DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Status
Claims 1-20 are pending.
Priority
This application which claims benefit of application no. 63/038,691, filed 06/12/2020 and application no. 63/403,894, filed 09/06/2022. The instant application has the effective filing date of 06 September 2022.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 08/15/2023 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement has been considered by the examiner.
Drawings
The drawings, submitted on 08/15/2023, are accepted by the examiner.
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 1-20 are rejected under U.S.C 101 because the claimed invention is directed to abstract ideas without significantly more, as detailed in the analysis below.
Eligibility Step 1: Subject matter eligibility evaluation in accordance with MPEP § 2106:
Claims 1-7 are directed to a statutory category (method).
Claim 15-20 are directed to a statutory category (system).
Claims 8-14 are non-statutory as they recite “a computer program product”. The claims as instantly recited read on carrier waves, which are transitory propagating signals and therefore are not proper patentable subject matter because they do not fit within any of the four statutory categories of invention (In re Nuijten, Federal Circuit, 2007). It is noted that the recitation of a "non-transitory computer program product" would overcome the rejection with respect to claims 8-14 reading on signals. However, the amendment to only "non-transitory computer program product" would not overcome the rejection under 35 U.S.C. 101 since the claims would still be directed to a judicial exception without significantly more (see below).
Claims 1-7 and 15-20 [Eligibility Step 1: YES]
Claims 8-14 [Eligibility Step 1: NO]
Though claims 8-14 are not directed to statutory subject matter, in the interest of compact prosecution, Alice/Mayo Evaluation via MPEP 2143 continues below on all claims.
Eligibility Step 2A: This step determines whether a claim is directed to a judicial exception in accordance with MPEP § 2106.
Eligibility Step 2A -- Prong One: Limitations are analyzed to determine if the claims recite any concepts that could equate to a judicial exception (i.e. abstract idea, law of nature, or natural phenomenon). Possible judicial exceptions are explored below.
Recitations of Judicial Exceptions:
Claims 1, 8, and 15: optionally introducing a minimal number of mutations to a T-cell receptor (TCR) sequence to enable the TCR sequence to bind to a peptide; (mental process)
using a disentangled Wasserstein autoencoder to separate an embedding space of the TCR sequence into functional embeddings and structural embeddings; (mathematical concept)
using an auxiliary classifier to predict a probability of a positive binding label from the functional embeddings and the peptide; (mathematical concept)
generating new TCR sequences with enhanced binding affinity for immunotherapy to target a particular virus or tumor. (mental process)
Claims 2, 9, and 16: wherein the functional embeddings include information about a generic sequential context and the structural embeddings encode patterns that are responsible for peptide recognition. (mathematical concept)
Claims 3, 10, and 17: wherein a first and second auxiliary loss are employed to ensure the functional embeddings encode functional information while being independent of the structural embeddings. (mathematical concept)
Claims 4, 11, and 18: wherein the first auxiliary loss is a Wasserstein loss based on a maximum mean discrepancy (MMD) between a marginal distribution of concatenated embeddings. (mathematical concept)
Claims 5, 12, and 19: wherein the second auxiliary loss is an isotropic multivariate normal distribution loss. (mathematical concept)
Claims 6 and 13: wherein the functional embeddings correspond to functional patterns and the structural embeddings correspond to structural patterns. (mathematical concept)
Claims 7 and 14: wherein the functional embeddings are encoded by a functional encoder and the structural embeddings are encoded by a structural encoder. (mathematical concept)
Claim 20: wherein the functional embeddings correspond to functional patterns and the structural embeddings correspond to structural patterns. (mathematical concept)
wherein the functional embeddings are encoded by a functional encoder and the structural embeddings are encoded by a structural encoder. (mathematical concept)
Step 2A – Prong One Analysis:
Analysis techniques such as generically “generating” a sequence or introducing mutations require nothing more than the human mind and pen/paper, and thus fall under the mental process grouping of abstract ideas.
Analysis techniques such as scoring encoding, loss functions, and distributions recite mathematical calculations, formulas and/or relationships that fall under the mathematical concept grouping of abstract ideas. Limitations that merely provide additional information regarding the data being analyzed in this manner are similarly categorized (claims 2, 6, 9, 13, and 16).
Therefore, the claims are found to recite judicial exceptions.
[Eligibility Step 2A – Prong One: YES]
Eligibility Step 2A – Prong Two: 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. If the claim contains no additional claim elements beyond the abstract idea, the claim fails to integrate the abstract idea into a practical application (MPEP 2106.04(d)). Additional elements are recited, categorized, and analyzed below.
Data Gathering Elements:
Claims 1, 8, and 15: feeding the functional embeddings and the structural embeddings to a long short-term memory (LSTM) or transformer decoder
Computer Components Elements:
Claims 1: computer-implemented method
Claim 8: computer program product
Claim 15: computer processing system for learning disentangled representations for T- cell receptors to improve immunotherapy, comprising: a memory device for storing program code; and a processor device operatively coupled to the memory device, for running the program code to
Step 2A – Prong Two Analysis:
Performing data input necessary to carry out the judicial exceptions is a data gathering activity which equates to insignificant extra-solution activity per MPEP 2106.05 (g).
Generic computer components and implementations provide mere instructions to implement the abstract ideas onto a technological environment per Alice Corp., 573 U.S. at 223, 110 USPQ2d at 1983. See also 573 U.S. at 224, 110 USPQ2d at 1984.
As such, the additional elements, when viewed separately and in the context of a whole claimed invention, do not integrate the judicial exceptions into practical application.
[Eligibility Step 2A – Prong Two: NO]
Eligibility Step 2B: Claim elements are probed for inventive concept equating to significantly more than the judicial exception (MPEP 2106.04(II)).
Step 2B Analysis:
Data gathering activities are further found to be well-understood, routine, and conventional per Mayo, 566 U.S. at 79, 101 USPQ2d at 1968; OIP Techs., Inc. v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1092-93 (Fed. Cir. 2015) of MPEP 2106.05(g); and Akbar et al. (mAbs; Vol. 14: 1; 2022), which reviews machine learning-based design of fit-for-purpose monoclonal antibodies.
The computer components are further found to be well-understood, routine, and conventional per Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93 for storing and retrieving information in memory and Alice Corp., 573 U.S. at 225, 110 USPQ2d at 1984 (MPEP 2106.05 (a)).
As such, the additional elements are further found to lack inventive concept.
[Eligibility Step 2B: NO]
Therefore, claims 1-20 are directed to judicial exceptions without significantly more and are rejected under 35 U.S.C 101.
Claim Rejections - 35 USC § 112
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 1-20 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.
The term “minimal number” in claims 1, 8, and 15 is a relative term which renders the claim indefinite. The term “minimal number” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. The dependent claims are rejected on similar grounds for failing to remedy the issue herein.
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.
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.
Claims 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Lu et al. (2023/0349914) in view of Li et al. (Briefings in Bioinformatics; Vol. 22: 6; 2021), Chen et al. (Mol Syst Des Engineering; vol. 6: 406; 2021), and Han et al. (ICLR; arxiv: 2101.07496; 2021).
Lu et al. describes a deep learning system for predicting the T-cell receptor binding specificity of neoantigens.
Claims 1, 8, and 15 are directed to methods, computer-readable mediums and systems for learning disentangled representations for T-cell receptors to improve immunotherapy, in which the methods include: optionally introducing a minimal number of mutations to a T-cell receptor (TCR) sequence to enable the TCR sequence to bind to a peptide; and using a disentangled Wasserstein autoencoder to separate an embedding space of the TCR sequence into functional embeddings and structural embeddings.
Lu et al. teaches performing residue mutations for all the 619 TCRs included in the testing cohort of the validation data [0095]; and determining TCR embeddings using an auto-encoder that includes multiple encoder layers and multiple decoder layers [0012].
Lu et al. further teaches the embedding network first determines numeric embeddings of pMHCs that represent the protein sequences of neoantigens and the MHCs numerically [0060]; second, the stacked auto-encoder determines an embedding of TCR sequences that encode text strings of TCR sequences numerically [0060]; and the two step approach to numerically encoding pMHCs and TCR sequences provides several advantages that improve the computational efficiency of the training process and the flexibility of the trained models [0060].
Therefore Lu et al. teaches using an embedding network and auto-encoder to create a TCR embedding space separated into functional and structural embeddings.
Claims 1, 8, and 15 are further directed to feeding the functional embeddings and the structural embeddings to a long short-term memory (LSTM) or transformer decoder; using an auxiliary classifier to predict a probability of a positive binding label from the functional embeddings and the peptide; and generating new TCR sequences with enhanced binding affinity for immunotherapy to target a particular virus or tumor.
Lu et al. teaches the encoded pMHCs and neoantigens may be input into the LSTM layers [0068]; and a multi-layer neural network determines a probability that a particular pMHC molecule binds to one or more neo-antigen protein sequences [0010].
Claims 2, 9, and 16 are directed to wherein the functional embeddings include information about a generic sequential context and the structural embeddings encode patterns that are responsible for peptide recognition.
Lu et al. teaches determining a set of TCR embeddings that encode TCR data for a plurality of TCR sequences [0007]; and determining a set of MHC embeddings that encode neoantigen and major histocompatibility complex (MHC) data for a plurality of MHC proteins (pMHC) [0007], in which neoantigens serve as recognition markers for cytotoxic T cells via their interactions with T cell receptors (TCRs) [0003].
Claims 6 and 13 are directed to wherein the functional embeddings correspond to functional patterns and the structural embeddings correspond to structural patterns.
Lu et al. teaches the TCR training dataset may include TCR data, for example, a matrix or other structured data representation of one or more biochemical properties of amino acids included in each of the training TCR protein sequences [0053]; and although the output of the embedding network can be dedicated to predicting antigen and MHC binding, the internal layers of the network may contain important information regarding the overall structure of the pMHC complex [0066].
Claims 7 and 14 are directed to wherein the functional embeddings are encoded by a functional encoder and the structural embeddings are encoded by a structural encoder.
Lu et al. teaches at step 102, a set of MHC embeddings is determined; the MHC embeddings may encode neoantigen and MHC data for a plurality of pMHCs; and each of the MHC embeddings may include a numeric representation of one or more pMHCs generated by a multi-layer neural network or other MHC numeric embedding layer [0052].
Lu et al. teaches at step 104, a set of TCR embeddings is determined; the TCR embeddings may include a numeric representation of TCR sequences generated by an auto-encoder or other TCR numeric embedding layer [0053].
Claim 20 is directed to wherein the functional embeddings correspond to functional patterns and the structural embeddings correspond to structural patterns; and wherein the functional embeddings are encoded by a functional encoder and the structural embeddings are encoded by a structural encoder.
Lu et al. teaches the TCR training dataset may include TCR data, for example, a matrix or other structured data representation of one or more biochemical properties of amino acids included in each of the training TCR protein sequences [0053]; and although the output of the embedding network can be dedicated to predicting antigen and MHC binding, the internal layers of the network may contain important information regarding the overall structure of the pMHC complex [0066].
Lu et al. teaches at step 102, a set of MHC embeddings is determined; the MHC embeddings may encode neoantigen and MHC data for a plurality of pMHCs; and each of the MHC embeddings may include a numeric representation of one or more pMHCs generated by a multi-layer neural network or other MHC numeric embedding layer [0052].
Lu et al. teaches at step 104, a set of TCR embeddings is determined; the TCR embeddings may include a numeric representation of TCR sequences generated by an auto-encoder or other TCR numeric embedding layer [0053].
Lu et al. further teaches additional insights into the interactions between pMHCs and TCR sequences could be used to enhance the design or implementation of various types of immunotherapies [0004]; and the prediction model could predict the TCRs that would be most effective at targeting specific tumors allowing for preparation of a vaccine including neoantigens that can activate the targeted T cells with these TCRs [0058].
Lu et al. does not explicitly teach generating new TCR sequences with enhanced binding affinity for immunotherapy to target a particular virus or tumor (claims 1, 8, and 15).
Li et al. describes DeepImmuno, a deep learning-empowered prediction and generation of immunogenic peptides for T-cell immunity.
Li et al. teaches a T cell-specific immune response will be triggered if a peptide is capable of binding with a cognate MHC molecule, and the resultant peptide–MHC complex can further interact with selected TCR sequences (page 2, column 1); peptides that meet these criteria are referred to as immunogenic peptides (page 2, column 1); DeepImmuno-GAN is able to learn and produce synthetic immunogenic pseudo-sequences (page 7, fig. 4); and increase the sensitivity for detection of valid neoantigens, such as tumor-specific mutations (page 8, column 2).
Therefore Li et al. teaches generating immunogenic peptides, capable of binding with MHCs and interacting with TCR sequences. Lu et al. provides sufficient motivation for one of ordinary skill in the art to use the prediction method to generate an enhanced design of an immunotherapy that would be most effective at targeting specific tumors. As such, it would be obvious to one of ordinary skill in the art to combine the prediction method of Lu et al. with the production method of Li et al. in order to predict and generate immunotherapies in the form of immunogenic peptides with enhanced binding and interacting abilities.
Li et al. further teaches training a generative adversarial network (GAN) (page 6, column 2) to accurately simulate immunogenic peptides with physicochemical properties and immunogenicity predictions similar to that of real antigens (page 1, column 1). Li et al. further teaches the generative adversarial network may be a Wasserstein GAN (page 6, column 2).
Lu et al. in view of Li et al. do not teach using a disentangled Wasserstein autoencoder to separate the embedding space (claims 1, 8, and 15).
Chen et al. reviews sequence-based peptide identification, generation, and property prediction with deep learning.
Chen et al. teaches focusing our attention peptide identification, property prediction, and new sequence generation (page 3, column 1); vaccine development efforts directly benefit from the accurate prediction of the binding affinity between MHC alleles and specific peptides, since neoantigens are ideal targets for immunotherapy (page 14, column 1); and the development of versatile deep learning pan-specific models for different MHC alleles or allele-specific models with extremely high accuracy is a future direction of MHC–peptide binding studies (page 14, column 1).
Chen et al. further teaches although RNNs have shown potential for peptide generation, they can be further integrated to form more complex generative models; representatives of these models are variational autoencoders (VAE) and generative adversarial networks (GAN) (page 12, column 1).
Chen et al. teaches on the basis of VAE, another variant called conditional VAE (CVAE) is also gaining attention in various areas like image classification, intrusion detection, molecule generation, and peptide design (page 12, column 2); a CVAE model had the potential to learn a disentangled latent representation of data, which means the manipulation of latent code will only influence the desired property without resulting in unexpected variation of other properties (page 13, column 1); and disentangled generation of peptides is attractive since in many cases only select properties are wanted while others can be kept unchanged (page 13, column 1).
Therefore Chen et al. teaches both generative adversarial networks and disentangled variational autoencoder that learn disentangled latent representations of data are techniques known to be effective for protein sequence generation in the field of protein-MHC binding interaction prediction and immunotherapy production. As such, it would be obvious to one of ordinary skill in the art to complete the predicting and generation of an enhanced peptide-MHC binding sequence using either of these known techniques with a reasonable expectation of success.
Chen et al. does not teach using a disentangled Wasserstein autoencoder to separate the embedding space.
Han et al. describes disentangled recurrent Wasserstein autoencoders.
Han et al. teaches disentangled representation learning, which further separates the latent embedding space into exclusive explainable factors such that each factor only interprets one of semantic attributes of sensory data, has received a lot of interest and achieved many empirical successes on static data such as images (page 1, column 1); recurrent Wasserstein Autoencoder (R-WAE) learns disentangled representations of sequential data (page 2, column 1); and our proposed R-WAE(MMD) achieves better disentanglement than disentangled variational autoencoder, DS-VAE (page 7, column 1).
Claims 3, 10, and 17 are directed to wherein a first and second auxiliary loss are employed to ensure the functional embeddings encode functional information while being independent of the structural embeddings.
Han et al. teaches updating p0 and py with loss given by equations 9 and 11 (page 15, column 2); defining the following probabilistic generative model by assuming Zmt and Zc are independent (page 3, column 1); and using the model to further disentangle the attribute factors within content latent variables (page 7, column 1).
Claim 4, 11, and 18 are directed to wherein the first auxiliary loss is a Wasserstein loss based on a maximum mean discrepancy (MMD) between a marginal distribution of concatenated embeddings.
Han et al. teaches R-WAE minimizes a penalized form of a Wasserstein distance (page 4, column 1); the details of optimizing the first term Maximum Mean discrepancy MMDkγ (QZc,PZc) in Eq. (11) is provided in Appendix D based on scaled MMD, a principled and stable technique for training MMD-based critic; and we call the resulting model R-WAE(MMD) (see Algorithm 2 in Appendix for details) (page 5, column 1).
Claim 5, 12, and 19 are directed to wherein the second auxiliary loss is an isotropic multivariate normal distribution loss.
Han et al. teaches the prior distribution of Zc is chosen as a multivariate unit-variance Gaussian, N(0,I) (page 5, column 1).
Therefore Chen et al. teaches using a disentangled variational autoencoder for peptide prediction and sequence generation useful for immunotherapy vaccine development. Han et al. provides sufficient motivation for one of ordinary skill in the art to substitute the variational autoencoder with a Wasserstein autoencoder as it achieves better disentanglement. As such, it would be obvious to one of ordinary skill in the art to apply the Wasserstein autoencoder methods to the method of Lu et al. in view of Li et al. with a reasonable expectation of success and improvement to the system.
Conclusion
No claims are currently allowed.
Correspondence
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Milana Thompson whose telephone number is (571)272-8740. The examiner can normally be reached Monday - Friday, 9:00-6:00 ET.
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) 272-1113. 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.
/M.K.T./Examiner, Art Unit 1687
/Karlheinz R. Skowronek/Supervisory Patent Examiner, Art Unit 1687
Prosecution Insights