DETAILED ACTION
Applicant’s response, filed 4/23/2026, has been fully considered. The following rejections and/or objections are either reiterated or newly applied. They constitute the complete set presently being applied to the instant application.
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 1-10 are currently pending and examined on the merits.
Claims 1-10 are rejected.
Priority
The instant application claims foreign priority to Application CN 2022105370156 filed on 17 May 2022, in China. At this point in examination, the effective filing date of claims 1-10 is 17 May 2022.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 7 July 2022 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.
The listing of references in the specification is not a proper information disclosure statement. 37 CFR 1.98(b) requires a list of all patents, publications, or other information submitted for consideration by the Office, and MPEP § 609.04(a) states, "the list may not be incorporated into the specification but must be submitted in a separate paper." Therefore, unless the references have been cited by the examiner on form PTO-892, they have not been considered.
Claim Objections
The objection to claim 10 is withdrawn, in view of the claim amendments.
New ground for objection was necessitated by amendment to the claims, received 4/23/2026. Claim 9 is objected to because of the following informalities:
In claim 9, line 16, “using a deep learning model obtain sequence” should read “using a deep learning model to obtain sequence”.
There is a typographical error. Appropriate correction is required.
Claim Interpretation – 35 USC § 112(f)
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f):
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f). The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f). The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) because the claim limitations use a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitations are:
A: “a calculation module configured to: evaluate a mutation space of the at least one of the antibody and the macromolecular drug based on the template sequence information and modification requirements; narrow a mutation range for screening if the mutation space exceeds a predetermined upper limit” in claims 1-3.
Because these claim limitations are being interpreted under 35 U.S.C. 112(f), they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
Para. [0102] of the published specification indicates that the system provided by the present invention and its various devices, modules, and units can be regarded as hardware components, and the devices, modules, and units included in the system for implementing various functions can also be regarded as structures within the hardware components. MPEP 2181.II.B. indicates that for computer-implemented means-plus-function limitations, the structure is an algorithm coupled with a microprocessor or computer. The above paragraph provides support for the processor or computer. The individual algorithms for each 112(f) invocation are detailed as follows:
A: “a calculation module” - para. [0092] of the instant specification states steps for a calculation module that evaluate the mutation space of the antibody and narrow the mutation range for screening if the mutation space exceeds an upper limit. The calculation module is also described to preprocess candidate mutation amino acid sequences one by one, calculate antibody antigen affinities based on a deep learning model and score them, where either the highest or lowest affinity sequence becomes the final modification sequence. Therefore, the instant specification discloses the specific steps of an algorithm. Thus, the description in the specification for the claimed calculation module has adequate corresponding structure.
If applicant does not intend to have these limitations interpreted under 35 U.S.C. 112(f), applicant may: (1) amend the claim limitations to avoid them being interpreted under 35 U.S.C. 112(f) (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitations recite sufficient structure to perform the claimed function so as to avoid them being interpreted under 35 U.S.C. 112(f).
Response to Arguments
Applicant’s arguments, see pages 7-8, filed 4/23/2026, with respect to claims 1-8 have been fully considered and are persuasive. The interpretation of the claims under 35 U.S.C. § 112(f) has been withdrawn in view of the amendments to the claims, filed 4/23/2026.
However, Applicant argues that the amended claims now recite the operational workflow performed by the calculation module, thereby providing sufficient acts for performing the claimed function (pg. 7-8, para. 5 of Applicant’s Remarks). As stated above, the structure of computer-implemented means-plus-function limitations requires an algorithm to be coupled with a microprocessor or computer components. Therefore, the calculation module recited in claims 1-3 is interpreted under 35 U.S.C. 112(f).
Claim Rejections - 35 USC § 112(a)
The previous rejections to claims 1-8 under 35 U.S.C. § 112(a) are withdrawn in view of the claim amendments.
Claim Rejections - 35 USC § 112(b)
The previous rejections to claims 1-8 under 35 U.S.C. § 112(b) are withdrawn in view of the claim amendments.
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 1-10 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 1-8 are directed to a system (machine). Claims 9-10 are directed to a method (process) for affinity modification of antibody/macromolecular drug. 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.
Claims 1, 5, and 8-10 recite the following steps which fall within the mental processes and/or mathematical concepts groups of abstract ideas, as noted below.
Independent claim 1 further recites:
a calculation module configured to: evaluate a mutation space of the at least one of the antibody and the macromolecular drug based on the template sequence information and modification requirements (i.e., mental processes);
a calculation module configured to: narrow a mutation range for screening if the mutation space exceeds a predetermined upper limit (i.e., mental processes);
perform at least one of corresponding partial and exhaustive numeration of possible sequences in a part of a full variable range to obtain a mutation library (i.e., mental processes);
perform a sequence-based affinity prediction on candidate sequences within the mutation library (i.e., mental processes).
Dependent claim 5 further recites:
at least one element of a set comprising marking the variable range and specifying the variable range (i.e., mental processes);
defining a modification direction (i.e., mental processes).
Dependent claim 8 further recites:
wherein, the visual analysis display module further comprises a comparative analysis of the template sequence information of the at least one of the antibody and the macromolecular drug and the sequence information of the at least one of the modified antibody and the modified macromolecular drug in a variable range (i.e., mental processes).
Dependent claim 9 further recites:
evaluate a mutation space of the at least one of the antibody and the macromolecular drug based on the template sequence information and modification requirements (i.e., mental processes);
narrow a mutation range for screening if the mutation space exceeds a predetermined upper limit (i.e., mental processes);
perform at least one of: corresponding partial and exhaustive numeration of possible sequences in a part of a full variable range to obtain a mutation library (i.e., mental processes);
perform a sequence-based affinity prediction on candidate sequences within the mutation library (i.e., mental processes).
Dependent claim 10 further recites:
when performing the at least one of partial and exhaustive numeration of possible sequence in a part of the full variable range (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 2-4, 6, and 10 recite information further limiting the judicial exceptions indicated above.
Therefore, claims 1, 5, and 8-10 recite 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)).
The judicial exceptions identified in Eligibility Step 2A, Prong One are not integrated into a practical application because of the reasons noted below.
Claims 5 and 10 do not recite any elements in addition to the judicial exception, and thus are part of the judicial exception.
Claims 1 and 9 recite perform a sequence-based affinity prediction on candidate sequences within the mutation library using a deep learning model. The limitation recites “using a deep learning model”, which provides nothing more than mere instructions to implement an abstract idea on a generic computer. See MPEP 2106.05(f). Therefore, the claimed additional element does not integrate the abstract ideas into a practical application.
Claims 1, 7, and 9 recite the additional non-abstract elements of data gathering:
a user input interface set to: input template sequence information of at least one of the antibody and the macromolecular drug, modification requirements of at least one target of at least one of the antibody and the macromolecular drug and optional user-defined screening requirements to generate interaction sequence information of the at least one of the antibody and the macromolecular drug (claim 1);
an output module configured to output the sequence information of the at least one of the modified antibody and the modified macromolecular drug (claim 1);
the visual analysis display module provides the complete sequence information of at least one of the modified antibody and the modified macromolecular drug (claim 7);
output the sequence information of the at least one of the modified antibody and the modified macromolecular drug (claim 9).
which are each a data gathering step, or a description of the data gathered.
Data gathering steps are not an abstract idea, they are extra-solution activity, as they collect the data needed to carry out the JE. The data gathering does not impose any meaningful limitation on the JE, or how the JE is performed. The additional limitation (data gathering) must have more than a nominal or insignificant relationship to the identified judicial exception. (MPEP 2106.04/.05, citing Intellectual Ventures LLC v. Symantee Corp, McRO, TLI communications, OIP Techs. Inc. v. Amason.com Inc., Electric Power Group LLC v. Alstrom S.A.).
Claim 1 recites the additional non-abstract element (EIA) of a general-purpose computer system or parts thereof:
an affinity modification system (claim 1);
a user input interface (claim 1).
The EIA do not provide any details of how specific structures of the computer elements are used to implement the JE. The claims require nothing more than a general-purpose computer to perform the functions that constitute the judicial exceptions. The computer elements of the claims do not provide improvements to the functioning of the computer itself (as in DDR Holdings, LLC v. Hotels.com LP); they do not provide improvements to any other technology or technical field (as in Diamond v. Diehr); nor do they utilize a particular machine (as in Eibel Process Co. v. Minn. & Ont. Paper Co.). Hence, these are mere instructions to apply the JE using a computer, and therefore the claim does not recite integrate that JE into a practical application.
Thus, the additionally recited elements merely invoke a computer as a tool, and/or amount to insignificant extra-solution data gathering activity, and as such, when all limitations in claims 1-10 have been considered as a whole, the claims are deemed to not recite any additional elements that would integrate a judicial exception into a practical application. Claims 1, 7, and 9 contain additional elements that would not integrate a judicial exception into a practical application and are further probed for inventive concept in Step 2B.
[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).
The claims do not include any additional elements that are sufficient to amount to significantly more than the judicial exception(s) because of the reasons noted below.
With respect to claims 1, 7, and 9: The limitations identified above as non-abstract elements (EIA) related to data gathering do not rise to the level of significantly more than the judicial exception. Activities such as data gathering do not improve the functioning of a computer, or comprise an improvement to any other technical field. The limitations do not require or set forth a particular machine, they do not affect a transformation of matter, nor do they provide an unconventional step (citing McRO and Trading Technologies Int’l v. IBG). Data gathering steps constitute a general link to a technological environment. Simply appending well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception are insufficient to provide significantly more (as discussed in Alice Corp.,).
With respect to the recited affinity modification system and user input interface in claim 1: The limitations identified above as non-abstract elements (EIA) related to general-purpose computer systems do not rise to the level of significantly more than the judicial exception. These elements do not improve the functioning of the computer itself, or comprise an improvement to any other technical field (Trading Technologies Int’l v. IBG, TLI Communications). They do not require or set forth a particular machine (Ultramercial v. Hulu, LLC., Alice Corp. Pty. Ltd v. CLS Bank Int’l), they do not affect a transformation of matter, nor do they provide an unconventional step. Simply appending well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception are insufficient to provide significantly more (as discussed in Alice Corp., CyberSource v. Retail Decisions, Parker v. Flook, Versata Development Group v. SAP America).
The additional element of perform a sequence-based affinity prediction on candidate sequences within the mutation library using a deep learning model (claims 1 and 9) is conventional. Evidence for conventionality is shown by Dhakal et al. (Briefings in Bioinformatics, 2021, 23(1), 1-23). Dhakal et al. reviews a list of deep learning methods for predicting protein-ligand binding affinity, including a convolutional neural network that uses only sequence information of both targets and drugs to predict drug target interaction binding affinities (pg. 15, Table 5). This shows that there are several deep learning models that predict affinities. Therefore, the deep learning model is a conventional element in the art.
[Step 2B: NO]
Therefore, claims 1-10 are patent ineligible under 35 U.S.C. § 101.
Response to Arguments
Applicant's arguments, see pages 11-15, filed 4/23/2026, with respect to claims 1-10, have been fully considered but they are not persuasive.
With respect to the Applicant’s argument that the amended claims 1 and 9 do not recite abstract ideas, and instead define a structured computational workflow for exploring and optimizing large sequence spaces that cannot be performed mentally or by using pen and paper (pg. 12-13 of Applicant’s Remarks), this argument is not persuasive. The claimed operations recite evaluating mutation spaces, narrowing mutation ranges based on a threshold limit, enumerating sequences, and performing affinity predictions on sequences, all of which are judgements that can be made mentally or by using pen and paper. Furthermore, the size of mutation search spaces does not prevent one skilled in the art from performing these steps, regardless of time. Therefore, the claims are directed to an abstract idea.
With respect to the Applicant’s argument that the amended claims 1 and 9 integrate into a practical application and recite elements that operate together as a coordinated computational architecture to automate large-scale sequence optimization, providing a technological improvement in computational protein engineering (pg. 13-14 of Applicant’s Remarks), this argument is not persuasive. The elements recited in the amended claims are directed to abstract ideas, regardless of being operated together. Furthermore, claimed additional elements are observed for integration of judicial exceptions into a practical application in Step 2A, Prong Two. In this case, using a deep learning model to perform sequence-based affinity predictions is merely using a computer as a tool to perform an abstract idea. Inputting template sequence information, modification requirements, and optional user-defined screening requirements is mere data gathering. Therefore, the claims do not integrate into a practical application.
With respect to the Applicant’s argument that the amended claims 1 and 9 recite technical elements that are significantly more than any alleged abstract idea (pg. 14 of Applicant’s Remarks), this argument is not persuasive. Step 2B evaluates whether additional elements amount to an inventive concept. In this case, the elements recited in the amended claims are directed to abstract ideas. Furthermore, using a deep learning model to perform sequence-based affinity predictions and inputting template sequence information, modification requirements, and optional user-defined screening requirements are well-understood, routine, and conventional activities as explained above.
Therefore, the rejection to claims 1-10 under 35 U.S.C. § 101 is maintained with modifications as necessitated by amendment of the claims, filed 4/23/2026.
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 1-10 are rejected under 35 U.S.C. 103 as being unpatentable over Kang et al. (arXiv preprint, 2021, 1-9), in view of Warszawski et al. (PLOS Computational Biology, 2019, 16(10), 1-24).
With respect to claim 1:
Regarding the recited a user input interface set to: input template sequence information of at least one of the antibody and the macromolecular drug, modification requirements of at least one target of at least one of the antibody and the macromolecular drug, and optional user-defined screening requirements to generate interaction sequence information of the at least one of the antibody and the macromolecular drug, Kang et al. discloses “Antibody-Bind (AB-Bind) is a manually curated and organized database that includes 1101 mutants across 32 complexes” (Page 2, Section “2.1 Data Collection”, line 1). Also, further discloses “we explore the graph representation of antibody-antigen complex with three different representation strategies. 1) Full-seq model: the full-seq model simply takes antibody and antigen sequences as two separated graph sequences (Figure 2(A)). The intuition is to incorporate both interact contacts (ICs) and non-interacting surface (NIS) into modeling as the binding strength between antibody and antigen relies on the full conformation of the formed complex [9, 12]. 2) Contacts-only model: the contacts-only model produces a compact representation of the complex by utilizing residues on the interfacial surface (distance <5 Angstrom) only (Figure 2(B)); Given limited high-quality training data, this approach is presumably more adequate since it models the most relevant information for antibody-antigen interactions. The identification of interface contacts was obtained using Prodigy-based prediction service [9] based on complex’s 3D structure. 3) Antibody-only model: the antibody-only model aims to address the promiscuous binding capability of antibodies cross diverse antigens [15, 16, 17, 18]. We investigated on antibody-only modeling for binding affinity prediction, and evaluate if Hag-Net based network structure captures the enabling features for antibodies’ natural binding capability.” (Page 3, Section “2.2.2 Baseline study: binding affinity change (
∆
∆
G
) prediction”, paragraph 2, lines 1-12). Kang et al. discloses “Antibody maturation aims to optimize the binding affinity based on known antibody leads targeting specific antigens. To this purpose, we construct the pairwise problem and study the binary relations between mutated variants of each therapeutic lead. Specifically, for each wildtype antibody, its associated mutations are grouped into unique ordered pairs. Each pair consists of two complexes with same target antigen but different mutated variants. If the first complex ranks higher than the second complex with respect to their binding affinities (i.e. the first complex possesses higher affinity), the pair is labeled as 1, otherwise as 0. Therefore, our goal is to obtain classifier f:
f
a
,
b
=
1
i
f
∆
∆
G
a
<
∆
∆
G
b
0
e
l
s
e
” (Page 3, Section “2.2.3 Pairwise-study: binding affinity pairwise rank prediction”, paragraph 1, lines 1-7). Antibody-Bind (AB-Bind) is a database of mutants across complexes, which teaches an input template sequence information of antibody drugs. The three graph representation strategies for antibody-antigen complexes indicate modification requirements for at least one target of antibody drugs because they model bindings and interactions between antibodies and antigens. The pairwise problem teaches a user-defined screening requirement to generate a classifier f, which represents interaction sequence information of an antibody drug.
Regarding the recited perform a sequence-based affinity prediction on candidate sequences within the mutation library using a deep learning model to obtain sequence information of at least one of a modified antibody and a modified macromolecular drug based on a ranking of the candidate sequences, Kang et al. discloses “We performed antibody-antigen complex affinity prediction based on sequence data with Hag-Net network structure.” (Page 3, Section “2.2.2 Baseline study: binding affinity change (
∆
∆
G
) prediction”, paragraph 1, line 1). Also, further discloses “In the proposed modeling approach, each node represents a single residue, and its edges represent connections with associated/interacting residues. This modeling approach represents each antibody-antigen complex as a single graph structure that corresponds to the amino acid sequences of antibody and antigen. The resulting graph was then converted to an input matrix by one-hot encoding, where each row represents a specific residue, as well as an adjacent matrix, where each 1 represents connection between amino acids in two positions. Thus, a 200 amino acid sequence will result in a 200x22 matrix as the node input and a 200x200 adjacent matrix as edge input.” (Page 3, Section “2.2.2 Baseline study: binding affinity change (
∆
∆
G
) prediction”, paragraph 1, lines 2-7). Kang et al. discloses “In the presented table, each wildtype (e.g. 1AK4) are listed with their associated mutated and binding affinity measurement. To construct an ordered pair, we take two mutations (mutation A and mutation B) and generate corresponding graph representations respectively. The outputs of Hag-Net are used as affinity scores for the comparison between mutations. The difference between affinity scores then goes through sigmoid function to predict the binary relation label, specifically in this example, label is set to 0 since mutate A has lower binding affinity than mutate B.” (Page 4, Figure 3, lines 1-6). Kang et al. further discloses “Antibody maturation aims to optimize the binding affinity based on known antibody leads targeting specific antigens. To this purpose, we construct the pairwise problem and study the binary relations between mutated variants of each therapeutic lead. Specifically, for each wildtype antibody, its associated mutations are grouped into unique ordered pairs. Each pair consists of two complexes with same target antigen but different mutated variants. If the first complex ranks higher than the second complex with respect to their binding affinities (i.e. the first complex possesses higher affinity), the pair is labeled as 1, otherwise as 0. Therefore, our goal is to obtain classifier f:
f
a
,
b
=
1
i
f
∆
∆
G
a
<
∆
∆
G
b
0
e
l
s
e
” (Page 3, Section “2.2.3 Pairwise-study: binding affinity pairwise rank prediction”, paragraph 1, lines 1-7). This teaches that the Hag-Net network structure is the deep learning model used to perform sequence-based affinity prediction on mutated antibody-antigen sequences. The affinity scores of the complexes generated from Hag-Net are ranked against each other to obtain their respective binary relation labels, which represents the sequence information of the modified antibody drug.
Regarding the recited an output module configured to output the sequence information of the at least one of the modified antibody and the modified macromolecular drug, Kang et al. discloses “Overall, 97711 pairs were generated from the original 1101 mutants of 32 complexes.” (Page 6, Section “3.2 Pairwise Study”, paragraph 1, line 1). Also, further discloses “Antibody maturation aims to optimize the binding affinity based on known antibody leads targeting specific antigens. To this purpose, we construct the pairwise problem and study the binary relations between mutated variants of each therapeutic lead. Specifically, for each wildtype antibody, its associated mutations are grouped into unique ordered pairs. Each pair consists of two complexes with same target antigen but different mutated variants. If the first complex ranks higher than the second complex with respect to their binding affinities (i.e. the first complex possesses higher affinity), the pair is labeled as 1, otherwise as 0. Therefore, our goal is to obtain classifier f:
f
a
,
b
=
1
i
f
∆
∆
G
a
<
∆
∆
G
b
0
e
l
s
e
” (Page 3, Section “2.2.3 Pairwise-study: binding affinity pairwise rank prediction”, paragraph 1, lines 1-7). This teaches outputting sequence information of modified antibodies, which can be represented by binary relation labels.
Kang et al. does not disclose a calculation module configured to: evaluate a mutation space of the at least one of the antibody and the macromolecular drug based on the template sequence information and modification requirements.
However, Warszawski et al. discloses AbLIFT, which is an automated design protocol for improving molecular interactions across the light-heavy chain (vL-vH) interface (pg. 7, para. 2, lines 1-5). To validate AbLIFT, two antibodies were chosen as subjects for design: the synthetic antibody G6 and an engineered variant of the 492.1 antibody (pg. 8, para. 1, lines 1-4). Then, AbLIFT computes a mutational-tolerance map of G6 starting from its bound structure, PDB entry 2FJG, where 30 vL-vH interface positions defined 26 affinity-enhancing mutations at 11 positions (pg. 8, para. 2, lines 1-5). Combinations of native-state energy and point mutation evolutionary-conservation score (PSSMs) thresholds were screened to determine which combination optimally discriminates enhancing from deleterious mutations and predicts which mutations in the vL-vH interface were likely to enhance affinity and stability, resulting in a tolerated sequence space of 203,835 unique multipoint mutants (pg. 5, para. 3, lines 1-5; pg. 6, para. 2, lines 1-3; pg. 7, para. 2, lines 1-5; pg. 8, para. 2, lines 1-5; pg. 18, para. 6). This teaches evaluating a mutation space of an antibody based on the antibody’s original bound structure as a template and modification threshold requirements.
Kang et al. does not disclose a calculation module configured to: narrow a mutation range for screening if the mutation space exceeds a predetermined upper limit.
However, Warszawski et al. discloses combinations of native-state energy and point mutation evolutionary-conservation score (PSSMs) thresholds were screened to determine which combination optimally discriminates enhancing from deleterious mutations and predicts which mutations in the vL-vH interface were likely to enhance affinity and stability, resulting in a tolerated sequence space of 203,835 unique multipoint mutants (pg. 5, para. 3, lines 1-5; pg. 6, para. 2, lines 1-3; pg. 7, para. 2, lines 1-5; pg. 8, para. 2, lines 1-5; pg. 18, para. 6). This includes PSSM (
≥
-
1
) and
∆
∆
G
(
≤
+
1
R
.
e
.
u
.
)
filters. This teaches narrowing a mutation space if it exceeds an upper limit.
Kang et al. does not disclose perform at least one of corresponding partial and exhaustive numeration of possible sequences in a part of a full variable range to obtain a mutation library.
However, Warszawski et al. discloses exhaustively enumerating on a subset of stable multipoint mutants within the vast hypothetical sequence space of mutants at the vL-vH interface (
20
30
=
10
39
unique sequences) (pg. 7, para. 2, lines 1-5; pg. 8, para. 2, lines 9-15). This teaches exhaustive enumeration of possible tolerated mutations at positions in the mutational-tolerance map to obtain a subset mutation library.
With respect to claim 2:
Kang et al. does not disclose wherein, in the calculation module, a single quantity level of the mutation library is not less than
10
10
.
However, Warszawski et al. discloses a hypothetical sequence space of mutants at the vL-vH interface (
20
30
=
10
39
unique sequences) (pg. 7, para. 2, lines 1-5; pg. 8, para. 2, lines 9-15). This teaches a quantity level of a mutation library that is greater than
10
10
.
With respect to claim 3:
Kang et al. does not disclose wherein, in the calculation module, the variable range includes one or more variable regions, variable spaces, variable number of sites, or combinations thereof.
However, Warszawski et al. discloses AbLIFT computing a mutational-tolerance map of G6 starting from its bound structure, PDB entry 2FJG, where 30 vL-vH interface positions defined 26 affinity-enhancing mutations at 11 positions (pg. 8, para. 2, lines 1-5). This teaches a variable range including a variable number of sites from the vL-vH interface regions.
With respect to claim 4:
Warszawski et al. does not disclose wherein, the template sequence information of at least one of the antibody and the macromolecular drug includes at least one element of a set comprising an antigen-antibody template sequence, a protein-protein template sequence, and a protein-polypeptide template sequence.
However, Kang et al. discloses “Antibody-Bind (AB-Bind) is a manually curated and organized database that includes 1101 mutants across 32 complexes” (Page 2, Section “2.1 Data Collection”, line 1). Also, further discloses “In the proposed modeling approach, each node represents a single residue, and its edges represent connections with associated/interacting residues. This modeling approach represents each antibody-antigen complex as a single graph structure that corresponds to the amino acid sequences of antibody and antigen.” (Page 3, Section “2.2.2 Baseline study: binding affinity change (
∆
∆
G
) prediction”, paragraph 1, lines 2-4). This teaches an input template sequence information of an antibody including elements of residues and connections with interacting residues of a set of antigen-antibody complexes from the Antibody-Bind database.
With respect to claim 5:
Warszawski et al. does not disclose wherein, the modification requirements of at least one target of the at least one of the antibody and the macromolecular drug, further comprising: at least one element of a set comprising marking the variable range and specifying the variable range.
However, Kang et al. discloses “2) Contacts-only model: the contacts-only model produces a compact representation of the complex by utilizing residues on the interfacial surface (distance <5 Angstrom) only” (Page 3, Section “2.2.2 Baseline study: binding affinity change (
∆
∆
G
) prediction”, paragraph 2, lines 5-6). This suggests marking and specifying the variable range of residues with distance <5 Angstrom.
Warszawski et al. does not disclose wherein, the modification requirements of at least one target of the at least one of the antibody and the macromolecular drug, further comprising: defining a modification direction.
However, Kang et al. discloses “1) Full-seq model: the full-seq model simply takes antibody and antigen sequences as two separated graph sequences (Figure 2(A)). The intuition is to incorporate both interact contacts (ICs) and non-interacting surface (NIS) into modeling as the binding strength between antibody and antigen relies on the full conformation of the formed complex” (Page 3, Section “2.2.2 Baseline study: binding affinity change (
∆
∆
G
) prediction”, paragraph 2, lines 2-5). This indicates defining a modification direction as towards both interact contacts and non-interacting surfaces for modeling based on binding strength.
With respect to claim 6:
Claim 6 recites wherein, the output module further comprises a visual analysis display module. It would be obvious to one of ordinary skill in the art to incorporate a visual output device or monitor in the affinity modification system because a display is a well-known component used to provide information to the user.
With respect to claim 7:
Warszawski et al. does not disclose wherein, the visual analysis display module provides the complete sequence information of at least one of the modified antibody and the modified macromolecular drug.
However, Kang et al. discloses “Overall, 97711 pairs were generated from the original 1101 mutants of 32 complexes.” (Page 6, Section “3.2 Pairwise Study”, paragraph 1, line 1). Also, further discloses “Antibody maturation aims to optimize the binding affinity based on known antibody leads targeting specific antigens. To this purpose, we construct the pairwise problem and study the binary relations between mutated variants of each therapeutic lead. Specifically, for each wildtype antibody, its associated mutations are grouped into unique ordered pairs. Each pair consists of two complexes with same target antigen but different mutated variants. If the first complex ranks higher than the second complex with respect to their binding affinities (i.e. the first complex possesses higher affinity), the pair is labeled as 1, otherwise as 0. Therefore, our goal is to obtain classifier f:
f
a
,
b
=
1
i
f
∆
∆
G
a
<
∆
∆
G
b
0
e
l
s
e
” (Page 3, Section “2.2.3 Pairwise-study: binding affinity pairwise rank prediction”, paragraph 1, lines 1-7). This teaches outputting sequence information of modified antibodies, which can be represented by binary relation labels.
With respect to claim 8:
Warszawski et al. does not disclose wherein, the visual analysis display module further comprises a comparative analysis of the template sequence information of the at least one of the antibody and the macromolecular drug and the sequence information of at least one of the modified antibody and the modified macromolecular drug in a variable range.
However, Kang et al. discloses “In the presented table, each wildtype (e.g. 1AK4) are listed with their associated mutated and binding affinity measurement. To construct an ordered pair, we take two mutations (mutation A and mutation B) and generate corresponding graph representations respectively. The outputs of Hag-Net are used as affinity scores for the comparison between mutations. The difference between affinity scores then goes through sigmoid function to predict the binary relation label, specifically in this example, label is set to 0 since mutate A has lower binding affinity than mutate B.” (Page 4, Figure 3, lines 1-6). This suggests a comparative analysis of the sequence information between mutation A and mutation B, which represent the template antibody and modified antibody drug in a variable range, respectively.
With respect to claim 9:
Warszawski et al. does not disclose an input template sequence information of at least one of the antibody and the macromolecular drug, modification requirements of at least one target of at least one of the antibody and the macromolecular drug, and optional user-defined screening requirements to generate interaction sequence information of the at least one of the antibody and the macromolecular drug.
However, Kang et al. discloses “Antibody-Bind (AB-Bind) is a manually curated and organized database that includes 1101 mutants across 32 complexes” (Page 2, Section “2.1 Data Collection”, line 1). Also, further discloses “we explore the graph representation of antibody-antigen complex with three different representation strategies. 1) Full-seq model: the full-seq model simply takes antibody and antigen sequences as two separated graph sequences (Figure 2(A)). The intuition is to incorporate both interact contacts (ICs) and non-interacting surface (NIS) into modeling as the binding strength between antibody and antigen relies on the full conformation of the formed complex [9, 12]. 2) Contacts-only model: the contacts-only model produces a compact representation of the complex by utilizing residues on the interfacial surface (distance <5 Angstrom) only (Figure 2(B)); Given limited high-quality training data, this approach is presumably more adequate since it models the most relevant information for antibody-antigen interactions. The identification of interface contacts was obtained using Prodigy-based prediction service [9] based on complex’s 3D structure. 3) Antibody-only model: the antibody-only model aims to address the promiscuous binding capability of antibodies cross diverse antigens [15, 16, 17, 18]. We investigated on antibody-only modeling for binding affinity prediction, and evaluate if Hag-Net based network structure captures the enabling features for antibodies’ natural binding capability.” (Page 3, Section “2.2.2 Baseline study: binding affinity change (
∆
∆
G
) prediction”, paragraph 2, lines 1-12). Kang et al. discloses “Antibody maturation aims to optimize the binding affinity based on known antibody leads targeting specific antigens. To this purpose, we construct the pairwise problem and study the binary relations between mutated variants of each therapeutic lead. Specifically, for each wildtype antibody, its associated mutations are grouped into unique ordered pairs. Each pair consists of two complexes with same target antigen but different mutated variants. If the first complex ranks higher than the second complex with respect to their binding affinities (i.e. the first complex possesses higher affinity), the pair is labeled as 1, otherwise as 0. Therefore, our goal is to obtain classifier f:
f
a
,
b
=
1
i
f
∆
∆
G
a
<
∆
∆
G
b
0
e
l
s
e
” (Page 3, Section “2.2.3 Pairwise-study: binding affinity pairwise rank prediction”, paragraph 1, lines 1-7). Antibody-Bind (AB-Bind) is a database of mutants across complexes, which teaches an input template sequence information of antibody drugs. The three graph representation strategies for antibody-antigen complexes indicate modification requirements for at least one target of antibody drugs because they model bindings and interactions between antibodies and antigens. The pairwise problem teaches a user-defined screening requirement to generate a classifier f, which represents interaction sequence information of an antibody drug.
Kang et al. does not disclose evaluate a mutation space of the at least one of the antibody and the macromolecular drug based on the template sequence information and modification requirements.
However, Warszawski et al. discloses AbLIFT, which is an automated design protocol for improving molecular interactions across the light-heavy chain (vL-vH) interface (pg. 7, para. 2, lines 1-5). To validate AbLIFT, two antibodies were chosen as subjects for design: the synthetic antibody G6 and an engineered variant of the 492.1 antibody (pg. 8, para. 1, lines 1-4). Then, AbLIFT computes a mutational-tolerance map of G6 starting from its bound structure, PDB entry 2FJG, where 30 vL-vH interface positions defined 26 affinity-enhancing mutations at 11 positions (pg. 8, para. 2, lines 1-5). Combinations of native-state energy and point mutation evolutionary-conservation score (PSSMs) thresholds were screened to determine which combination optimally discriminates enhancing from deleterious mutations and predicts which mutations in the vL-vH interface were likely to enhance affinity and stability, resulting in a tolerated sequence space of 203,835 unique multipoint mutants (pg. 5, para. 3, lines 1-5; pg. 6, para. 2, lines 1-3; pg. 7, para. 2, lines 1-5; pg. 8, para. 2, lines 1-5; pg. 18, para. 6). This teaches evaluating a mutation space of an antibody based on the antibody’s original bound structure as a template and modification threshold requirements.
Kang et al. does not disclose narrow a mutation range for screening if the mutation space exceeds a predetermined upper limit.
However, Warszawski et al. discloses combinations of native-state energy and point mutation evolutionary-conservation score (PSSMs) thresholds were screened to determine which combination optimally discriminates enhancing from deleterious mutations and predicts which mutations in the vL-vH interface were likely to enhance affinity and stability, resulting in a tolerated sequence space of 203,835 unique multipoint mutants (pg. 5, para. 3, lines 1-5; pg. 6, para. 2, lines 1-3; pg. 7, para. 2, lines 1-5; pg. 8, para. 2, lines 1-5; pg. 18, para. 6). This includes PSSM (
≥
-
1
) and
∆
∆
G
(
≤
+
1
R
.
e
.
u
.
)
filters. This teaches narrowing a mutation space if it exceeds an upper limit.
Kang et al. does not disclose perform at least one of: corresponding partial and exhaustive numeration of possible sequences in a part of a full variable range to obtain a mutation library.
However, Warszawski et al. discloses exhaustively enumerating on a subset of stable multipoint mutants within the vast hypothetical sequence space of mutants at the vL-vH interface (
20
30
=
10
39
unique sequences) (pg. 7, para. 2, lines 1-5; pg. 8, para. 2, lines 9-15). This teaches exhaustive enumeration of possible tolerated mutations at positions in the mutational-tolerance map to obtain a subset mutation library.
Warszawski et al. does not disclose perform a sequence-based affinity prediction on candidate sequences within the mutation library using a deep learning model obtain sequence information of at least one of a modified antibody and a modified macromolecular drug based on a ranking of the candidate sequences.
However, Kang et al. discloses “We performed antibody-antigen complex affinity prediction based on sequence data with Hag-Net network structure.” (Page 3, Section “2.2.2 Baseline study: binding affinity change (
∆
∆
G
) prediction”, paragraph 1, line 1). Also, further discloses “In the proposed modeling approach, each node represents a single residue, and its edges represent connections with associated/interacting residues. This modeling approach represents each antibody-antigen complex as a single graph structure that corresponds to the amino acid sequences of antibody and antigen. The resulting graph was then converted to an input matrix by one-hot encoding, where each row represents a specific residue, as well as an adjacent matrix, where each 1 represents connection between amino acids in two positions. Thus, a 200 amino acid sequence will result in a 200x22 matrix as the node input and a 200x200 adjacent matrix as edge input.” (Page 3, Section “2.2.2 Baseline study: binding affinity change (
∆
∆
G
) prediction”, paragraph 1, lines 2-7). Kang et al. discloses “In the presented table, each wildtype (e.g. 1AK4) are listed with their associated mutated and binding affinity measurement. To construct an ordered pair, we take two mutations (mutation A and mutation B) and generate corresponding graph representations respectively. The outputs of Hag-Net are used as affinity scores for the comparison between mutations. The difference between affinity scores then goes through sigmoid function to predict the binary relation label, specifically in this example, label is set to 0 since mutate A has lower binding affinity than mutate B.” (Page 4, Figure 3, lines 1-6). Kang et al. further discloses “Antibody maturation aims to optimize the binding affinity based on known antibody leads targeting specific antigens. To this purpose, we construct the pairwise problem and study the binary relations between mutated variants of each therapeutic lead. Specifically, for each wildtype antibody, its associated mutations are grouped into unique ordered pairs. Each pair consists of two complexes with same target antigen but different mutated variants. If the first complex ranks higher than the second complex with respect to their binding affinities (i.e. the first complex possesses higher affinity), the pair is labeled as 1, otherwise as 0. Therefore, our goal is to obtain classifier f:
f
a
,
b
=
1
i
f
∆
∆
G
a
<
∆
∆
G
b
0
e
l
s
e
” (Page 3, Section “2.2.3 Pairwise-study: binding affinity pairwise rank prediction”, paragraph 1, lines 1-7). This teaches that the Hag-Net network structure is the deep learning model used to perform sequence-based affinity prediction on mutated antibody-antigen sequences. The affinity scores of the complexes generated from Hag-Net are ranked against each other to obtain their respective binary relation labels, which represents the sequence information of the modified antibody drug.
Warszawski et al. does not disclose output the sequence information of the at least one of the modified antibody and the modified macromolecular drug.
However, Kang et al. discloses “Overall, 97711 pairs were generated from the original 1101 mutants of 32 complexes.” (Page 6, Section “3.2 Pairwise Study”, paragraph 1, line 1). Also, further discloses “Antibody maturation aims to optimize the binding affinity based on known antibody leads targeting specific antigens. To this purpose, we construct the pairwise problem and study the binary relations between mutated variants of each therapeutic lead. Specifically, for each wildtype antibody, its associated mutations are grouped into unique ordered pairs. Each pair consists of two complexes with same target antigen but different mutated variants. If the first complex ranks higher than the second complex with respect to their binding affinities (i.e. the first complex possesses higher affinity), the pair is labeled as 1, otherwise as 0. Therefore, our goal is to obtain classifier f:
f
a
,
b
=
1
i
f
∆
∆
G
a
<
∆
∆
G
b
0
e
l
s
e
” (Page 3, Section “2.2.3 Pairwise-study: binding affinity pairwise rank prediction”, paragraph 1, lines 1-7). This teaches outputting sequence information of modified antibodies, which can be represented by binary relation labels.
With respect to claim 10:
Kang et al. does not disclose wherein, when performing the at least one of partial and exhaustive numeration of possible sequence in a part of the full variable range, a single quantity level of the mutation library is not less than
10
10
.
However, Warszawski et al. discloses exhaustively enumerating on a subset of stable multipoint mutants within the vast hypothetical sequence space of mutants at the vL-vH interface (
20
30
=
10
39
unique sequences) (pg. 7, para. 2, lines 1-5; pg. 8, para. 2, lines 9-15). This teaches exhaustive enumeration of possible tolerated mutations at positions in the mutational-tolerance map, where a mutation library is greater than
10
10
.
It would have been prima facie obvious to one of ordinary skill in the art to modify the deep learning model for predicting antibody-antigen binding affinity disclosed by Kang et al. to incorporate mutation library enumeration disclosed by Warszawski et al. One would be motivated to incorporate sequence enumeration into the deep learning model for affinity prediction because the AbLIFT server disclosed by Warszawski et al. implements more accurate atomistic scoring and enables greater user control (pg. 14, para. 3, lines 4-9). This means that incorporating sequence enumeration into the deep learning model for affinity prediction will allow for greater accuracy and user control. There is a likelihood of success, since affinity prediction using deep learning models and optimization of antibody affinity through automated design are well known techniques in the field of computational biology.
Response to Arguments
Applicant’s arguments, see pg. 16-18, filed 4/23/2026, with respect to the rejection(s) of claim(s) 1-10 under 35 U.S.C. § 103 have been fully considered and are persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made in view of Kang et al. (arXiv preprint, 2021, 1-9).
Conclusion
No claims are allowed.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
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
/Karlheinz R. Skowronek/Supervisory Patent Examiner, Art Unit 1687
Prosecution Insights