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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 29 February 2026 has been entered.
Status of the claims
The claim set received 08 August 2025 has been entered into the application.
Claims 1-11, 13-15, and 21 are amended.
Claims 16-20 are canceled.
Claims 1-15 and 21-25 are pending.
Election/Restrictions
Claims 16-20 withdrawn from further consideration pursuant to 37 CFR 1.142(b) as being drawn to a nonelected invention, there being no allowable generic or linking claim. Election was made without traverse in the reply filed on 07 February 2025.
New claims 21-25 are drawn to similar methods as elected claims 1-15.
Claims 1-15 and 21-25 are pending examination.
Priority
This Application is a 371 of PCT/CA2020/050005 filed 20 January 2020 which claims benefit to U.S. Provisional Application 62/788,682 filed 04 January 2019.
Information Disclosure Statement
The information disclosure statements (IDS’s) submitted on 19 February 2026 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement has been considered by the examiner.
Claim Rejections - 35 USC § 112
35 USC § 112(b)
It is noted the amendments received 19 February 2026 necessitated new ground(s) of rejection.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-15 and 21-25 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 1 recites a step of performing a three-dimensional structure alignment for proteins in the protein cluster. Claim 1 recites obtaining ghost ligands by projection a ligand onto a plurality of corresponding binding-site region of other proteins in the cluster “identified by the three-dimensional structure alignment”. However, these limitations are not clear because the initial “performing” alignment step does not recite any clusters of proteins being “identified” and the subsequent and previous steps do not provide identifying steps for “identifying” clusters of proteins from the three-dimensional structure alignments. Therefore, it is unclear what is being identified, for example, proteins in the protein cluster or binding site regions of protein in the cluster.
Claim 1 recites “obtaining the plurality of ghost ligands” by projection of a ligand onto other plurality of binding site regions of other proteins. It is unclear how this step results in a “ghost ligand”. Merely projecting a ligand onto binding regions of a protein(s) is insufficient to result in a ghost ligand. Clarification is needed with respect to what the ghost ligand is and how it is obtained.
Claims 1, 15, and 21 generating scores step recites “…representing projection certainties…”. The term renders the claims indefinite because the metes and bounds of the term/limitation is not clear as to what encompasses projection certainties. The specification does not provide a definition for the term or provide as to what data or values the “projection certainty” is to encompass. It is recommended to amend the claims to provide terms and language consistent with the specification. It is noted claims 8-9 also utilizes the term “projection certainties”. Thus, claims 8-9 are also rejected.
Furthermore, claim 1 generating confidence scores step recites the limitation " “generating…the plurality of corresponding binding-site regions of the plurality of ghost ligands.” [lines 13-14]. There is insufficient antecedent basis for “the plurality of corresponding binding-site regions of the plurality of ghost ligands.” It is recommended to amend the claimed step to recite “a plurality of corresponding binding-site regions of the plurality of ghost ligands.”
Additionally, the claim sets forth “…plurality of corresponding binding-site regions of” proteins but not for ghost ligands. Furthermore, the projection is between a ligand(s) and protein(s) so it is not clear what relationship the ghost ligand has with this step. Perhaps applicants have made a typographical error and intended to recite “…the plurality of corresponding binding-site regions of the plurality of ghost ligand’s protein.”
Claim 1 recites “generating synthetic training samples” for a selected protein and selected ligand. The claim subsequently recites “selecting the synthetic training samples” from the subset of ghost ligands utilizing the confidence vectors. These steps are not clear because the generated synthetic training samples are for a selected protein and ligand, not comprised of ghost ligands. It is therefore not clear what is meant by “selecting the synthetic training samples from the subset of ghost ligands”.
Moreover, and in the generating confidence vector step, the claim recites generating confidence vectors for a subset of ghost ligands “utilizing the confidence score and by comparing the subset of ghost ligands to the selected ligand.” The step sets forth a use without active steps reciting how that use is achieved. It is unclear how the confidence scores are “utilized” to generate a confidence vector. See MPEP 2173.05(q).
Claims 2-14 and 22-25 are rejected because they fail to provide limitations to overcome the deficiencies of the base claim(s).
Examiner’s Note
It is noted the Applicant is invited to contact Examiner to schedule an interview to discuss amendments with respect to the rejection under 35 U.S.C § 112(b) to expedite further prosecution.
Claim Rejections - 35 USC § 103
The instant rejection is maintained for reason for record in the Office Action mailed 19 November 2025 and modified in view of the amendments filed 19 February 2026. It is noted the amendments received 19 February 2026 necessitated new ground(s) of rejection.
The rejection of claim(s) 1-2, 4, 12, 15, 21-23, and 25 under 35 U.S.C. 103 as being unpatentable over Tong et al. (Patent Pub: US 2012/0239367, Patent Pub Date: 09 September 2012) in view of Zhou et al. (US Patent Pub: US 2007/0244651, Patent Pub Date: 18 Oct 20074) in view of Fan et al. (Patent Pub: US 2018/0341754, Patent Pub Date: 29 Nov 2018) in view of Chiang et al. (The Journal of chemical physics, 2016-12, Vol.145 (23), p.234109-234109) in the Office Action mailed 19 November 2025 is withdrawn in view of the amendments filed 19 February 2026.
The rejection of claim(s) 3 and 24 under 35 U.S.C. 103 as being unpatentable over Tong in view of Zhou in view of Fan in view of Chiang, as applied to claims 1-2, 4, 12, 15, 21-23, and 25 above, and in further view of Shah et al. (US Patent Pub: US 2018/0312999, Patent Pub Date: 01 Nov 2018) in the Office Action mailed 19 November 2025 is withdrawn in view of the amendments filed 19 February 2026.
The rejection of claim(s) 5 and 7 under 35 U.S.C. 103 as being unpatentable over Tong in view of Zhou in view of Fan in view of Chiang, as applied to claims 1-2, 4, 12, 15, 21-23, and 25 above, and in further view of Maggiore et al. (Journal of medicinal chemistry, 2014-04, Vol.57 (8), p.3186-3204) (Cited in the Office Action mailed 20 May 2025) in the Office Action mailed 19 November 2025 is withdrawn in view of the amendments filed 19 February 2026.
The rejection of claim(s) 10 under 35 U.S.C. 103 as being unpatentable over Tong in view of Zhou in view of Fan in view of Chiang, as applied to claims 1-2, 4, 12, 15, 21-23, and 25 above, and in further view of Shi et al. (Genomics (San Diego, Calif.), 2019-12, Vol.111 (6), p.1839-185) (Cited in the Office Action mailed 20 May 2025) in the Office Action mailed 19 November 2025 is withdrawn in view of the amendments filed 19 February 2026.
The rejection of claim(s) 13-14 under 35 U.S.C. 103 as being unpatentable over Tong in view of Zhou in view of Fan in view of Chiang, as applied to claims 1-2, 4, 12, 15, 21-23, and 25 above, and in further view of Da et al. (Journal of chemical information and modeling, 2014-09, Vol.54 (9), p.2555-2561) in the Office Action mailed 19 November 2025 is withdrawn in view of the amendments filed 19 February 2026.
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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claim(s) 1-9, 12, 15, and 21-24 are rejected under 35 U.S.C. 103 as being unpatentable over Tong et al. (Patent Pub: US 2012/0239367, Patent Pub Date: 09 September 2012) in view of Chiang et al. (The Journal of chemical physics, 2016-12, Vol.145 (23), p.234109-234109) in view of Zhou et al. (Patent Pub: US U.S 2008/0059077; Patent Pub Date: 06 March 2008) in view Gonen (Bioinformatics, 2012-09, Vol.28 (18), p.2304-2310) in view of Liu et al. (Liu, Yong ; Wu, Min ; Miao, Chunyan ; Zhao, Peilin ; Li, Xiao-Li).
Claims 1 is drawn to a method for predicting drug-target binding using synthetically augmented data while claim 15 is drawn to non-transitory computer readable medium storing the method of claim 1 and claim 21 is drawn to a system. Claims 1, 15, and 21 recite similar limitations. Therefore, claims 1, 15, and 21 are examined similarly.
Claim 1 recites generating a plurality of ghost ligands for a plurality of proteins in a protein structure database, wherein generating the plurality of ghost ligands.
Claim 1 recites for a cluster of proteins, selected from the plurality of proteins by performing a three-dimensional structure alignment for proteins in the cluster of proteins.
Claim 1 recites obtaining the plurality of ghost ligands by projecting a ligand that interacts with a protein in the cluster of proteins at a binding site onto a plurality of corresponding binding-site regions of other proteins in the cluster of proteins identified by the three-dimensional structure alignment.
Claim 1 recites generating confidence scores representing projection certainties for the ligand and the plurality of corresponding binding-site regions of the plurality of ghost ligands.
Claims 1 recites generating synthetic training samples for training a neural network to generate protein ligand interaction likelihood predictions by, for a selected protein and a selected ligand.
Claim 1 recites selecting, from the plurality of ghost ligands, a subset of ghost ligands corresponding to the selected protein.
Claim 1 recites generating confidence vectors for the subset of ghost ligands utilizing the confidence scores and by comparing the subset of ghost ligands to the selected ligand.
Claim 1 recites selecting the synthetic training samples from the subset of ghost ligands utilizing the confidence vectors.
Claim 1 recites generating a plurality of drug-target interaction (DTI) training features for the synthetic training samples.
Claims 1 recites training the generating a neural network using the plurality of DTI training features for the synthetic training samples.
Clams 1 recites predicting a likelihood of interaction for a combination of a query protein and a query ligand using the neural network.
Tong et al. (Tong) discloses deriving input instances of ligand-receptor interactions with known 3D structures or with theoretical models with the facilitation using machine-learning on 3-D structures or theoretical models for prediction of binding activities between ligands and receptors [Tong, disclosure, page 6, right col para 0080]. Tong discloses a method for using a ligand known to interact with a receptor to generate a plurality of additional ligands (i.e., ghost like ligand) [Tong, claim 1]. Tong discloses generating additional ligands using a source ligand [Tong, claim 1], as in instant claim 1 generating a plurality of ghost ligands for a plurality of proteins in a protein structure database, wherein generating the plurality of ghost ligands.
Tong discloses, in an embodiment, the present invention uses machine learning (i.e., ANN, SVM) on 3D structure or with theorical models (i.e., synthetic training samples) for predicting binding activities between ligand (i.e., drug, hormones, peptides, drugs, small molecule, etc., [Tong, page 1 col para 0003]) and receptors [Tong, page 6 right col para 0080]. Tong discloses representations for each ligand-receptor interaction is in the form LIS: TP-RIS-BA wherein LIS represents the ligand contact elements of the interaction, TP represents the chemical bonds involved in the interaction, RIS represents the receptor contact elements of the interaction and BA represents the strength of the interaction [Tong, claim 13]. Tong discloses training the predictive model [Tong, claim 1], as in instant claim 1 generating a plurality of drug-target interaction (DTI) training features for the synthetic training samples.
Tong discloses, in an embodiment, the present invention can use machine learning (i.e., ANN, SVM) to process 3D structure structures or theorical models (i.e., synthetic training samples) for predicting binding activities between ligand (i.e., drug, hormones, peptides, drugs, small molecule, etc., [Tong, page 1 col para 0003]) and receptors [Tong, page 6 right col para 0080]. Tong discloses characteristics of the ligand-receptor interaction comprised one or more of the following ligand contact elements of the inter action, receptor contact elements of the interaction, chemical bonds involved in the interaction and a strength of the interaction [Tong, claim 12].Tong discloses representations for each ligand-receptor interaction is in the form LIS: TP-RIS-BA wherein LIS represents the ligand contact elements of the interaction, TP represents the chemical bonds involved in the interaction, RIS represents the receptor contact elements of the interaction and BA represents the strength of the interaction [Tong, claim 13], as in claim 1 training a neural network using the plurality of DTI training features for the synthetic training samples.
Tong does not teach claim 1 for a cluster of proteins, selected from the plurality of proteins by performing a three-dimensional structure alignment for proteins in the cluster of proteins.
Tong does not teach claim 1 obtaining the plurality of ghost ligands by projecting a ligand that interacts with a protein in the cluster of proteins at a binding site onto a plurality of corresponding binding-site regions of other proteins in the cluster of proteins identified by the three-dimensional structure alignment.
Tong does not teach claim 1 generating confidence scores representing projection certainties for the ligand and the plurality of corresponding binding-site regions of the plurality of ghost ligands.
Tong does not teach claims 1 generating synthetic training samples for training a neural network to generate protein ligand interaction likelihood predictions by, for a selected protein and a selected ligand.
Tong does not teach claim 1 selecting, from the plurality of ghost ligands, a subset of ghost ligands corresponding to the selected protein.
Tong does not teach claim 1 generating confidence vectors for the subset of ghost ligands utilizing the confidence scores and by comparing the subset of ghost ligands to the selected ligand.
Tong does not teach claim 1 selecting the synthetic training samples from the subset of ghost ligands utilizing the confidence vectors.
Tong does not teach claim 1 predicting a likelihood of interaction for a combination of a query protein and a query ligand using the neural network.
Gonen
Gonen et al. (Gonen) teaches using projecting methods for determining drug-target interactions [page 2305 left col second para]. Gonen teaches exploratory analysis using low dimensional projections. Gonen teaches predicting projections for the held-out drugs and target proteins. Gonen teaches providing interactions between clusters of drugs and target protein (i.e., ion channel, enzymes, GPCR) [page 2308 fig 2]. Gonen teaches clusters of proteins target and drugs [page 2308 fig 2]. Zhou discloses identifying binding pockets [Zhou, claim 1] and 3D-structure alignment [Zhou, claim 12], as in claim 1 obtaining a plurality of ghost ligands by projecting a ligand that interacts with a protein in a cluster of proteins at a binding site onto a plurality of binding sites corresponding binding site of regions of other proteins in the cluster of proteins identified by the three-dimensional alignment.
Gonen teaches projecting drug compounds and target proteins [page 2308 fig 2]. Gonen teaches prediction for a new target protein and a joint prediction for a new drug compound and a new target protein (i.e., confidence scores based on representing projections) [page 2307 right col]. Zhou discloses identifying binding pockets [Zhou, claim 1] and 3D-structure alignment [Zhou, claim 12], as in claim 1 generating obtaining, for each of the plurality of ghost ligands, a confidence scores representing projection certainties for the ligand and the plurality of corresponding binding-site regions of the plurality of ghost ligands.
Dependent claim(s): 4
Gonen teaches projecting drug compounds and target proteins [page 2308 fig 2]. Gonen teaches prediction for a new target protein and a joint prediction for a new drug compound and a new target protein (i.e., confidence scores quantifies uncertainty). [page 2308 fig 2] as in instant claim 4.
Zhou
Zhou et al. (Zhou) using sets of three-dimensional proteins based on local homologous based on three-dimensional protein structure and global alignment [Zhou, claim 12]. Zhou discloses sequence and structure-based clustering [Zhou, page 10 right col para 0135-0136], as in instant claim 1 for a cluster of proteins, selected from the plurality of proteins, performing a three-dimensional structure alignment for proteins in the cluster of proteins. Here, it is obvious the three-dimensional structure alignment was performed on the set of aligned three-dimensional structure of the target polypeptide and residues comprising a near-neighbor polypeptide.
Dependent claim(s): 3, 7, 9, 12, and 22-24.
Zhou discloses analyzing similarity of sequences [Zhou, claim 1]. Zhou discloses spans or domains maybe identified by Class Architechture Topology Homology database [Zhou, page 7, para 0085], as in instant claim 3.
Zhou discloses using predetermined distances for one-to-one set of corresponding residues from the aligned three-dimensional structures [Zhou, claim 10]. Zhou discloses using less than 5 angstroms for predetermined distances [Zhou, claim 11]. Zhou discloses selecting a set of target pockets using cluster similarities [Zhou, claim 2] as in instant claim 7.
Zhou discloses in the method a set of superpositions are created in order to detect regions are most similar. Zhou discloses using LGA scoring function has two components, LCS (longest continuous segments) and GDT (global distance test), established for the detection of regions of local and global structure similarities between proteins [Zhou, page 13 left col. para 0168], as in claim 9.
Zhou discloses structure motif module uses scored residues in the span correspondence to indefinity the subset of residues in the span which form a motif based on cutoff conservation score value [Zhou, page 7 left col para 0089], as in instant claim 12.
Zhou discloses generating a cluster based on a plurality of pockets similarity values in a target pocket [Zhou, claim 2]. Zhou discloses three-dimensional structure alignments [Zhou, claim 12]. Zhou discloses using topological database Class Architecture Topology Homology database and using CATH structures grouped by secondary structure, gross orientation of secondary structure (architecture), topology (folds and connections) and homology [Zhou, page 7 left col para 0085], as in instant claim 22.
Zhou discloses generating clusters based on pocket similarity values [Zhou, claim 2]. Zhou discloses using Pocket Cluster Module that generates representative binding profiles by combining the binding profiles of a set of pockets associated with a protein family, a group of organisms or any other user specified set of pockets [Zhou, page 8, right col 0106], as in instant claim 23.
Zhou discloses generating a cluster based on a plurality of pockets similarity values in a target pocket [Zhou, claim 2]. Zhou discloses three-dimensional structure alignments [Zhou, claim 12]. Zhou discloses using topological database Class Architecture Topology Homology database and using CATH structures grouped by secondary structure, gross orientation of secondary structure (architecture), topology (folds and connections) and homology [Zhou, page 7 left col para 0085]. Zhou discloses generating clusters based on pocket similarity values [Zhou, claim 2]. Zhou discloses using Pocket Cluster Module that generates representative binding profiles by combining the binding profiles of a set of pockets associated with a protein family, a group of organisms or any other user specified set of pockets [Zhou, page 8, right col 0106]. Zhou discloses database which contains protein families, functional sites and domains which may be applied to characterize unknown protein sequences [Zhou, page 8 left col para 0099], as in instant claim 24. Here, it would be obvious for a user to manually create a cluster of known and unknown proteins that share similarities such as topology.
Obvious claim(s): 2, 5-6, and 8
Zhou discloses aligning the three-dimensional structural alignments on proteins [Zhou, claim 12]. Zhou discloses using a Protein Set Selection Module for selecting of a group of known homologs [Zhou, page 9 left para 0117]. Gonen teaches projection [page 2305 left col second para], as in claim 2.
Tong discloses, in an embodiment, the present invention uses machine learning (i.e., ANN, SVM) on 3D structure or with theorical models (i.e., synthetic training samples) for predicting binding activities between ligand (i.e., drug, hormones, peptides, drugs, small molecule, etc., [Tong, page 1 col para 0003]) and receptors [Tong, page 6 right col para 0080].Gonen teaches projection of ligands interactions with target proteins (i.e., ghost ligands) [page 2308 fig 2]. Tong discloses representations for each ligand-receptor interaction is in the form LIS: TP-RIS-BA wherein LIS represents the ligand contact elements of the interaction, TP represents the chemical bonds involved in the interaction, RIS represents the receptor contact elements of the interaction and BA represents the strength of the interaction [Tong, claim 13]. Tong discloses training the predictive model [Tong, claim 1]. Zhou discloses generating a cluster based on pocket similarities and selecting a set of target pockets based on the cluster [Zhou, claim 2]. Zhou discloses using a Pocket Binding Module that uses methods of computational protein ligand docking to evaluate binding of a set of ligands or Small molecules to a pocket [Zhou, page 8 right col para 0104], as in instant claim 5.
Zhou discloses using global alignment of homologous three-dimensional protein structures or any combination thereof [Zhou, claim 12], as in instant claim 6.
Gonen et al. (Gonen) teaches using projecting methods for determining drug-target interactions [page 2305 left col second para]. Gonen teaches exploratory analysis using low dimensional projections. Gonen teaches predicting projections for the held-out drugs and target proteins. Gonen teaches providing interactions between clusters of drugs and target protein (i.e., ion channel, enzymes, GPCR) [page 2308 fig 2]. Gonen teaches clusters of proteins target and drugs [page 2308 fig 2]. Zhou discloses identifying binding pockets [Zhou, claim 1] and 3D-structure alignment [Zhou, claim 12], Zhou discloses colored bars represent distance deviation of the alpha carbons between superimposed PDB structures and Dengue virus (PDB code: 2fom B). Zhou discloses the alignment is 150 residues in length from the left (N terminal) to the right (C terminal)) [Zhou, page 10 left para 0139], as in instant claim 8.
Liu
Liu et al. (Liu) teaches the top 30 novel interactions predicted by NRLMF on GPCR dataset of different drug and target interactions and also provides predicted probabilities of interaction between drug (i.e., D00283) and target (i.e., hsa1814) [Liu, page 21, table 5].Gonen teaches projection of ligands interactions with target proteins (i.e., ghost ligands) [page 2308 fig 2] , as in instant claim 1 selecting a subset of ligands corresponding to a selected protein and predicting a likelihood of interaction for a combination of a query protein and a query ligand using the neural network.
Liu et al. (Liu) teaches a method that focuses on predicting the probability a drug would interact with a target. Liu teaches using the properties of a drug and a target are represented by two latent vectors in the shared low dimensional latent space with for each drug target pair, the interaction probability (certainty) is modeled by a logistic function of the drug-specific and target-specific latent vectors [Liu, page 3 third para].Gonen teaches projection of ligands interactions with target proteins (i.e., ghost ligands) [page 2308 fig 2], as in instant claim 1 generating confidence vectors for the subset of ghost ligands utilizing the confidence scores and by comparing the subset of ghost ligands to the selected ligand.
Liu teaches the top 30 novel interactions predicted by NRLMF on GPCR dataset of different drug and target interactions and also provides predicted probabilities of interaction between drug (i.e., D00283) and target (i.e., hsa1814) [Liu, page 21, table 5]. Gonen teaches a joint prediction for new drug and new target protein interactions [Gonen, page 2307 right col section 3.3] as in instant claim 1 predicting a likelihood of interaction for a combination of a query protein and a query ligand using the neural network.
Obvious claim 1 step(s):
Gonen teaches projecting drug compounds and target proteins [page 2308 fig 2]. Gonen teaches prediction for a new target protein and a joint prediction for new drug compounds and a new target protein [page 2307 right col]. Tong discloses using at least one training ligand and a training receptor are identified and using these training ligands and receptors or database management system. Tong discloses a predictive model is trained using the representations of the ligand-receptor interactions [Tong, disclosure, page 2 para 0020]. Tong discloses the predictive model can be trained using artificial neural network [page 5 left para 0062]. Tong discloses using theoretical models for training sets and tests [Tong, disclosure page 6 left col table 2 and right col para 0080]. Liu teaches using CVS1: CV on drug-target pairs as random entries Y and using Y elements as training data and the remaining 10% of the element were used as test data [page 9]. Liu teaches unknown interactions are ranked using interaction probabilities. Liu teaches the probability of interactions between drug and target pairs [page 21 table 5], as in claim 1 generating synthetic training samples for training a neural network to generate protein ligand interaction likelihood predictions by, for a selected protein and a selected ligand. Thus, it would be obvious to use the data of Tong and Liu to construct a theoretical dataset containing ligands and proteins for generating protein-ligand interactions.
Tong discloses using at least one training ligand and a training receptor are identified and using these training ligands and receptors or database management system. Tong discloses a predictive model is trained using the representations of the ligand-receptor interactions [Tong, disclosure, page 2 para 0020]. Tong discloses the predictive model can be trained using artificial neural network [page 5 left para 0062]. Liu teaches the top 30 novel interactions [page 21 tables 5]. Liu teaches using positive and negative training samples [page 5 In DTI prediction para]. Tong discloses using at least one training ligand and a training receptor are identified and using these training ligands and receptors or database management system. Tong discloses a predictive model is trained using the representations of the ligand-receptor interactions [Tong, disclosure, page 2 para 0020]. Tong discloses the predictive model can be trained using artificial neural network [page 5 left para 0062].Gonen teaches projection of ligands interactions with target proteins (i.e., ghost ligands) [page 2308 fig 2]. Liu teaches using the properties of a drug and a target are represented by two latent vectors in the shared low dimensional latent space with for each drug target pair, the interaction probability (certainty) is modeled by a logistic function of the drug-specific and target-specific latent vectors [Liu, page 3 third para], as in instant claim 1 selecting the synthetic training samples from the subset of ghost ligands utilizing the confidence vectors. Here, is it obvious the confidence vectors were utilized for selection of the top 30 novel interactions because the probability analysis of the selected top 30 novel interactions required using the modeled interaction probability (certainty) uses latent vectors.
It would be obvious to one of ordinary skill in the art by the effective filing date of the claimed invention to modify Tong in view of Gonen because Gonen teaches methods for projecting drug ligands and target projection [Gonen, page 2308 fig 2]. One of ordinary skill in the art would recognize Gonen teaches methods that can be used to project the additional ligands of Tong/Gonen to target protein receptors of Tong/Gonen to yield ghost ligands. Thus, one of ordinary skill in the art would be motivated to combine Tong in view of Gonen because Gonen provides different prediction methods for drug target interactions such as predictions for new drug compound, prediction for new target protein, and a joint prediction for a new drug and a new target protein [page 2307 section 3.3] which can used to determine/predict drug-target interactions (DTI’s). Therefore, combining the known elements of ligand-receptor binding affinity analysis of Tong in view of the projection methods of Gonen would construct a predictable method step for obtaining ghost ligands for DTI analysis.
It would be obvious to one of ordinary skill in the art by the effective filing date of the claimed invention to modify Tong in view Gonen in view of Zhou because Zhou discloses methods for analyzing protein target pockets for broad spectrum drug development using three-dimensional pocket concavity in protein structure and three-dimensional structure alignment methods. Here, one of ordinary skill in the art be motivated to combine Tong in view of Gonen in view of Zhou because Zhou discloses methods for evaluating protein target pockets by different parameters such as similarity, clustering, and binding profiles which could be utilized for analyzing either protein ligands with respect to a protein target receptor or with respect to small molecule drug(s) to protein target receptor. One of ordinary skill in the art would be motivated to combine Tong in view of Gonen in view of Zhou because the analysis of Zhou in conjunction with the method of Tong for predicting ligand and receptor binding affinity could be used to construct a method step for performing three-dimensional structure alignment of protein receptors to determine binding properties of a ligand and receptor (i.e., drug-target interactions). Thus, one of ordinary skill in the art would have a reasonable expectation success combining ligand-receptor binding affinity analysis of Tong, the projection analysis of Gonen (i.e., ghost ligands), and the pocket analysis of Zhou because Zhou’s target protein pockets analysis is used for drug development with respect to ligand-receptors interactions. Therefore, one of ordinary skill in the art would combine the known elements of Tong, Gonen, and Zhou to construct a predictable method step for performing three-dimensional structure alignment processing of ligands/ligand proteins and target proteins for determining binding affinities for subsequent determination of drug-target interactions using synthetic data.
It would be obvious to one of ordinary skill in the art by the effective filing date of the claimed invention to modify Tong in view of Gonen view of Zhou in view of Liu because Liu teaches a mathematical analysis of drug target interactions [Liu, title]. One of ordinary skill in the art would recognize that Tong, Gonen, Zhou, and Liu are in similar field of endeavor (i.e., drug-target/ligand-receptor interaction analysis). One of ordinary skill in the art would be motivated to combine Tong in view of Gonen in view of Zhou in view of Liu because Liu teaches analyzing drug-target pairs and modeling interaction probabilities using logistic function drug-specific and target-specific latent vectors [Liu, page 3 third para]. One of ordinary skill in the art would be motivated to combine Tong in view of Gonen in view of Zhou in view of Liu because Liu teaches the mathematical concepts using matrices and vectors for determining probabilities of DTI’s [Liu, pages 11-12 table 2-4]. Thus, one of ordinary skill in the art would have a reasonable expectation of success combining affinity assessment of Tong, ghost ligand projection method of Gonen, and pocket analysis of Zhou with the matrices and vector concepts of Liu because Liu teaches analyzing proteins and ligand interactions for selecting 30 novel DTI. Therefore, combining the binding affinity prediction of Tong, the mathematical theory and concepts of Gonen and Liu, and the identification of protein pocket properties for drug develop of Zhou would construct a predictable method step for obtaining confidence vectors for comparing subsets of ligands for determining DTI’s using synthetic data.
Claim(s) 10-11, 13-14, and 25 are rejected under 35 U.S.C. 103 as being unpatentable over Tong in view Gonen in view of Zhou in view of Liu, as applied to claims 1-9, 12, 15, and 21-24 above, and in further view of in further view of Shi et al. (Genomics (San Diego, Calif.), 2019-12, Vol.111 (6), p.1839-185).
Tong in view of Gonen in view of Zhou in view of Liu teaches claims1-8, 12, 15, and 21-24.
Tong in view of Gonen in view of Zhou in view of Liu teach a method for predicting ligand-target (i.e., drug-target) binding that generates ghost ligands, ghost ligand drug-target interactions (DTI), and generating a neural network using DTI features for predicting a likelihood of interaction between a ligand and protein.
Tong in view of Gonen in view of Zhou in view of Liu does not teach claims 10-11, 13-14, and 25.
Shi et al. (Shi) teach using chemical structure and amino acid sequences for generating positive samples [page 1844 fig 1]. Shi teaches feature selection using known and unknown interactions and processing the interaction that are processed using Lasso, PCA, ReliefF, and Elastic net which is further processes using SMOTE method to produce positive and negative samples [page 1844 fig 1]. Shi teaches Synthetic Minority Oversampling Technique (SMOTE) is a method for randomly under sampling a large scale of samples while randomly oversampling a small scale of samples [page 1843 left col top para]. Shi teaches SMOTE processes random oversampling on a small scale of positive samples and random under sampling on a large scale of negative samples so that the positive and negative samples are balanced [page 1842]. Liu teaches using positive and negative training samples [page 5 In DTI prediction para]. Tong discloses, in an embodiment, the present invention can use machine learning (i.e., ANN, SVM) to process 3D structure structures or theorical models (i.e., synthetic training samples) for predicting binding activities between ligand (i.e., drug, hormones, peptides, drugs, small molecule, etc., [Tong, page 1 col para 0003]) and receptors [Tong, page 6 right col para 0080], as in instant claim 10.
Shi teaches removing useless information and extracting the most discriminating features from the descriptors of drug-target pair using the Lasso method to reduce the dimensionality of the original data features [Shi, page 1842 right col section 2.6]. Tong discloses, in an embodiment, the present invention uses machine learning (i.e., ANN, SVM) on 3D structure or with theorical models (i.e., synthetic training samples) for predicting binding activities between ligand (i.e., drug, hormones, peptides, drugs, small molecule, etc., [Tong, page 1 col para 0003]) and receptors [Tong, page 6 right col para 0080]. Tong discloses representations for each ligand-receptor interaction is in the form LIS: TP-RIS-BA wherein LIS represents the ligand contact elements of the interaction, TP represents the chemical bonds involved in the interaction, RIS represents the receptor contact elements of the interaction and BA represents the strength of the interaction [Tong, claim 13]. Tong discloses training the predictive model [Tong, claim 1]. Liu et al. (Liu) teaches a method that focuses on predicting the probability a drug would interact with a target. Liu teaches using the properties of a drug and a target are represented by two latent vectors in the shared low dimensional latent space with for each drug target pair, the interaction probability (certainty) is modeled by a logistic function of the drug-specific and target-specific latent vectors [Liu, page 3 third para]. Liu teaches the top 30 novel interactions predicted by NRLMF on GPCR dataset of different drug and target interactions and also provides predicted probabilities of interaction between drug (i.e., D00283) and target (i.e., hsa1814) [Liu, page 21, table 5] as in instant claim 11. Here, it is obvious that DTI’s were filtered using confidence vectors because Liu teaches predicting DTI using interaction probabilities that encompasses latent vectors and teach the top 30 novel interactions were selected based on the probabilities [Liu page 17 page 21 table 5].
Furthermore, based on the obviousness of claim 11, it would be obvious that once the data is filtered and removed the data would be ignored and thus not providing a signal to the neural network about obvious patterns, as in instant claim 25.
Tong discloses, in an embodiment, the present invention can use machine learning (i.e., ANN, SVM) to process 3D structure structures or theorical models (i.e., synthetic training samples) for predicting binding activities between ligand (i.e., drug, hormones, peptides, drugs, small molecule, etc., [Tong, page 1 col para 0003]) and receptors [Tong, page 6 right col para 0080]. Tong discloses characteristics of the ligand-receptor interaction comprised one or more of the following ligand contact elements of the inter action, receptor contact elements of the interaction, chemical bonds involved in the interaction and a strength of the interaction [Tong, claim 12].Tong discloses representations for each ligand-receptor interaction is in the form LIS: TP-RIS-BA wherein LIS represents the ligand contact elements of the interaction, TP represents the chemical bonds involved in the interaction, RIS represents the receptor contact elements of the interaction and BA represents the strength of the interaction [Tong, claim 13]. Shi teaches steps for FP2 molecular fingerprinting and teaches fingerprinting between positive and negative dataset [page 1842 section 2.3 and section 2.5], as in instant claim 13. Here, it would be obvious to use the fingerprinting methods of Shi to processes the ligand and receptors descriptors of Tong.
Zhou discloses determining a set of three-dimensional protein structures based on local alignment and global alignment of homologous three-dimensional structures [Zhou, claim 12]. Zhou discloses determining functional annotations [Zhou, Fig 7]. Zhou discloses a Structure Motif Module that identify motifs to cluster based on Sequence Activity Relationship information or functional sequence annotations generated by the Sequence Activity Relationship Module [Zhou, page 7 left col para 0091], as in claim 14.
It would be obvious to one of ordinary skill in the art by the effective filing date of the claimed invention to modify Tong in view Gonen in view of Zhou in view of Liu,, and in further view of Shi because Shi teaches analyzing drug-target interactions using Lasso with random forest based on evolutionary information and chemical structure [title]. One of ordinary skill would be motivated to combine Tong in view Gonen in view of Zhou in view of Liu, and in further view of Shi because Shi teaches using chemical data (i.e., ligand/drug) and amino acid data (i.e., target or protein), and combining and separating the interaction data into positive and negative datasets for predicting drug-target interactions (DTI). Thus, there would be a reasonable expectation of success combining the methods of Tong in view Gonen in view of Zhou in view of Liu,, and in further view of Shi to construct a DTI system that can predict DTI interaction because Shi teaches methods using random forest machine learning for predicting DTI using positive and negative datasets. Here, combining the methods Tong, Gonen, Zhou, and Liu and Shi would yield predictable method steps for predicting drug-target binding that uses positive and negative datasets for determining DTI’s using synthetic data.
Response to Arguments
Applicant’s arguments, filed 19 February 2026, have been fully considered and the rejection is maintained. However, upon further consideration, a new ground(s) of rejection is made in view of amendments received 19 February 2026. Here, Tong, Gonen, Zhou, Liu, and Shi address the amendments received 19 February 2026.
Conclusion
Claims 1-15 and 21-25 are rejected.
No claims are allowed.
Finality
This Office action is a Non-Final action. A shortened statutory period for reply to this action is set to expire THREE MONTHS from the mailing date of this action.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JOSEPH C PULLIAM whose telephone number is (571)272-8696. The examiner can normally be reached 0730-1700 M-F.
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, Karlheinz Skowronek can be reached at (571) 272-9047. 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.C../ Examiner, Art Unit 1687
/Anna Skibinsky/
Primary Examiner, AU 1635
Prosecution Insights