DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Status
Claims 1-6 are cancelled.
Claims 7-15 are newly added.
Claims 7-15 are currently pending and examined on the merits.
Priority
The instant application is a 371 of PCT/KR2020/016951 filed on 11/26/2020 and claims foreign priority to KR10-2020-0138236 filed on 10/23/2020, in Korea. At this point in examination, the effective filing date of claims 7-15 is 10/23/2020.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 4/12/2023 is in compliance with the provisions of 37 CFR 1.97. A signed copy of the corresponding 1449 form has been included with this Office Action.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 7-15 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claims recite: (a) mathematical concepts, (e.g., mathematical relationships, formulas or equations, mathematical calculations); and (b) mental processes, i.e., concepts performed in the human mind, (e.g., observation, evaluation, judgement, opinion).
Subject matter eligibility evaluation in accordance with MPEP 2106:
Eligibility Step 1: Claims 7-15 are directed to a method (process) for deciding availability as a new drug by creating a new compound. Therefore, these claims are encompassed by the categories of statutory subject matter, and thus satisfy the subject matter eligibility requirements under Step 1.
[Step 1: YES]
Eligibility Step 2A: First, it is determined in Prong One whether a claim recites a judicial exception, and if so, then it is determined in Prong Two whether the recited judicial exception is integrated into a practical application of that exception.
Eligibility Step 2A, Prong One: In determining whether a claim is directed to a judicial exception, examination is performed that analyzes whether the claim recites a judicial exception, i.e., whether a law of nature, natural phenomenon, or abstract idea is set forth described in the claim.
Claim 7 recites the following steps which fall within the mental processes and/or mathematical concepts groups of abstract ideas, as noted below.
Independent claim 7 further recites:
determining, by a drug prediction device, at least one first new compound (i.e., mental processes);
determining, by the drug prediction device, a first target new compound among the at least one first new compound (i.e., mental processes);
determining, by the drug prediction device, a first decision information about a target protein and the first target new compound (i.e., mental processes);
determining, by the drug prediction device, an availability as a new drug of the first target new compound based on the first decision information (i.e., mental processes);
generating, by the drug prediction device, a reward function result value of the first target new compound by applying a reward function based on the first decision information (i.e., mental processes);
determining, by the drug prediction device, at least one second new compound based on the reward function result value (i.e., mental processes).
The abstract ideas recited in the claims are evaluated under the broadest reasonable interpretation (BRI) of the claim limitations when read in light of and consistent with the specification. As noted in the foregoing section, the claims are determined to contain limitations that can practically be performed in the human mind with the aid of a pencil and paper, and therefore recite judicial exceptions from the mental process grouping of abstract ideas. Additionally, the recited limitations that are identified as judicial exceptions from the mathematical concepts grouping of abstract ideas are abstract ideas irrespective of whether or not the limitations are practical to perform in the human mind. Dependent claims 8-15 recite information further limiting the judicial exceptions indicated above.
Therefore, claim 7 recites an abstract idea.
[Step 2A, Prong One: YES]
Eligibility Step 2A, Prong Two: In determining whether a claim is directed to a judicial exception, further examination is performed that analyzes if the claim recites additional elements that, when examined as a whole, integrates the judicial exception(s) into a practical application (MPEP 2106.04(d)). A claim that integrates a judicial exception into a practical application will apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception. The claimed additional elements are analyzed to determine if the abstract idea is integrated into a practical application (MPEP 2106.04(d)(I); MPEP 2106.05(a-h)). If the claim contains no additional elements beyond the abstract idea, the claim fails to integrate the abstract idea into a practical application (MPEP 2106.04(d)(III)).
Claims 7-15 do not recite any additional elements in addition to the judicial exception and therefore fail to integrate the abstract ideas into a practical application. See MPEP 2106.04.II.A.2.
[Step 2A, Prong Two: NO]
Eligibility Step 2B: Because the claims recite an abstract idea, and do not integrate that abstract idea into a practical application, the claims are probed for a specific inventive concept. The judicial exception alone cannot provide that inventive concept or practical application (MPEP 2106.05). Identifying whether the additional elements beyond the abstract idea amount to such an inventive concept requires considering the additional elements individually and in combination to determine if they amount to significantly more than the judicial exception (MPEP 2106.05A i-vi).
Claims 7-15 are drawn to a judicial exception and do not recite any additional elements that amount to significantly more than the judicial exception. Furthermore, an inventive concept cannot be furnished by a judicial exception. See MPEP 2106.05.I.
[Step 2B: NO]
Therefore, claims 7-15 are patent ineligible under 35 U.S.C. § 101.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 7-9 and 11-13 are rejected under 35 U.S.C. 103 as being unpatentable over Khemchandani et al. (Journal of Cheminformatics, 2020, 12(53), 1-17), in view of Popova et al. (Science Advances, 2018, 4(7), 1-14).
With respect to claim 7:
Regarding the recited determining, by a drug prediction device, at least one first new compound, Khemchandani et al. discloses a molecular generation task that uses reinforcement learning to incorporate both molecular constraints and desired properties using reward functions along with graph convolutional policy networks for generating molecules (pg. 3, col. 1, para. 1, lines 14-28; pg. 2, Fig. 1). This teaches determining a new compound.
Regarding the recited determining, by the drug prediction device, a first target new compound among the at least one first new compound, Khemchandani et al. discloses selecting top 10 molecules generated by DeepGraphMolGen for analysis (pg. 3, col. 1, para. 1, lines 14-28; pg. 11, col. 1, para. 1, lines 1-2). This teaches determining a target compound from the new compounds.
Regarding the recited determining, by the drug prediction device, a first decision information about a target protein and the first target new compound, Khemchandani et al. discloses predicting property scores of molecules, specifically the binding constant of various molecules at the dopamine and norepinephrine transporters (pg. 3, col. 1, para. 1, lines 5-9; pg. 2, Fig. 1). This teaches determining a decision information about a target protein and a target compound.
Regarding the recited generating, by the drug prediction device, a reward function result value of the first target new compound by applying a reward function based on the first decision information, Khemchandani et al. discloses that their reinforcement learning strategy for molecular generation can be implemented using any graph-based reinforcement learning generator since it would just need to use the predicted property score as the reward function (pg. 5, col. 1, para. 1, lines 12-16; pg. 2, Fig. 1). This teaches generating a reward function value of a compound, where the reward function is applied based on the predicted property score.
Regarding the recited determining, by the drug prediction device, at least one second new compound based on the reward function result value, Khemchandani et al. discloses that final property rewards are calculated by the previously trained model for newly generated molecules (pg. 5, col. 1, para. 1, lines 1-5; pg. 2, Fig. 1). Property scores are used as reward feedback for the reinforcement learning pathway, which leads to molecular generation (pg. 2, Fig. 1). This teaches a cycle where property scores from the previous molecular generation are fed into the reinforcement learning pathway to generate a second new compound.
Khemchandani et al. does not disclose determining, by the drug prediction device, an availability as a new drug of the first target new compound based on the first decision information.
However, Popova et al. discloses that orally bioavailable compounds should have an octanol-water partition coefficient logP less than 5, therefore aiming to design a library of compounds with logP values within a favorable drug-like range (pg. 7, col. 1, para. 3). This teaches determining drug available compounds based on a decision information.
It would have been prima facie obvious to one of ordinary skill in the art to combine the drug generation method disclosed by Khemchandani et al. with determining drug availability disclosed by Popova et al. One would be motivated to combine the drug generation method with drug availability because Popova et al. discloses success in generating a library with 88% of the molecules falling within the drug-like region of logP values (pg. 7, col. 1, para. 3). This indicates a high percentage of success for determining drug availability in the drug generation method. There is a likelihood of success, since both teachings are of compound generation methods using generative or predictive models, which are well known techniques in the field of cheminformatics.
With respect to claim 8:
Popova et al. does not disclose wherein the first decision information comprises: at least one of a binding affinity information between the target protein and the first target new compound, an absorption prediction information, a distribution prediction information, a metabolism prediction information, an excretion prediction information, a toxicity prediction information, QED (quantitative estimate of druglikeness) information, or a SAS (synthetic accessibility scores) information as sub decision information.
However, Khemchandani et al. discloses adding metrics such as the quantitative estimation of drug-likeness (QED) and synthetic accessibility (SA) scores to the reward function to generate drug-like molecules that can be synthetically generated (pg. 6, col. 1, para. 1, lines 6-9). This teaches a decision information comprising QED information and SAS information.
With respect to claim 9:
Popova et al. does not disclose wherein the first decision information comprises a binding affinity information between the target protein and the first target new compound.
However, Khemchandani et al. discloses training a model to predict property scores of molecules, specifically the binding constant of various molecules at the dopamine and norepinephrine transporters (pg. 3, col. 1, para. 1, lines 5-9). This teaches a decision information comprising binding affinity information between a target protein and a target compound.
With respect to claim 11:
Popova et al. does not disclose wherein at least one second new compound is generated by an additional learning path that applies a reward function based on at least one of a sub decision information among the plurality of sub decision information, considering reliability or importance of the plurality of sub decision information included in the first decision information, wherein the plurality of sub decision information includes at least one of a binding affinity information between the target protein and the first target new compound, an absorption prediction information, a distribution prediction information, a metabolism prediction information, an excretion prediction information, a toxicity prediction information, a QED (quantitative estimate of druglikeness) information, or a SAS (synthetic accessibility scores) information.
However, Khemchandani et al. discloses modifying the original reward function to be a weighted combination of p
K
i
values for two different receptors when generating molecules that have high binding affinity to the dopamine transporter, but low binding affinity to the norepinephrine transporter (pg. 11, col. 2, para. 2, lines 1-20). This teaches generating a new compound by applying a modified reward function that emphasizes importance on binding affinity between target proteins and target compounds.
With respect to claim 12:
Popova et al. does not disclose wherein at least one first new compound is generated by bond addition, bond deletion, atom addition, or atom deletion using a Markov decision process.
However, Khemchandani et al. discloses building a molecule step by step with an addition of a new bond in each step, and graph generation is treated as a Markov decision process (pg. 5, col. 2, para. 1, lines 1-6). This teaches generating a compound by bond addition using a Markov decision process.
With respect to claim 13:
Popova et al. does not disclose wherein the at least one second new compound is determined by a Markov decision process considering a reward function result value generated by applying a reward function based on the first decision information, and determined by applying bond addition, bond deletion, atom addition, or atom deletion for the first target new compound.
However, Khemchandani et al. discloses building a molecule step by step with an addition of a new bond in each step, and graph generation is treated as a Markov decision process (pg. 5, col. 2, para. 1, lines 1-6). Also, further discloses final property rewards are calculated by the previously trained model for newly generated molecules (pg. 5, col. 1, para. 1, lines 1-5; pg. 2, Fig. 1). Property scores are used as reward feedback for the reinforcement learning pathway, which leads to molecular generation (pg. 2, Fig. 1). This teaches a cycle where reward values from the previous molecular generation are fed into the reinforcement learning pathway to generate a second new compound. The process of generating this compound is determined by a Markov decision process and applying bond addition.
Claim 10 is rejected under 35 U.S.C. 103 as being unpatentable over Khemchandani et al. (Journal of Cheminformatics, 2020, 12(53), 1-17) and Popova et al. (Science Advances, 2018, 4(7), 1-14) as applied to claims 7-9 and 11-13 above, in view of Ozturk et al. (Bioinformatics, 2018, 34, i821-i829), as provided in the IDS filed 4/12/2023.
Khemchandani et al. and Popova et al. are applied to claims 7-9 and 11-13 above.
With respect to claim 10:
Khemchandani et al. and Popova et al. do not disclose wherein the binding affinity information is predicted by preprocessing the first target new compound and the target protein, and concatenating the preprocessed first target new compound and the preprocessed target protein.
However, Ozturk et al. discloses using integer/label encoding to represent compound SMILES sequences and a similar label encoding for protein sequences (pg. i823-i824, col. 2, para. 3, lines 1-11). Also, further discloses concatenating the max-pooling layers of the compound SMILES representation and the protein sequence representation (pg. i824, col. 1, para. 3, lines 3-9; pg. i824, Fig. 2). This teaches preprocessing target compounds and proteins and concatenating the preprocessed compounds and proteins.
It would have been prima facie obvious to one of ordinary skill in the art to modify the drug generation method disclosed by Khemchandani et al. and Popova et al. to incorporate preprocessing and concatenating target compounds and proteins disclosed by Ozturk et al. One would be motivated to incorporate preprocessing and concatenating target compounds and proteins into the drug generation method because Ozturk et al. discloses that performance significantly increased in comparison to baseline methodologies when two CNN-blocks that learn representations of proteins and drugs based on raw sequence data are used in conjunction with DeepTA (pg. i827, col. 2, para. 2). This indicates that preprocessing and concatenating target compounds and proteins would improve the performance of the drug generation method. There is a likelihood of success, since the teachings are either compound generation methods using generative models or drug target binding affinity prediction methods, which are well known techniques in the field of cheminformatics.
Claim 14 is rejected under 35 U.S.C. 103 as being unpatentable over Khemchandani et al. (Journal of Cheminformatics, 2020, 12(53), 1-17) and Popova et al. (Science Advances, 2018, 4(7), 1-14) as applied to claims 7-9 and 11-13 above, in view of Jin et al. (Proceedings of the 37th International Conference on Machine Learning, 2020, 1-11).
Khemchandani et al. and Popova et al. are applied to claims 7-9 and 11-13 above.
With respect to claim 14:
Popova et al. does not disclose wherein the second new compound is determined based on a first type reward function for a first decision group including at least one first sub decision information of the plurality of sub decision information included in the first decision information or a second type reward function for a second decision group including at least one second sub decision information of the plurality of sub decision information.
However, Khemchandani et al. discloses modifying the original reward function to be a weighted combination of p
K
i
values for two different receptors when generating molecules that have high binding affinity to the dopamine transporter, but low binding affinity to the norepinephrine transporter (pg. 11, col. 2, para. 2, lines 1-20). This teaches generating a new compound by applying a modified reward function that emphasizes importance on binding affinity between target proteins and target compounds.
Khemchandani et al. and Popova et al. do not disclose wherein the first decision group and the second decision group are set to be different according to a target disease.
However, Jin et al. discloses combinations of multi-property constraints that are biologically relevant, such as jointly inhibiting JNK3 and GSK3β, which could provide potential benefit for the treatment of Alzheimer’s disease, and further requiring the dual inhibitors to be drug-like and synthetically accessible (pg. 6, para. 1, lines 1-13). This teaches inhibition of two properties that could be different according to a disease.
It would have been prima facie obvious to one of ordinary skill in the art to modify the drug generation method disclosed by Khemchandani et al. and Popova et al. to incorporate setting decision groups to be different based on target disease disclosed by Jin et al. One would be motivated to modify the drug generation method to incorporate different decision groups because the rationale based generative model for molecular design disclosed by Jin et al. significantly outperforms the baseline in terms of exact matching rationales in molecules (pg. 9, col. 1-2, para. 4). This means there is a high accuracy of matching molecular substructures, therefore the incorporation of different decision groups used in this method would make the drug generation method more accurate. There is a likelihood of success, since all teachings are of compound generation methods using generative or predictive models, which are well known techniques in the field of cheminformatics.
Claim 15 which recites wherein the at least one first sub decision information is information for determining availability as a new drug based on a specific range criterion, wherein the at least one second sub decision information is information for determining availability as a new drug based on a presence, wherein the at least one first sub decision information and the at least one second sub decision information are prioritized based on availability as a new drug, wherein each of the at least one first sub decision information corresponds to a first type reward function considering priority, wherein each of the at least one second sub decision information corresponds to a second type reward function considering priority, wherein the first type reward function is a function considering the specific range criterion, wherein the second type reward function is a function considering the presence, is free of the art.
Conclusion
No claims are allowed.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Jammy Luo whose telephone number is (571)272-2358. The examiner can normally be reached Monday - Friday, 9:00 AM - 5:00 PM EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Larry D Riggs can be reached at (571)270-3062. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/J.N.L./Examiner, Art Unit 1686
/LARRY D RIGGS II/Supervisory Patent Examiner, Art Unit 1686
Prosecution Insights