

the discriminator machine learning model is trained to receive the candidate drug compound as input and to predict, based on biomedical activity data pertaining to the plurality of protein drug compounds, an output related to the effectiveness of the biomedical feature which the candidate drug compound provides; and

using at least the output related to the effectiveness of the biomedical feature which the candidate drug compound provides, training, by the one or more processing devices, the generator machine learning model to remove the candidate drug compound from consideration, when the consideration occurs during an iteration of a subsequent generation.

2 . The computer-implemented method of claim 1 , further comprising:

including the candidate drug compound as a node in a knowledge graph; and

generating, via a descriptor module of the artificial intelligence engine, a description for the candidate drug compound at a node in the knowledge graph, wherein the knowledge graph comprises the multi-dimensional representation and the respective description comprises the drug compound structural information, the drug compound activity information, and the drug compound semantic information.

3 . The computer-implemented method of claim 2 , further comprising, based on the description, performing, via a scientist module of the artificial intelligence engine, at least one benchmark analysis of a parameter of the creator module, wherein the scientist module comprises a machine learning model trained to perform the benchmark analysis of the parameter of the creator module, wherein the at least one benchmark analysis comprises assigning a score to the parameter of the creator module, and the parameter relates at least to an ability of the creator module to generate candidate drug compounds.

4 . The computer-implemented method of claim 3 , further comprising tuning, based on the score of the parameter, an attribute of the creator module.

5 . The computer-implemented method of claim 4 , wherein the attribute comprises a weight, an activation function, a hidden layer number, a loss function, or some combination thereof.

6 . The computer-implemented method of claim 3 , wherein the parameter comprises a validity of the candidate drug compound, uniqueness of the candidate drug compound, a novelty of the candidate drug compound, a similarity of the candidate drug compound to other candidate drug compounds, or some combination thereof, and

the computer-implemented method further comprises ranking a plurality of creator modules based on the score, wherein the plurality of creator modules comprises the creator module.

7 . The computer-implemented method of claim 1 , further comprising training, using at least the effectiveness of the biomedical activity which the candidate drug compound provides, by the one or more processing devices, a machine learning model to remove at least one candidate drug compound from consideration in an iteration of a subsequent generation.

8 . The computer-implemented method of claim 1 , wherein the determining, using the decoder by the artificial intelligence engine, which of the dimensions are included in the candidate drug compound further comprises:

transforming the candidate drug compound from the lower-dimensional vector to the higher-dimensional vector.

9 . The computer-implemented method of claim 8 , wherein the decoder transforms the candidate drug compound by obtaining a set of coordinates from the higher-dimensional vector.

10 . The computer-implemented method of claim 9 , wherein the set of coordinates may be used in a back-calculation operation to determine whether the dimensions pertain to the drug compound structural information, the drug compound activity information, the drug compound semantic information, or some combination thereof.

11 . A system for using an artificial intelligence engine to generate candidate drug compounds, wherein the system comprises:

a memory device storing computer instructions;

a processing device communicatively coupled to the memory device, wherein the processing device executes the computer instructions to:

generate a multi-dimensional representation of a plurality of protein drug compounds, wherein for each protein drug compound of the plurality of protein drug compounds, the multi-dimensional representation includes one or more relationships between protein drug compound structural information, protein drug compound activity information, and protein drug compound semantic information;

translate, by the artificial intelligence engine, the multi-dimensional representation to a plurality of encodings;

concatenate, by the artificial intelligence engine, a plurality of encodings to form a concatenated vector, wherein:

the encodings are each respective protein sequences represented in a vector,

a first encoding of the plurality of encodings pertains to the protein drug compound structural information,

a second encoding of the plurality of encodings pertains to the protein drug compound activity information, and

a third encoding of the plurality of encodings pertains to the protein drug compound semantic information;

use an autoencoder of the artificial intelligence engine to compress the concatenated vector from a higher-dimensional vector to a lower-dimensional vector, wherein compressing the concatenated vector from the higher-dimensional vector to the lower-dimensional vector reduces processing complexity;

generate, using the lower-dimensional vector, a candidate drug compound comprising a protein sequence via a creator module of the artificial intelligence engine, wherein the artificial intelligence engine is executed by one or more processing devices;

determine, using a decoder of the artificial intelligence engine, which dimensions are included in the candidate drug compound by converting the candidate drug compound to the higher-dimensional vector and obtaining a set of coordinates from the higher-dimensional vector, wherein the coordinates represent dimensions pertain to the protein drug compound structural information, the protein drug compound activity information, the protein drug compound semantic information, or some combination thereof; and

based on the dimensions, determine, by the artificial intelligence engine, an effectiveness of a biomedical feature of the candidate drug compound, wherein the creator module comprises a generator machine learning model and a discriminator machine learning model, wherein:

the generator machine learning model is trained to receive the lower-dimensional vector and to generate, based on a counterfactual comprising a specification for modifying a different candidate drug compound, the candidate drug compound, and

the discriminator machine learning model is trained to receive the candidate drug compound as input and to predict, based on biomedical activity data pertaining to the plurality of protein drug compounds, an output related to the effectiveness of the biomedical feature which the candidate drug compound provides; and

use at least the output related to the effectiveness of the biomedical feature which the candidate drug compound provides, training, by the one or more processing devices, the generator machine learning model to remove the candidate drug compound from consideration, when the consideration occurs during an iteration of a subsequent generation.

12 . The system of claim 11 , wherein the processing device is further to:

include the candidate drug compound as a node in a knowledge graph; and

generate, via a descriptor module of the artificial intelligence engine, a description for the candidate drug compound at a node in the knowledge graph, wherein the knowledge graph comprises the multi-dimensional representation and the respective description comprises the drug compound structural information, the drug compound activity information, and the drug compound semantic information.

13 . The system of claim 12 , wherein the processing device is further to, based on the description, perform, via a scientist module of the artificial intelligence engine, at least one benchmark analysis of a parameter of the creator module, wherein the scientist module comprises a machine learning model trained to perform the benchmark analysis of the parameter of the creator module, wherein the at least one benchmark analysis comprises assigning a score to the parameter of the creator module, and the parameter relates at least to an ability of the creator module to generate candidate drug compounds.

14 . The system of claim 13 , further comprising tuning, based on the score of the parameter, an attribute of the creator module.

15 . The system of claim 14 , wherein the attribute comprises a weight, an activation function, a hidden layer number, a loss function, or some combination thereof.

16 . The system of claim 13 , wherein the parameter comprises a validity of the candidate drug compound, uniqueness of the candidate drug compound, a novelty of the candidate drug compound, a similarity of the candidate drug compound to other candidate drug compounds, or some combination thereof; and

the processing device is further configured to rank a plurality of creator modules based on the score, wherein the plurality of creator modules comprises the creator module.

17 . The system of claim 11 , wherein the processing device is further configured to train, using at least the effectiveness of the biomedical activity which the candidate drug compound provides, by the one or more processing devices, a machine learning model to remove at least one candidate drug compound from consideration in an iteration of a subsequent generation.

18 . A tangible, non-transitory computer-readable medium storing instructions that, when executed, cause a processing device to:

generate a multi-dimensional representation of a plurality of protein drug compounds, wherein for each protein drug compound of the plurality of protein drug compounds, the multi-dimensional representation includes one or more relationships between protein drug compound structural information, protein drug compound activity information, and protein drug compound semantic information;

translate, by the artificial intelligence engine, the multi-dimensional representation to a plurality of encodings;

concatenate, by the artificial intelligence engine, a plurality of encodings to form a concatenated vector, wherein:

the encodings are each respective protein sequences represented in a vector,

a first encoding of the plurality of encodings pertains to the protein drug compound structural information,

a second encoding of the plurality of encodings pertains to the protein drug compound activity information, and

a third encoding of the plurality of encodings pertains to the protein drug compound semantic information;

use an autoencoder of the artificial intelligence engine to compress the concatenated vector from a higher-dimensional vector to a lower-dimensional vector, wherein compressing the concatenated vector from the higher-dimensional vector to the lower-dimensional vector reduces processing complexity;

generate, using the lower-dimensional vector, a candidate drug compound comprising a protein sequence via a creator module of the artificial intelligence engine, wherein the artificial intelligence engine is executed by one or more processing devices;

determine, using a decoder of the artificial intelligence engine, which dimensions are included in the candidate drug compound by converting the candidate drug compound to the higher-dimensional vector and obtaining a set of coordinates from the higher-dimensional vector, wherein the coordinates represent the protein drug compound structural information, the protein drug compound activity information, the protein drug compound semantic information, or some combination thereof; and

based on the dimensions, determine, by the artificial intelligence engine, an effectiveness of a biomedical feature of the candidate drug compound, wherein the creator module comprises a generator machine learning model and a discriminator machine learning model, wherein:

the generator machine learning model is trained to receive the lower-dimensional vector and to generate, based on a counterfactual comprising a specification for modifying a different candidate drug compound, the candidate drug compound, and

the discriminator machine learning model is trained to receive the candidate drug compound as input and to predict, based on biomedical activity data pertaining to the plurality of protein drug compounds, an output related to the effectiveness of the biomedical feature which the candidate drug compound provides; and

use at least the output related to the effectiveness of the biomedical feature which the candidate drug compound provides, training, by the one or more processing devices, the generator machine learning model to remove the candidate drug compound from consideration, when the consideration occurs during an iteration of a subsequent generation.

19 . The computer-readable medium of claim 18 , wherein the processing device is further configured to:

include the candidate drug compound as a node in a knowledge graph; and

generate, via a descriptor module of the artificial intelligence engine, a description for the candidate drug compound at a node in the knowledge graph, wherein the knowledge graph comprises the multi-dimensional representation and the respective description comprises the drug compound structural information, the drug compound activity information, and the drug compound semantic information.

20 . The computer-readable medium of claim 19 , wherein the processing device is further configured to, based on the description, perform, via a scientist module of the artificial intelligence engine, at least one benchmark analysis of a parameter of the creator module, wherein the scientist module comprises a machine learning model trained to perform the benchmark analysis of the parameter of the creator module, wherein the at least one benchmark analysis comprises assigning a score to the parameter of the creator module, and the parameter relates at least to an ability of the creator module to generate candidate drug compounds.

21 . The computer-readable medium of claim 20 , wherein the processing device is further configured to tune, based on the score of the parameter, an attribute of the creator module.

22 . The computer-readable medium of claim 21 , wherein the attribute comprises a weight, an activation function, a hidden layer number, a loss function, or some combination thereof.

23 . An artificial intelligence engine for generating candidate drug compounds, wherein a processing device executes the artificial intelligence engine to:

generate a multi-dimensional representation of a plurality of protein drug compounds, wherein for each protein drug compound of the plurality of protein drug compounds, the multi-dimensional representation includes one or more relationships between protein drug compound structural information, protein drug compound activity information, and protein drug compound semantic information;

translate, by the artificial intelligence engine, the multi-dimensional representation to a plurality of encodings;

concatenate, by the artificial intelligence engine, a plurality of encodings to form a concatenated vector, wherein:

the encodings are each respective protein sequences represented in a vector,

a first encoding of the plurality of encodings pertains to the protein drug compound structural information,

a second encoding of the plurality of encodings pertains to the protein drug compound activity information, and

a third encoding of the plurality of encodings pertains to the protein drug compound semantic information;

use an autoencoder of the artificial intelligence engine to compress the concatenated vector from a higher-dimensional vector to a lower-dimensional vector, wherein compressing the concatenated vector from the higher-dimensional vector to the lower-dimensional vector reduces processing complexity;

generate, using the lower-dimensional vector, a candidate drug compound comprising a protein sequence via a creator module of the artificial intelligence engine, wherein the artificial intelligence engine is executed by one or more processing devices;

determine, using a decoder of the artificial intelligence engine, which dimensions are included in the candidate drug compound by converting the candidate drug compound to the higher-dimensional vector and obtaining a set of coordinates from the higher-dimensional vector, wherein the coordinates represent the protein drug compound structural information, the protein drug compound activity information, the protein drug compound semantic information, or some combination thereof; and

based on the dimensions, determine, by the artificial intelligence engine, an effectiveness of a biomedical feature of the candidate drug compound, wherein the creator module comprises a generator machine learning model and a discriminator machine learning model, wherein:

the generator machine learning model is trained to receive the lower-dimensional vector and to generate, based on a counterfactual comprising a specification for modifying a different candidate drug compound, the candidate drug compound, and

the discriminator machine learning model is trained to receive the candidate drug compound as input and to predict, based on biomedical activity data pertaining to the plurality of protein drug compounds, an output related to the effectiveness of the biomedical feature which the candidate drug compound provides; and

use at least the output related to the effectiveness of the biomedical feature which the candidate drug compound provides, training, by the one or more processing devices, the generator machine learning model to remove the candidate drug compound from consideration, when the consideration occurs during an iteration of a subsequent generation.

24 . The artificial intelligence engine of claim 23 , wherein the processing device is further configured to:

include the candidate drug compound as a node in a knowledge graph; and

generate, via a descriptor module of the artificial intelligence engine, a description for the candidate drug compound at a node in the knowledge graph, wherein the knowledge graph comprises the multi-dimensional representation and the respective description comprises the drug compound structural information, the drug compound activity information, and the drug compound semantic information.

25 . The artificial intelligence engine of claim 24 , wherein the processing device is further configured to, based on the description, perform, via a scientist module of the artificial intelligence engine, at least one benchmark analysis of a parameter of the creator module, wherein the scientist module comprises a machine learning model trained to perform the benchmark analysis of the parameter of the creator module, wherein the at least one benchmark analysis comprises assigning a score to the parameter of the creator module, and the parameter relates at least to an ability of the creator module to generate candidate drug compounds.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 20, 2023
From: LEE, FRANCIS; STECKBECK, JONATHAN D., DR.; HOLSTE, HANNES
To: PEPTILOGICS, INC.
Reel/Frame 063386/0716 →
Continuity (3)
Division 16870611 · May 8, 2020
Provisional Application 62975470 · Feb 12, 2020
Related Publication 20210366581A1 · Nov 25, 2021
References Cited (172)
US 6547994B1 · Monkhouse et al. · 2003 [cited by applicant]
US 6849403B1 · Shuber · 2005 [cited by applicant]
US 9169287B2 · Simon et al. · 2015 [cited by applicant]
US 9695214B2 · Simon et al. · 2017 [cited by applicant]
US 9868759B2 · Simon et al. · 2018 [cited by applicant]
US 10683325B2 · Thomas, III et al. · 2020 [cited by applicant]
US 10889613B2 · Simon et al. · 2021 [cited by applicant]
US 10988504B2 · Simon et al. · 2021 [cited by applicant]
US 11049590B1 · Lee et al. · 2021 [cited by applicant]
US 11152125B2 · Spiro · 2021 [cited by applicant]
US 11615324B2 · Pabrinkis · 2023 [cited by applicant]
US 20030028327A1 · Brunner et al. · 2003 [cited by applicant]
US 20030103675A1 · Endo · 2003 [cited by examiner]
US 20040115726A1 · Nagashima · 2004 [cited by applicant]
US 20040115762A1 · Zotchev · 2004 [cited by applicant]
US 20050032119A1 · Tickle · 2005 [cited by applicant]
US 20050060305A1 · Hopkins et al. · 2005 [cited by applicant]
US 20060052943A1 · Ramani et al. · 2006 [cited by applicant]
US 20060106545A1 · Balaji · 2006 [cited by applicant]
US 20070178473A1 · Chen · 2007 [cited by applicant]
US 20090104638A1 · Hebert · 2009 [cited by applicant]
US 20100184034A1 · Bankaitis-Davis et al. · 2010 [cited by applicant]
US 20110207113A1 · Hebert · 2011 [cited by applicant]
US 20120296090A1 · Wong · 2012 [cited by applicant]
US 20140047365A1 · Clark et al. · 2014 [cited by applicant]
US 20150065357A1 · Fox · 2015 [cited by applicant]
US 20150142408A1 · Futamura · 2015 [cited by applicant]
US 20150307947A1 · Basu et al. · 2015 [cited by applicant]
US 20150371009A1 · Chen · 2015 [cited by applicant]
US 20160063180A1 · Weaver et al. · 2016 [cited by applicant]
US 20170081359A1 · Thomas, III et al. · 2017 [cited by applicant]
US 20170147743A1 · Bixby · 2017 [cited by applicant]
US 20180261331A1 · Futamura · 2018 [cited by applicant]
US 20180372724A1 · Khine · 2018 [cited by applicant]
US 20190018933A1 · Oono et al. · 2019 [cited by applicant]
US 20190034586A1 · Pirrotte et al. · 2019 [cited by applicant]
US 20190065677A1 · Gifford et al. · 2019 [cited by applicant]
US 20190095584A1 · Hu · 2019 [cited by applicant]
US 20190220524A1 · Costabello et al. · 2019 [cited by applicant]
US 20190279737A1 · Na · 2019 [cited by applicant]
US 20190304568A1 · Wei et al. · 2019 [cited by applicant]
US 20200082916A1 · Polykovskiy · 2020 [cited by applicant]
US 20200327963A1 · Ain et al. · 2020 [cited by applicant]
US 20200342960A1 · Liu · 2020 [cited by applicant]
US 20200365238A1 · Kazemi et al. · 2020 [cited by applicant]
US 20200392178A1 · Manica · 2020 [cited by applicant]
US 20210047365A1 · Thomas et al. · 2021 [cited by applicant]
US 20210183467A1 · Goodman · 2021 [cited by applicant]
US 20210188901A1 · Thomas, III et al. · 2021 [cited by applicant]
US 20210202047A1 · Kang · 2021 [cited by applicant]
US 20210217498A1 · Pae · 2021 [cited by applicant]
US 20210249108A1 · Lee et al. · 2021 [cited by applicant]
US 20210287763A1 · Sharma · 2021 [cited by applicant]
US 20210295955A1 · Lee et al. · 2021 [cited by applicant]
US 20210304854A1 · Lee et al. · 2021 [cited by applicant]
US 20210365795A1 · Pae · 2021 [cited by applicant]
US 20210407642A1 · Xia et al. · 2021 [cited by applicant]
US 20220009966A1 · Das et al. · 2022 [cited by applicant]
US 20220036969A1 · De Oliveira · 2022 [cited by applicant]
US 20220059196A1 · Lee et al. · 2022 [cited by applicant]
US 20220101972A1 · Bajpai · 2022 [cited by applicant]
US 20220122692A1 · Feala et al. · 2022 [cited by applicant]
US 20220157403A1 · Mason et al. · 2022 [cited by applicant]
US 20220199198A1 · Rooney et al. · 2022 [cited by applicant]
US 20220237930A1 · Rando · 2022 [cited by applicant]
US 20220238179A1 · Sarmiento et al. · 2022 [cited by applicant]
US 20220246239A1 · Cho · 2022 [cited by applicant]
US 20220328140A1 · Apgar · 2022 [cited by applicant]
US 20220348903A1 · Ranganathan et al. · 2022 [cited by applicant]
US 20220367002A1 · Sellwood · 2022 [cited by applicant]
US 20220375539A1 · Alvarez et al. · 2022 [cited by applicant]
US 20220415433A1 · Xu · 2022 [cited by applicant]
US 20220415452A1 · Ye et al. · 2022 [cited by applicant]
US 20230004862A1 · Xiang · 2023 [cited by applicant]
US 20230022022A1 · Laniado et al. · 2023 [cited by applicant]
US 20230034559A1 · Fu et al. · 2023 [cited by applicant]
US 20230050156A1 · Yu · 2023 [cited by applicant]
US 20230083769A1 · Ma · 2023 [cited by applicant]
US 20230094323A1 · Cho et al. · 2023 [cited by applicant]
US 20230098285A1 · Jhe et al. · 2023 [cited by applicant]
US 20230098833A1 · Li · 2023 [cited by applicant]
US 20230207066A1 · Lee · 2023 [cited by applicant]
CN 1876172A · 2006 [cited by applicant]
CN 102027373A · 2011 [cited by applicant]
CN 103733065A · 2014 [cited by applicant]
CN 107810416A · 2018 [cited by applicant]
CN 108323184A · 2018 [cited by applicant]
CN 109036571A · 2018 [cited by applicant]
CN 110379459A · 2019 [cited by applicant]
CN 111164704A · 2020 [cited by applicant]
CN 113838536A · 2021 [cited by applicant]
CN 114187979A · 2022 [cited by applicant]
WO 2005051303A1 · 2005 [cited by applicant]
WO 2009049045A1 · 2009 [cited by applicant]
WO 2019006022A1 · 2019 [cited by applicant]
WO 2022026551A1 · 2022 [cited by applicant]
WO 2023049466A2 · 2023 [cited by applicant]
Guney et al. “Network-based in silico drug efficacy screening” Nature Communications, 7 (1) (2016), p. 1-13. [cited by examiner]
Vamathevan et al. Nature Reviews Drug Discovery, 18, pp. 463-477, 2019. [cited by examiner]
Rifaioglu et al. “Recent applications of deep learning and machine intelligence on in silico drug discovery: methods, tools and databases” Brief Bioinform, 20 (5), 2019, pp. 1878-1912. [cited by examiner]
Liu et al. “Retrosynthetic reaction prediction using neural sequence-to-sequence models” ACS Cent Sci, 3 (10), 2017, pp. 1103-1113. [cited by examiner]
Leamer Carnegie-Rochester Conference Series on Public Policy 22 (1985) 255304v. [cited by examiner]
Hodgkin Mol. Modell. Drug Des. (1994), 137-69. [cited by examiner]
Peiretti et al. “Artificial Intelligence: The Future for Organic Chemistry?” ACS Omega, 2018, 3 (10), 13263-13266. [cited by examiner]
Sadiq, Saad, Mei-Ling Shyu, and Daniel J. Feaster. “Counterfactual autoencoder for unsupervised semantic learning.” International Journal of Multimedia Data Engineering and Management (IJMDEM) 9.4 (2018): 1-20. [cited by examiner]
Atherton et al., A physically supported gel polymer for low pressure, continuous flow solid phase reactions. Application to solid phase peptide synthesis., J. Chem. Soc., Chem. Commun., 1981, pp. 1151-1152. [cited by applicant]
Atherton et al., Internal association in solid phase peptide synthesis. Synthesis of cytochrome C residues 66-104 on polyamide supports, J. Chem. Soc., Chem. Commun., 1980, pp. 970-971. [cited by applicant]
Badowski et al., Synergy Between Expert and Machine-Learning Approaches Allows for Improved Retrosynthetic Planning, Angewandte Chemie International Edition, 2020, vol. 59, pp. 725-730. [cited by applicant]
Baker, 1,500 scientists lift the lid on reproducibility, Nature News, 2016, vol. 533, p. 452. [cited by applicant]
Bédard et al., Reconfigurable system for automated optimization of diverse chemical reactions, Science, 2018, vol. 361, p. 1220. [cited by applicant]
Bedford et al., Amino acid structure and difficult sequences in solid phase peptide synthesis, International Journal of Peptide and Protein Research, 1992, vol. 40, pp. 300-307. [cited by applicant]
Bondalapati et al., Expanding the chemical toolbox for the synthesis of large and uniquely modified proteins, Nature Chemistry, 2016, vol. 8, pp. 407-418. [cited by applicant]
Brown et al., Analysis of Past and Present Synthetic Methodologies on Medicinal Chemistry: Where Have All the New Reactions Gone?, J. Med. Chem., 2016, vol. 59, pp. 4443-4458. [cited by applicant]
Cameron et al., Peptide synthesis. Part 13. Feedback control in solid phase synthesis. Use of fluorenylmethoxycarbonyl amino acid 3,4-dihydro-4-oxo-1,2,3-benzotriazin-3-yl esters in a fully automated system., J. Chem. S… [cited by applicant]
Coley et al., A robotic platform for flow synthesis of organic compounds informed by AI planning, Science, 2019, vol. 365, p. eaax1566. [cited by applicant]
Coley et al., Autonomous discovery in the chemical sciences part II: Outlook. Angewandte Chemie International, 2019, Edition 0, doi:10.1002/anie.201909989. [cited by applicant]
Coley et al., Machine Learning in Computer-Aided Synthesis Planning, Acc. Chem. Res., 2018, vol. 51, pp. 1281-1289. [cited by applicant]
Coley et al., Prediction of Organic Reaction Outcomes Using Machine Learning, ACS Cent. Sci., 2017, vol. 3, pp. 434-443. [cited by applicant]
Dryland et al., Peptide synthesis. Part 8. A system for solid-phase synthesis under low pressure continuous flow conditions., J. Chem. Soc., Perkin Trans., 1986, vol. 1, pp. 125-137. [cited by applicant]
Gao et al., Using Machine Learning To Predict Suitable Conditions for Organic Reactions, ACS Cent. Sci., 2018, vol. 4, pp. 1465-1476. [cited by applicant]
Hartrampf et al., Synthesis of proteins by automated flow chemistry, Science, 2020, vol. 368, p. 980. [cited by applicant]
Häse et al., Phoenics: A Bayesian Optimizer for Chemistry, ACS Cent. Sci., 2018, vol. 4, pp. 1134-1145. [cited by applicant]
Kent, Total chemical synthesis of proteins, Chem. Soc. Rev., 2009, vol. 38, pp. 338-351. [cited by applicant]
Li et al., Synthesis of many different types of organic small molecules using one automated process, Science, 2015, vol. 347, p. 1221. [cited by applicant]
Lukas et al., Solid-phase peptide synthesis under continuous-flow conditions, Proc Natl Acad Sci USA, 1981, vol. 78, p. 2791. [cited by applicant]
MacLeod et al., Self-driving laboratory for accelerated discovery of thin-film materials, Sci Adv., 2020, vol. 6, p. eaaz8867. [cited by applicant]
Merrifield, Solid phase peptide synthesis. I. The synthesis of a tetrapeptide, Journal of the American Chemical Society, 1963, vol. 85, pp. 2149-2154. [cited by applicant]
Mijalis et al., A fully automated flow-based approach for accelerated peptide synthesis, Nature Chemical Biology, 2017, vol. 13, p. 464. [cited by applicant]
Milton et al., Prediction of difficult sequences in solidphase peptidesynthesis, J. Am. Chem. Soc., 1990, vol. 112, pp. 6039-6046. [cited by applicant]
Mohapatra et al., Deep Learning for Prediction and Optimization of Fast-Flow Peptide Synthesis, ACS Cent. Sci., 2020, vol. 6, pp. 2277-2286. [cited by applicant]
Reizman et al., Feedback in Flow for Accelerated Reaction Development, Acc. Chem. Res., 2016, vol. 49, pp. 1786-1796. [cited by applicant]
Roch et al., ChemOS: Orchestrating autonomous experimentation, Sci. Robotics, 2018, vol. 3, p. eaat5559. [cited by applicant]
Sarin et al., Properties of swollen polymer networks. Solvation and swelling of peptide-containing resins in solid-phase peptide synthesis, J. Am. Chem. Soc., 1980, vol. 102, pp. 5463-5470. [cited by applicant]
Segler et al., Planning chemical syntheses with deep neural networks and symbolic AI, Nature, 2018, vol. 555, pp. 604-610. [cited by applicant]
Steiner et al., Organic synthesis in a modular robotic system driven by a chemical programming language, Science, 2019, vol. 363, p. eaav2211. [cited by applicant]
Trobe et al., The Molecular Industrial Revolution: Automated Synthesis of Small Molecules, Angewandte Chemie International Edition, 2018, vol. 57, pp. 4192-4214. [cited by applicant]
Waltz et al., Automating Science, Science, 2009, vol. 324, p. 43. [cited by applicant]
Wei et al., Neural Networks for the Prediction of Organic Chemistry Reactions, ACS Cent. Sci., 2016, vol. 2, pp. 725-732.a. [cited by applicant]
Woerkom et al., Difficult couplings in stepwise solid phase peptide synthesis: predictable or just a guess?, International Journal of Peptide and Protein Research, 1991, vol. 38, pp. 103-113. [cited by applicant]
Woerly et al., Synthesis of most polyene natural product motifs using just 12 building blocks and one coupling reaction, Nature Chemistry, 2014, vol. 6, pp. 484-491. [cited by applicant]
Zhou et al., Optimizing Chemical Reactions with Deep Reinforcement Learning, ACS Cent. Sci., 2017, vol. 3, pp. 1337-1344. [cited by applicant]
Zompra et al., Manufacturing peptides as active pharmaceutical ingredients, Future Medicinal Chemistry, 2009, vol. 1, pp. 361-377. [cited by applicant]
Inza et al., “Machine Learning: An Indispenaible Tool in Bioinformatics”, Bioinformatics Methods in Clinical Research, 24 pages. [cited by applicant]
Hinton, et al., “Reducing the Dimensionality of Data with Neural Networks”, Science, vol. 313, 2006, four pages. [cited by applicant]
Larranga, et al., “Briefings in Bioinformatics”, vol. 7, Issue 1, 2006, 27 pages. [cited by applicant]
Gomez-Bombarelli et al., “Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules”, Dec. 5, 2017, 26 pages. [cited by applicant]
International Searching Authority, Search Report and Written Opinion for International Application No. PCT/US21/15745, mailed Apr. 12, 2021. [cited by applicant]
Yang, et al. “Concepts of Artificial Intelligence for Computer-Assisted Drug Discovery”, in Chemical Reviews (online), pp. 10520-10573, published Jul. 11, 2019 (retrieved on May 6, 2021). Retrieved from the internet <Ur… [cited by applicant]
Ma, et al. “Development of a Dual-Acting Antibacterial Agent (TNP-2092) for the Treatment of Persistent Bacterial Infections”, in Journal of Medical Chemistry (online), pp. 6645-6657, published Jun. 23, 2016 (retrieved … [cited by applicant]
Hofler: “Casual inference based on counterfactuals”, in BioMed Central (Online), pp. 1-12, published Sep. 13, 2005 (retrieved on May 6, 2021). Retrieved from Internet <URL: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1… [cited by applicant]
Mak et al. “Artificial intelligence in drug development: present status and future prospects”, Drug Discovery Today, vol. 24, No. 3, Mar. 2019, pp. 773-780. [cited by applicant]
Biswas., “Principles of Machine Learning Guided Protein Engineering”, Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences 2020, 195 Pages. [cited by applicant]
Xiong et al.“Pushing the Boundaries of Molecular Representation for Drug Discovery with the Graph Attention Mechanism”, J. Med. Chem. 2020, vol. 63, pp. 8749-8760. [cited by applicant]
International Searching Authority, Search Report and Written Opinion for International Application No. PCT/US2021/047325, Mailing Date Dec. 14, 2021, 6 Pages. [cited by applicant]
Website https://en.wikipedia.org/wiki/Assay Assay Definition Wikipedia, Downloaded 2022, 10 Pages. [cited by applicant]
Website https://www.rxlist.com/anti-infective/definition.htm Anti-Infection definition, RxList downloaded 2022, 1 Page. [cited by applicant]
Bradley., “Anti-infective research development-problems, challenges and solutions”, Lancet Infectious Diseases 2007, vol. 7, pp. 68-78. [cited by applicant]
Modena et al. “Smart Cell Culture Systems: Integration of Sensors and Actuators into Microphysiological Systems”, ACS Chemical Biology 2018, vol. 13, pp. 1767-1784. [cited by applicant]
Kuhn et al. “Chapter 4: Statistical Methods for Drug Discovery”, IN: Zhang, L. (ed) Nonclinical statistics for pharmaceutical and biotechnology industries. Statistics for biology in health, Springer, Switzerland 2016, p… [cited by applicant]
Zang et al. “Cell-based Assays in High-throughput Screening for Drug Discovery”, International Journal of Biotechnology for Wellness Industries 2012, vol. 1, pp. 31-51. [cited by applicant]
Vasan, Biomarkers of Cardiovascular Disease: Molecular Basis and Practical Considerations, Circulation, May 16, 2006, 113(19):2335-62. [cited by applicant]
Sabo et al., Optimization of the Hemolysis Assay for the assessment of cytotoxicity, Int. Journal of Molecular Sciences, Feb. 2023, 24(3):2914, 20 pgs. [cited by applicant]
Assay definition, retrieved on Jun. 26, 2023, https://en.wikipedia.org/wiki/Assay#:˜: text=An%20assay%20usually%20aims%20to,a%20standard%2C%20etc.). [cited by applicant]
Artificial Intelligence definition, retrieved on Jun. 26, 2023, https://en.wikipedia.org/wiki/Artificial_intelligence. [cited by applicant]
Extended European Search Report for EP Application No. 21754623.3, dated Jul. 22, 2022, pp. 1-10. [cited by applicant]
Elton et al., “Deep learning for molecular generation and optimization—a review of the state of the art,” arxiv.org, Cornell University Library, Mar. 11, 2019, pp. 1-22, New York. [cited by applicant]
Sanchez-Lengeling et al., “Inverse molecular design using machine learning: Generative models for matter engineering,” Science, Jul. 27, 2018, pp. 360-365, vol. 361, No. 6400. [cited by applicant]
Luo et al., “A network integration approach for drug-target interaction prediction and computational drug repositioning from heterogeneous information,” Nature Communications, Sep. 18, 2017, pp. 1-13, vol. 8, No. 1. [cited by applicant]
Ji et al., “A Survey on Knowledge Graphs: Representation, Acquisition and Applications,” IEEE transaction on neural networks and learning systems, Feb. 2, 2020, pp. 1-25. [cited by applicant]
Dara, et al., “Machine Leanging in Drug Discovery: A Review,” Artificial Intelligence Review, Aug. 11, 2021, 53 pages, Date Accessed: May 2, 2023. [cited by applicant]
Hong et al., “Molecular Generative Model Based on Adversarially Regularized Autoencoder,” Journal of Chemical Information and Modeling, Dec. 10, 2019, pp. 29-36, Korea. [cited by applicant]
Sadiq et al., “Counterfactual Autoencoder for Unsupervised Semantic Learning,” International Journal of Multimedia Data Engineering and Management, 2018, 20 pages. [cited by applicant]
Cited By (1)
US 12,694,954