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 .
Status of Claims
Claims 9-11, 23-25, 33 and 34 are cancelled.
Claims 1-8, 12-22, 26-32 and 35-38 are pending and examined on the merits.
Information Disclosure Statement
The information disclosure statements filed 10/06/2021, 10/18/2021, 04/05/2022 and 08/08/2022 are acknowledged. A signed copy of the corresponding 1449 forms have been included with this Office action.
Priority
Priority As detailed on the 09/03/2021filing receipt, this application claims priority to as early as 03/06/2019. At this point in examination, all claims have been interpreted as being accorded this priority date.
Interpretations
The dependency score is defined in the specifications. Further, gh(·) is any function, is herein throughout referred to as a dependency function for convenience of description, and generates dependency scores for the allele-interacting variables Xn k based on a set of parameters Oh. gh(xt eh). (00325)
I interpreted the "dependency score" a score based on which the likelihood that a particular peptide sequence, especially a neoantigen, will bind to and be presented by a specific MHC-II allele. For example, This score can quantifies the strength of the interaction between the peptide and the MHC molecule, reflecting the peptide's ability to elicit an immune response.
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 1-8, 12-22, 26-32 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
The claims are found herein to recite abstract ideas to fall into the grouping of mathematical concepts.
Framework Analysis 0f Claims 1-8, 12-22, 26-32 and 35-38:
Step 1: Are the claims directed to a category of statutory subject matter (a process, machine, manufacture, or composition of matter? (See MPEP § 2106.03)
Claims 1 -8, 12-22, 26-32 and 35-38 are properly directed to a 101 statutory category, specifically: a Method.
[Step 1: claims 1 -8, 12-22, 26-32 and 35-38: YES]
Step 2A, Prong One: Do the claims recite a judicially recognized exception, i.e., an abstract idea, a law of nature, or a natural phenomenon? (See MPEP § 2106.04(a))
Claims 1-10 recite abstract ideas. The MPEP at 2106.04(a)(2) further explains that abstract ideas are defined as:
• mathematical concepts, (mathematical formulas or equations, mathematical
relationships and mathematical calculations);
• certain methods of organizing human activity (fundamental economic practices
or principles, managing personal behavior or relationships or interactions between people); and/or
• mental processes (procedures for observing, evaluating, analyzing/ judging and
organizing information).
The MPEP at 2106.04(b) defines natural law/ natural phenomena as:
• naturally occurring principles/ relations that are naturally occurring or that do not have
markedly different characteristics compared to what occurs in nature.
With respect to the instant claims, under Step 2A, Prong One evaluation, the claims are found to recite abstract ideas that fall into the grouping of mathematical concepts and mental processes.
Regarding claim 1
Regarding the recited in claim 1, identifying one or more T-cells that are antigen-specific. Identifying T cells is a mental process and is an abstract idea.
Regarding the recited in claim 1, encoding the peptide sequences of each of the neoantigens into a corresponding numerical vector. Encoding sequences into numerical vectors is a mental process and is an abstract idea.
Regarding the recited in claim 1, generate a set of presentation likelihoods using machine learning. The use of machine learning is a mathematical concept of a calculation and is an abstract idea.
Regarding the recited in claim 1, training the dataset using machine learning and training peptide sequences encoded as numerical vectors, is a mathematical concept of a calculation and is an abstract idea.
Regarding the recited in claim 1, selecting a subset of the set of neoantigens based on the set of presentation likelihoods to generate a set of selected neoantigens. Selecting a subset and generating a set of neoantigens, are mental processes and are abstract ideas.
Regarding claim 2
Regarding the recited in claim 2, applying the machine-learned presentation model to the peptide sequence of the neoantigen to generate a dependency score. Applying machine learning to generate a dependency score is a mathematical concept of a calculation and is an abstract idea.
Regarding claim3
Regarding the recited in claim 3, transforming the dependency scores to generate a corresponding per-allele likelihood. The specification discloses the use of a function to transform the dependency score (00379, Lines4-5). Using a function to transform a score is a mathematical concept and is an abstract idea.
Regarding the recited in claim 3, combining the per-allele likelihoods to generate the presentation likelihood of the neoantigen. Combining per-allele likelihoods to generate the presentation likelihood of the neoantigen is a mental process and is an abstract idea.
Regarding claim 6
Regarding the recited in claim 6, applying the machine-learned presentation model to the allele noninteracting features to generate a dependency score. Applying machine learning is a mathematical concept of a calculation and is an abstract idea.
Regarding claim 7
Regarding the recited in claim 7, transforming the combined dependency scores. The specification discloses “The dependency score may be transformed by the transformation function_f{·) to
generate a per-allele likelihood” (P.00328). Using a function is a mathematical calculation and is an abstract idea.
Regarding the recited in claim 7, combining the per-allele likelihoods to generate the presentation likelihood. Combining per-allele likelihoods using a function as disclosed in (P.00375), is a mathematical concept and is an abstract idea.
Regarding claim 8
Regarding the recited in claim 8, combining the dependency scores for each of the class II MHC alleles and the dependency score for the allele noninteracting features. The specification discloses the use of a function to combine the dependency (P.00385). The use of a function is a mathematical concept and is an abstract idea.
Regarding the recited in claim 8, transforming the combined dependency scores to generate the presentation likelihood. Using a function disclosed in (P.00362) to transform scores is a mathematical concept and is an abstract idea.
Regarding claim 12
Regarding the recited in claim 12, the use of hot encoding scheme. The hot encoding scheme can be
done using a mental process and is an abstract idea.
Regarding claim 15
Regarding the recited in claim 15, presentation likelihoods are further identified by at least expression levels. Identifying presentation likelihoods by expression levels is a mental process and is an abstract idea.
Regarding claim 16
Regarding the recited in claim 16, Using predicted affinity and stability to identify presentation, is a mental process and is an abstract idea.
Regarding claim 18
Regarding the recited in claim 18, selecting neoantigens that have an increased likelihood of being presented on the cell surface relative to unselected neoantigens based on the machine-learned presentation model. Selecting neoantigens based on data presentenced is a mental process and is an abstract idea. Using a machine learning to select the neoantigens is a mathematical concept and is an abstract idea.
Regarding claim 19
Regarding the recited in claim 19, selecting the set of selected neoantigens comprises selecting neoantigens that have an increased likelihood of being capable of inducing a tumor-specific immune response in the subject relative to unselected neoantigens based on the machine-learned presentation model. Selecting neoantigens based on data presentenced is a mental process and is an abstract idea. Using a machine learning to select the neoantigens is a mathematical concept and is an abstract idea.
Regarding claim 20
Regarding the recited in claim 20, selecting neoantigens that have an increased likelihood of being capable of being presented to native T-cells by professional antigen presenting cells (APCs) relative to unselected neoantigens based on the presentation model. Selecting neoantigens based on data presentenced is a mental process and is an abstract idea.
Regarding claim 21
Regarding the recited in claim 21, selecting neoantigens that have a decreased likelihood of being subject to inhibition via central or peripheral tolerance relative to unselected neoantigens based on the machine-learned presentation model. Selecting neoantigens based on data presentenced is a mental process and is an abstract idea. Using a machine learning to select the neoantigens is a mathematical concept and is an abstract idea.
Regarding claim 22
Regarding the recited in claim 22, selecting neoantigens that have a decreased likelihood of being capable of inducing an autoimmune response to normal tissue in the subject relative to unselected neoantigens based on the machine-learned presentation model. Selecting neoantigens based on data presentenced is a mental process and is an abstract idea. Using a machine learning to select the neoantigens is a mathematical concept and is an abstract idea.
Regarding the recited in claim 29, updating neural network parameters, is a mental process and is an abstract idea.
[Step 2A, Prong One, abstract idea: claims 1-3, 6-8, 12, 15, 16, 18, 19 and 20-22: YES]
Step 2A, Prong Two: If the claims recite a judicial exception under Prong One, then is the judicial exception integrated into a practical application? (See MPEP § 2106.04(d))
A claim can be said to integrate a judicial exception into a practical application when it applies, relies on, or uses the judicial exception in a manner that imposes a meaningful limit on the judicial exception.
Regarding claims 1-3, 5, 37 and 38: Additional elements have been added to the claims. Yet the claims fail to integrate the judicial exception into a practical application [see MPEP § 2106.04(d)(III)].
Claims 1-3 and 5 additional elements are input and obtaining data.
Additional elements, such as receiving data and descriptions of data, are not considered
abstract ideas, but perform functions of inputting, collecting, and outputting the data needed to
carry out the abstract idea. These steps are considered insignificant extra-solution activity, and
are not sufficient to integrate an abstract idea into a practical application as they do not impose
any meaningful limitation on the abstract idea or how it is performed. To integrate a judicial
exception into a practical application, the additional limitations must not be mere instructions to
apply the judicial exception [see MPEP § 2106.04(d) and MPEP § 2106.05(g)].
Claims 36 and 37 additional element cloning
The additional element cloning is a practical application.
Claim 38 additional element infusing the expanded T-cells
The additional element infusing the expanded T-cells is a practical application.
[Step 2A, 2nd prong: claims 1-3, 5: No, Claims 37 and 38: Yes]
Step 2B: If the claims do not integrate the judicial exception, do the claims provide an inventive concept? (See MPEP § 2106.05)
Step 2B, which evaluates whether the additional elements, individually and in combination, amount to significantly more than the judicial exception itself by providing an inventive concept. An inventive concept is furnished by an element or combination of elements that is recited in the claim in addition to the judicial exception, and is sufficient to ensure that the claim, as a whole, amounts to significantly more than the judicial exception itself (see MPEP § 2106.05).
Claims 1-3 and 5 additional elements are input and obtaining data.
Additional elements, such as receiving data and descriptions of data, do not cause the claims to rise to the level of significantly more than the judicial exception. The courts have recognized receiving or transmitting data over a network; storing and retrieving information in memory, and determining the level of a biomarker in blood by any means [see MPEP§2106.05(d)(II)], as well-understood, routine, conventional activity when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity. Jurtz (J Immunol. 2017 Oct 4) teaches, the use of neoantigen data from melanoma samples. Thus, in light of the prior art, and MPEP § 2106.05(d)(II), the data processing steps are shown to be routine, well-understood, and conventional in the art. As a result, the additional element of data gathering steps does not provide an inventive concept by amounting to significantly more than the judicial exception.
Considering these elements alone and in combination, they do not provide an inventive concept, and do not amount to significantly more than the judicial exception itself. Thus, the additional elements do not provide an inventive concept that transforms the judicial exception into a patent eligible application of the exceptions, and the claims do not amount to significantly more that the judicial exception itself.
[Step2B: claims 1-3, 5: NO.]
Therefore, the claims, when the limitations are considered individually and as a whole, are rejected under 35 U.S.C. § 101 as being directed to non-statutory subject matter
Claim Rejections - 35 USC § 103
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.
Claims 1, 13, 16, 18 and 26 are rejected under 35 U.S.C. 103) as being anticipated by Jurtz (J Immunol. 2017 Oct 4) in view of Rosati (BMC Biotechnology volume 17, Article number: 61 (2017)).
Regarding the recited in claim 1, Obtaining data representing peptide sequences of each of a
set of neoantigens.
Jurtz teaches, the use of neoantigen data from melanoma samples “Predictive performance evaluated in terms of rank of neo-antigens identified in four melanoma samples” (Under figure 10, Lines 1-2)
Regarding the recited in claim 1, encoding the peptide sequences of each of the neoantigens into a corresponding numerical vector.
Jurtz teaches, peptides were encoded using BLOSUM- based encoding “All peptides were represented as 9-mer binding cores by the use of insertions and deletions as described by Andreatta et al. (4) and encoded using BLOSUM encoding “(Neural network training section, 4th paragraph, Lines 2-4), which reads on the formation of numerical vector.
Regarding the recited in claim 1, to generate a set of presentation likelihoods for the set of neoantigens, each presentation likelihood in the set representing the likelihood that a corresponding neoantigen is presented by the one or more class II MHC alleles on the surface of the tumor cells of the subject.
Jurtz teaches NetMHCpan-4 to score peptides from tumor samples “Having demonstrated the increased predictive power of the BA+EL method when it comes to prediction of peptide binding affinity (the BA+EL BA model), likelihood of being an eluted ligand (BA+EL EL model), and the ability of capturing the MHC specific peptide length binding preferences (also the BA+EL EL model) ( The NetMHCpan-4.0 method section), the use of “likelihood of being an eluted ligand” reads on the presentation likelihood.
Regarding the recited in claim 1, a label obtained by mass spectrometry measuring presence of peptides bound to at least one class II MHC allele in a set of class II MHC alleles identified as present in the sample.
Jurtz teaches the use of spectrometry data from immunopeptidomes experiments “Recently an increasing amount of MHC presented peptides identified by mass spectrometry has been published containing information about peptide processing steps in the presentation pathway and the length distribution of naturally presented peptides.” (Abstract). The peptides are referred to by Jurtz by eluted peptides and each is associated with at least one MHC alle, “Data on all class I MHC ligand elution assays available in IEDB database (www.iedb.org) were collected including the ligand sequence, details of the source protein, position of the ligand in the source protein and the restricting allele of the ligand” (Data sets section, 1rst paragraph, Lines 1-3)
Regarding the recited in claim 1, training peptide sequences encoded as numerical vectors including information regarding a plurality of amino acids that make up the peptides and a set of positions of the amino acids in the peptides. Jurtz teaches ““Data on all class I MHC ligand elution assays available in IEDB database (www.iedb.org) were collected including the ligand sequence, details of the source protein, position of the ligand in the source protein and the restricting allele of the ligand” (Data sets section, 1rst paragraph, Lines 1-3)
Regarding the recited in claim1, a function representing a relation between the numerical vector received as input and the presentation likelihood generated as output based on the numerical
vector and the parameters.
Jurtz teaches the input to the neural network is “The NNAlign training approach with insertions and deletions (3) was extended by adding a second output neuron as shown in figure 1. This was done to allow combined training on binding affinity and MS eluted ligand data” (Neural network training section, 1rst paragraph, Lines 1-3) the input is represented by a BLOUSM encoding which reads on numerical vector “encoding “All peptides were represented as 9-mer binding cores by the use of insertions and deletions as described by Andreatta et al. (4) and encoded using BLOSUM encoding “(Neural network training section, 4th paragraph, Lines 2-4), which reads on the formation of numerical vector.
Regarding the use of a function, Jurtz teaches the use of a neural network and a hidden layer to pass input to output as shown in figure 1. Neural network is a function f(x, θ) where x is the input and θ is the parameters used by the neural network and the output is represented by y=f(x; θ). As evidence by Gibiansky .
Jurtz teaches an output layer that consists of 2 separate neurons one for binding affinities and one for eluted ligand likelihood as shown in figure one.
So, the use of neural network reads on the use of a function that links input vectors to output predictions with mapping governed by its internal parameters or weights.
Regarding the recited in claim 1, selecting a subset of the set of neoantigens based on the set of presentation likelihoods to generate a set of selected neoantigens.
Jurtz ranked candidate neoantigens based on predicted likelihood “Predictive performance evaluated in terms of rank of neo-antigens identified in four melanoma samples. A rank value of 1 corresponds to the ligand obtaining the highest score” (Figure10, Lines 1-2) in addition, Jartz teaches the top 25 neoantigens “Both the NetMHCpan-4.0 and MixMHCpred method proposed by Bassani-Sternberg et al. (14) identify the known neoantigens within the top 25 peptides in 6 out 10 cases.” (Identification of cancer neoantigens section, Lines 16-18), the top ranked neoantigens reads on selecting a subset of neoantigens based on the likelihoods.
Regarding the recited in claim1, identifying one or more T-cell receptors (TCR) of the one or more identified T-cells.
Jurtz identified T-cells epitopes “In this work, we focus on demonstrating the improved prediction performance not only on large sets of MS peptidome data but also on T cell epitope data independent from the data used to train the new predictor” Introduction, 7thParagraph, Lines 3-5).
Regarding the recited in claim 1, wherein the identification of one or more T-cell receptors comprises sequencing T-cell receptor sequences of the one or more identified T-cells.
Jurtz did not identify T-cells receptors by sequencing.
Rosati teaches “The extreme diversity of the TCR repertoire represents a major analytical challenge; this has led to the development of specialized methods which aim to characterize the TCR repertoire in-depth. Currently, next generation sequencing based technologies are most widely employed for the high-throughput analysis of the immune cell repertoire.” (Abstract)
Jurtez applied the method on MHC class I, a person of ordinary skill in the art, would apply the same method on HMC class II, as we apply the same methodology using different data and adjusted parameters.
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Rosati. The combination would create an end-to-end pipeline from computational prediction to biological confirmation, increasing confidence in the selected neoantigens. There is a likelihood of success since the methods used are well known in the art and the techniques have been used in different research.
Regarding claim 13
Regarding the recited in claim 13, the plurality of samples comprises at least one of:
one or more human cell lines obtained or derived from a plurality of patients.
Jutez teaches, “Predictive performance evaluated in terms of rank of neo-antigens identified in four melanoma samples.” (Figure 10, Lines 1-2)
Regarding claim 16
Regarding the recited in claim 16, (a) predicted affinity between a neoantigen in the set of neoantigens and the one or more class II MHC alleles.
Jurtz teaches “Binding affinity was then predicted for the original ligands and random peptide sets for their corresponding alleles” (Data sets, 3rd paragraph, Lines 7-8)
Regarding the recited in claim 16, (b) predicted stability of the neoantigen encoded peptide-MHC complex.
Jurtz teaches the ability to predict the complex stability “The machine-learning framework proposed here is not limited to the integration of MHC class I peptide binding affinity and MS peptidome data. The approach can readily be extended to integrate other types of relevant data including MHC binding stability” (Discussion, 7th paragraph, Lines 1-3).
Regarding claim 18
Regarding the recited in claim 18, selecting neoantigens that have an increased likelihood of being presented on the cell surface relative to unselected neoantigens based on the machine-learned presentation model.
Jurtez teaches the idea of selecting neoantigens with a higher likelihood of being presented based on a machine learning model “The results also confirm the earlier findings presented here, that NetMHCpan-4.0 achieves improved performance compared to that of version 3.0, and that the ligands in all cases are predicted with very strong eluted ligand likelihood values (all percentile rank values are less than 1, and the majority are less than or equal to 0.02).( Identification of cancer neoantigens, 1rst paragraph, last 4 lines).
Regarding the recited in claim 26, the machine-learned presentation model is a neural network model.
Jurtez teaches the use of neural network “A neural network ensemble was trained as described by Andreatta et al. (1) using 5-fold nested cross-validation. Networks with 60 and 70 hidden neurons were trained leading to an ensemble of 40 networks in total.” (Neural network training, 3rd paragraph)
Claim 2 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Nielsen (BMC Bioinformatics. 2007).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Regarding the recited in claim 2, applying the machine-learned presentation model to the peptide sequence of the neoantigen.
Jurtz teaches the use of neural network as a machine learning model “The inputs to the neural networks consisted of the peptide and the MHC molecule in terms of a pseudo sequence (8). All peptides were represented as 9-mer binding cores by the use of insertions and deletions as described by Andreatta et al.” (Neural network training section, 4th paragraph, Lines 1-4)
Regarding the recited in claim 2, generate a dependency score for each of the one or more class II MHC alleles indicating whether the class II MHC allele will present the neoantigen based on the particular
amino acids at the particular positions of the peptide sequence.
Regarding the use of dependency score.
Nielson generated IC50 score for each peptide “The SMM-align method seeks to identify a weight matrix that optimally reproduces the measured IC50 values for each peptide in the training set.” (Background, 4th paragraph)
Regarding the use of Per position scoring of peptide MHC.
Jurtz did not teach per position scoring.
Nielsen teaches “a weight matrix where weight is assigned to each amino acid at each of the 9 positions
For each allele, a weight matrix describing the binding motif was constructed based on the relevant data in the SYF data set and the set of binding peptides in the IEDB training set, using the Gibbs sampler method as described by Nielsen et al. [4]. An IC50 value of 500 nM was used identify peptide binders from the IEDB data set.” (Gibb’s sampler weight matrices section).
Regarding the recited, neoantigen based on the particular
amino acids at the particular positions of the peptide sequence.
Nielsen teaches an equation from which the binding affinity is calculated and the position scoring represented as weight at each position is included in the equation “The predicted binding affinity for a peptide sequence is determined as the highest nonamer peptide score within the peptide, where a nonamer peptide score is calculated as “EQUATION” where wla' is the binding motif weight at position l for amino acid a', and vala', is the sequence-encoding value for amino acid a' for amino acid a.”( The SMM-align method section, 2nd paragraph)
Regarding if the score indicates the class II allele will present a neoantigen or not, “Binding peptides were identified using an IC50 binding threshold of 500 nm.” (Under table 5)
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Nielson teachings. The combination would improve the accuracy of the model prediction by analyzing the impact of individual amino acids in peptides at each position. In addition, provides a deeper insight of MHC binding mechanics. There is a likelihood of success, since the combination is stacking multiple well known biologically grounded and independently validated filters to predict the most relevant candidates.
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 6, 18 and 26 above, in view of Stranzl (Immunogenetics. 2010 Ap) and in view of Łuksza (Nature. 2017 Nov 8) and Stranzl (Immunogenetics. 2010)
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Regarding claim 3
Regarding the recited in claim 3, transforming the dependency scores to generate a corresponding per-allele likelihood for each class II MHC allele indicating a likelihood that the corresponding class II MHC allele will present the corresponding neoantigen.
Stranzl calculated dependency scores from which presented.
Stranzl teaches “Here, we have developed a pan-specific MHC class I epitope predictor, NetCTLpan. The method integrates prediction of proteasomal cleavage, TAP transport efficiency, and MHC binding into a MHC class I pathway presentation likelihood score” (Discussion, 2nd paragraph, Lines 1-3). The determination of proteasomal cleavage, TAP transport efficiency, together form a prior probability that a peptide could be presented, reads on dependency scores. proteasomal cleavage and TAP transport efficiency scores are used to generate presentation likelihood score.
Regarding transforming the dependency score, the specification discloses “The scores for each MHC allele hare combined, and transformed by the transformation function. Fl"(P.00362) to generate the presentation likelihood that peptide sequence u1c = f(x~ • 0 2 + x~ • 0 3), where .,:l, X3k are the identified allele-interacting variables for alleles h 2, hc3, and h. 03 are the set of parameters determined for MHC alleles h=2. h=3.)” (P.00363), The limitation is interpreted as the weighted sum of dependency scores. Stranzl teaches “The NetCTLpan prediction value is defined as a weighted sum of the three individual prediction values for MHC class I affinity, TAP transport efficiency, and C-terminal proteasomal cleavage. (Combined class I pathway presentation prediction—NetCTLpan)
Stranzl applied the method on MHC class I, a person of ordinary skill in the art, would apply the same method on HMC class II, as we apply the same methodology using different data and adjusted parameters.
Jurtz and Stranzl did not generate pre-likelihood.
Tuksuza teaches the pre-likelihood of a neoantigen being presented by the MHC (Major Histocompatibility Complex) and the likelihood of T-cell recognition contribute to the overall fitness (or immunogenicity) of the neoantigen, which ultimately predicts response to immunotherapy. “Here, we present a fitness model for tumors based on immune interactions of neoantigens that predicts response to immunotherapy. Two main factors determine neoantigen fitness: its likelihood of presentation by the major histocompatibility complex (MHC) and its subsequent T-cell recognition. We estimate these two components using a neoantigen’s relative MHC binding affinity and a non-linear dependence on its sequence similarity to known antigens.” (Abstract)
Regarding the recited in claim 3, combining the per-allele likelihoods to generate the presentation likelihood of the neoantigen.
Tuksuza teaches a model that combines a neoantigen's relative MHC binding affinity and a non-linear dependence on its sequence similarity to known antigens to estimate the likelihood of neoantigen presentation by the major histocompatibility complex (MHC) and subsequent T-cell recognition. (Abstract)
It would have been obvious to a person of ordinary skill in the art to combine Jurtz, Stanzl and Tuksuza. The combination would include the dependency scores and pre-likelihood to calculate presented likelihood. The combination would improve the peptides predictions by reducing the false positives and improves the neoantigen ranking for therapies. There is a likelihood of success since the methods used are well known in the art and have been individually validated in the context of antigen processing and T cell epitope prediction. The concept of combining scores to generate a final presented likelihood is well known in the art and used in different research and capacities.
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Ozcan (Oncoimmunology. 2018) and Stranzl (Immunogenetics. 2010 Ap).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Regarding claim 4
Regarding the recited in claim 4, transforming the dependency scores models the presentation of the neoantigen as mutually exclusive across the one or more class II MHC alleles.
Jurtz did not explicitly consider neoantigen as mutually exclusive across the one or more class II MHC alleles.
Ozcan used mutual exclusive as a tool to ass weather mutations in different genes related to HLA class I antigen presentation tend to occur together “To evaluate further a possible functional role of the mutations affecting genes encoding components of HLA class I antigens, analysis of mutual exclusivity was performed. Mutual exclusivity was significant for B2M and HLA-B genes and co-occurence was observed for HLA-A and HLA-B; HLA-A and HLA-C; and HLA-B and HLA-C genes.” (Results section, Mutations of HLA class I-related genes in MSI cancer subsection, 2nd paragraph). If a method can be applied to MHC class 1 molecules, it can be extended to class MHC class II, since both classes share core features in piptide-MHC binding and T cell recognition.
Ozscan used the method on MHC class I, a person of ordinary skill in the art, would apply the same method on HMC class II, as we apply the same methodology using different data and adjusted parameters.
In addition, Stranzl implies the use of mutually exclusive in neoantigen presentation “we suggest using the per-protein measure, since pooling data from different proteins and HLA alleles will place ligands in a nonbiological competition for presentation.” (Combined class I pathway presentation prediction—NetCTLpan, 4th paragraph, Lines 5-6).
It would have been obvious to a person or ordinary skill in the art to combine Jurtz, Stranzl and Ozcan. The combination would increase the specificity of the prediction and supports the development of focused and clinically safe immunotherapies. There is a likelihood of success since the methods used are well known in the art and support by studies before the filling date of this application.
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Brown (Oncoimmunology. 2018 Dec 22)
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Regarding the recited in claim 5, wherein transforming the combination of the dependency scores models the presentation of the neoantigen as interfering between the one or more class II MHC alleles.
Jurtz did not consider neoantigen as interfering between the one or more class II MHC alleles.
Brown emphasizes the concept that a single neoantigen can be presented by multiple MHC molecules. Brown teaches “From these predictions, we observe that most peptides are able to be presented by relatively few (< 250) MHC, while some can be presented by upwards of 1,500 different MHC.” (Abstract). If a method can be applied to MHC class 1 molecules, it can be extended to class MHC class II, since both classes share core features in piptide-MHC binding and T cell recognition.
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Brown. The combination would account for real world HLA diversity in neoantigen presentation, capture competitive binding and enable multiple allele binding analysis. There is a likelihood of success since, the methods used are well known and established in the art.
Claims 6 and 8 are rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Stranzl (Immunogenetics. 2010 Ap).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Regarding claim 6
Regarding the recited in claim 6, applying the machine-learned presentation model to the allele noninteracting features to generate a dependency score for the allele noninteracting features indicating whether the peptide sequence of the corresponding neoantigen will be presented based on the allele noninteracting features.
Jurtz did not teach dependency scores.
Stranzl deliberately avoided modeling interactions between MHC alleles, so features used were non-interacting “we suggest using the per-protein measure, since pooling data from different proteins and HLA alleles will place ligands in a nonbiological competition for presentation” (Combined class I pathway presentation prediction—NetCTLpan, 4th paragraph, Lines 5-6)
Stranzl teaches predications are done for each allele independently and ranked them “The rank-score is calculated as the percent rank of a given NetCTLpan likelihood score to a set of 200,000 random natural 9-mer peptides.” (Discussion,5th paragraph, Lines 10-12), using the sum of three scores to do the prediction, which reads on the use of dependency score for each prediction separately “The NetCTLpan prediction value is defined as a weighted sum of the three individual prediction values for MHC class I affinity, TAP transport efficiency, and C-terminal proteasomal cleavage. Optimal relative weights on TAP transport efficiency and proteasomal cleavage were estimated using the training data set and based on the average AUC value per HLA class I ligand pair” (Combined class I pathway presentation prediction—NetCTLpan), The calculation of the weighted sum reads on the calculation of the presented likelihood.
Regarding claim 8
Regarding the recited in claim 8, combining the dependency scores for each of the class II MHC alleles and the dependency score for the allele noninteracting features; and transforming the combined dependency scores to generate the presentation likelihood.
Regarding transforming the dependency score, the specification discloses “The scores for each MHC allele hare combined, and transformed by the transformation function. Fl"(P.00362) to generate the presentation likelihood that peptide sequence u1c = f(x~ • 0 2 + x~ • 0 3), where .,:l, X3k are the identified allele-interacting variables for alleles h 2, hc3, and h. 03 are the set of parameters determined for MHC alleles h=2. h=3.)”(P.00363), The limitation is interpreted as the weighted sum of dependency scores. Stranzl teaches “The NetCTLpan prediction value is defined as a weighted sum of the three individual prediction values for MHC class I affinity, TAP transport efficiency, and C-terminal proteasomal cleavage.” (Combined class I pathway presentation prediction—NetCTLpan)
Regarding the use of noninteracting features, Stranzl deliberately avoided modeling interactions between MHC alleles, so features used were non-interacting “we suggest using the per-protein measure, since pooling data from different proteins and HLA alleles will place ligands in a nonbiological competition for presentation” (Combined class I pathway presentation prediction—NetCTLpan, 4th paragraph, Lines 5-6)
Stranzl calculated dependency scores for each of the class MHC alleles from which presented likelihood is calculated.
Stranzl teaches “Here, we have developed a pan-specific MHC class I epitope predictor, NetCTLpan. The method integrates prediction of proteasomal cleavage, TAP transport efficiency, and MHC binding into MHC class I pathway presentation likelihood score” (Discussion, 2nd paragraph, Lines 1-3). Reads on predicting MHC binding affinities
Stranzl applied the method on MHC class I, a person of ordinary skill in the art, would apply the same method on HMC class II, as we apply the same methodology using different data and adjusted parameters.
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Stranzl. The combination would improve the prediction accuracy by reducing the false positives, the dependency scores would provide structural and functional insights into the biological significance of peptides.
There is a likelihood of success since the combination merely involves applying know and complementary methods and techniques to improve the prediction specificity.
Claim 7 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Barra (Genome Medicine volume 10, Article number: 84 (2018)
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Regarding claim 7
Regarding the recited in claim 7, combining the dependency score for each class II MHC allele in the one or more class II MHC alleles with the dependency score for the allele noninteracting features; transforming the combined dependency scores for each class II MHC allele to generate a
per-allele likelihood for each class II MHC allele indicating a likelihood that the corresponding
class II MHC allele will present the corresponding neoantigen; and combining the per-allele likelihoods to generate the presentation likelihood.
Jurtz did mot teach combining dependency scores and generating pre-likelihood.
Barra implies the use of independent features “This observation was found consistently in all data sets independent of MHC II restriction and host species (human or mouse). (Discussion, 3rd paragraph, last 3 lines)
Barra also implied the use of dependency scores such as motif frequencies “we extracted the amino acid frequencies from the binding motifs displayed in Fig. 2” (Training prediction models on MHC class II ligand data, 2nd paragraph, Lines 13-15), identified binding affinities “The traditional approach to quantify peptide MHC-II binding relies on measuring binding affinity, either as the dissociation constant (Kd) of the complex [12, 13] or in terms of IC50 (concentration of the query peptide which displaces 50% of a bound reference peptide)” (Discussion, 2nd paragraph, Lines 5-8).
Barra teaches first stage of output which implies on the generation of a prei-ikelihood,
“Combined models were trained as described earlier [25] with both binding affinity and eluted ligand data as input.” (NNAlign modeling and architecture, 2nd paragraph, Lines 6-7). The combined models are then integrated to predict likelihood scores “Lastly, we integrated this information associated with antigen processing into a machine learning framework and demonstrated a consistently improved predictive performance” (Discussion, 3rd paragraph, Lines 1-2).
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Barra. The combination would include the dependency scores and pre-likelihood to calculate presented likelihood. The combination would improve the peptides predictions by reducing the false positives and improves the neoantigen ranking for therapies. There is a likelihood of success since the methods used are well known in the art and have been individually validated in the context of antigen processing and T cell epitope prediction. The concept of combining scores to generate a final presented likelihood is well known in the art and used in different research and capacities.
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Farrell (Create an MHC-Class I binding predictor in Python November 12 2018).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Regarding the recited in claim 12, wherein encoding the peptide sequence comprises encoding the peptide sequence using a one-hot encoding scheme.
Jurtz did not teach one-hot encoding.
Farrell teaches “The first and simplest is a so-called ‘one-hot encoding’ of the amino acids producing a 20-column vector for each position that only contains a 1 where the letter corresponds to that amino acid.” (One hot encoding).
It would be obvious to a person of ordinary skill in the art to combine Farrell and Jutz. The combination would transform biological sequence into machine readable format suitable for deep learning models. There is a likelihood of success since the methods used are well known in the art and have been widely used in computational immunology.
Claim 14 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Baumgartner (J Immunol (2010)).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Regarding the recited in claim 14, (a) data associated with peptide-MHC binding affinity measurements for at least one of the peptides.
Jutez teaches each peptide used in training has a binding affinity value (IC50) “Binding affinity values are measured as IC50 values in nM (aff) and can be rescaled to the interval [0,1] by applying 1-log(aff)/log(50,000), representing continuous target values” (Neural network training)
Regarding the recited in claim 14, (b) data associated with peptide-MHC binding stability measurements for at least one of the peptides.
Jurtz teaches the stability of the peptide:complex but did not teach peptide stability measurements.
Jurtz teaches “However, other factors including antigen processing and the stability of the peptide:MHC complex (13) could influence the likelihood of a given peptide to be presented as an MHC ligand.(Introduction, 4th paragraph, Lines 4-7)
Baumgartner measured peptide stability “we immunized mice with wild-type or mutated peptides displaying varying binding half-lives with MHC class II molecules to measure the impact of peptide-MHC class II stability on the clonal composition of the CD4 T cell response.” (Abstract).
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Baumgartner. The combination would ensure the predicted complex stability and the ability to be recognized by T cells. Integrating the stability measurements would yield more accurate and reliable predictions. There is a likelihood of success since the combination brings together 2 complementary, well known in the art and validated methodologies that enhance the prediction accuracy.
Claim 15 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16 and 18 above, in view of Sternberg (Nature Communications volume 7, Article number: 13404 (2016)).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Regarding the recited in claim 15, the set of presentation likelihoods are further identified by at least expression levels of the one or more class II MHC alleles in the subject, as measured by RNA-seq or mass spectrometry.
Jurtz teaches, the use of mass spectroscopy “(the eluted ligand likelihood prediction value of the model trained on the combined binding affinity and eluted ligand data) The methods were evaluated on all binding affinity (all_BA) data and all eluted ligand” (Under figure 2) the eluted ligand was generated from mass spectroscopy “This was done to allow combined training on binding affinity and MS eluted ligand data.” (Neural network training), the MS eluted ligand data reads on expression levelsof MHC.
Sternberg teaches “We conclude that direct identification of mutated peptide ligands from primary tumour material by MS is possible and yields true neoepitopes with high relevance for immunotherapeutic strategies in cancer.” (Abstract) and “We normalized for each patient the number of peptides derived from three selected TAA (PMEL, tyrosinase and PRAME) to the total amount of eluted peptides from the respective patient and correlated peptide presentation to RNA and protein expression of the defined TAA” (Results, Peptide ligands derived from tumour-associated antigens, Lines 8-10)
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Sternberg. The combination would enhance the prediction power with direct experiment validation from mass spectroscopy. The combination would quantify and identify which predicted peptides are truly presented on tumor cells. There is a likelihood of success, since the methods used are well known in the art and have been widely adapted in immunology and neoantigen discovery pipelines.
Claim 17 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16 and 18 above, in view of Andreatta (Immunogenetics. 2016).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Jurtz is applied to claim 15 above.
Regarding claim 17, (a) the C-terminal sequences flanking the neoantigen encoded peptide sequence within its source protein sequence. And (b) the N-terminal sequences flanking the neoantigen encoded peptide sequence within its source protein sequence.
Jurtz did not include the C-terminal.
Andreatta teaches “C- and N-terminal PFR lengths (LPFR) were each encoded using two input neurons with values LPFR/(LPFR+1) and 1-LPFR/(LPFR+1) respectively. The peptide length L was encoded with two input neurons taking the values LPEP and 1-LPEP, where LPEP=1/(1+exp((L-15)/2)).” (Neural network architecture and training, Lines 7-10)
It would have been obvious to one of person of ordinary skill in the art to combine Jurtz and Andretta. The combination would improve prediction power by improving the true positive rate, since it accounts for upstream and downstream biological processing events and would enable design of better peptide vaccines. There is a likelihood of success because the flanking sequence context is well established in the art and has been shown in numerous studies to enhance biological relevance.
Claim 19 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16 and 18 above, in view of Brendan (Nature Biotechnology. 2018).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Jurtz is applied to claim 15 above.
Jurtz is applied to claim 17 above.
Regarding the recited in claim 19
Regarding the recited in claim 19, selecting neoantigens that have an increased likelihood of being capable of inducing a tumor-specific immune response in the subject relative to unselected neoantigens based on the machine-learned presentation model.
Jurtz ranked neoantigens based on their predicted likelihood of being presented by MHC molecules, so inherently teaching the likelihood of being capable of inducing a tumor-specific immune response “Predictive performance evaluated in terms of rank of neo-antigens identified in four melanoma samples” (Figure1)
Brendan teaches a model that can be applied to select peptides that are confirmed to activated T-cells “We show that our model, named EDGE, increases the positive predictive value of HLA antigen prediction by up to ninefold. We apply EDGE to enable identification of neoantigens and neoantigen-reactive T cells using routine clinical specimens and small numbers of synthetic peptides for most common HLA alleles. EDGE could enable an improved ability to develop neoantigen-targeted immunotherapies for cancer patients.” (Abstract).
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Brendan. The combination would improve specificity by eliminating shared or non-immunogenic peptides, which lays a strong foundation for personalized immunotherapy. There is a likelihood of success since, all methods used are well known in the art before the filling date of this application.
Claim 20 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Zhuting (Nature Reviews Immunology volume 18, pages168–182 (2018))
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Jurtz is applied to claim 15 above.
Jurtz is applied to claim 17 above.
Jurtz is applied to claim 19 above.
Regarding the recited in claim 20, selecting neoantigens that have an increased likelihood of being capable of being presented to native T-cells by professional antigen presenting cells (APCs) relative to unselected neoantigens based on the presentation model, optionally wherein the APC is a dendritic cell (DC).
Jurtez did not teach increased likelihood of being capable of being presented to naitve T-cells by professional antigen presenting cells.
Zhuting teaches, identifying potentially immunogenic tumor-specific peptides (neoepitopes) that are likely to be recognized by the immune system. “Comprehensive mutational analysis is carried out through whole-exome sequencing and neoepitopes encoded by somatic mutations in the tumour are selected that have the highest probability of being presented by the individual's MHC molecules on the basis of affinity predictions” (Towards increasingly personalized vaccines, 3rd paragraph, Lines 2-5)
Zhuting also implies choosing neoantigens that T cells can recognize and respond to is a key strategy in effective immunotherapy and highlights that neoantigen-reactive T cells are crucial for tumor cell killing and that increased neoantigen load correlates with better patient outcome, Zhuting teaches “Increased neoantigen load is associated with improved patient outcomes, and neoantigen-reactive T cell populations expand in settings of effective immunotherapy and mediate tumour cell killing in preclinical models and patients.” (Under key points, 3th point)
Zhuting teaches dendritic cells loaded with peptides, reading on APC is a dendritic cell “Three clinical trials of neoantigen-based vaccines in patients with melanoma, using dendritic cells loaded with short peptides” (Under key points, 4th point), which reads on optionally wherein the APC is a dendritic cell (DC).
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Zhuting. The combination would improve the identification of functionally relevant neoantigens and ensures the presented peptides will be recognized by T-cells. The is a likelihood of success since, each component of the teachings is well established and backed by precedent research.
Claims 21 and 22 is rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Yarchoan (Nat Rev Cancer. 2017) and Zhuting (Nature Reviews Immunology volume 18, pages168–182 (2018))
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Jurtz is applied to claim 15 above.
Jurtz is applied to claim 17 above.
Jurtz is applied to claim 19 above.
Jurtz is applied to claim 20 above.
Regarding the recited in claim 21, selecting neoantigens that have a decreased likelihood of being subject to inhibition via central or peripheral tolerance relative to unselected neoantigens based on the machine-learned presentation model.
Jurtz did not select selecting neoantigens that have a decreased likelihood of being subject to inhibition via central or peripheral tolerance.
Selecting neoantigens with a lower chance of being inhibited by tolerance mechanisms increases their immunogenicity,
Zhuting teaches, developing personalized neoantigen vaccines, selecting neoantigens that are less likely to be subject to central or peripheral tolerance “Now, through integration of tumour sequencing with the prediction of MHC-binding epitopes, it is possible to identify candidate tumour neoantigens on a per patient basis. Growing evidence supports their immunogenicity and their use for developing personalized vaccines.” (Tumour antigens, 4th paragraph, Lines 6-9). In addition, Yarchoan teaches the concept of neoantigens can be selected based on the recognition by immune system “In this Review we discuss the emerging evidence that neoantigens are recognized by the immune system and can be targeted to increase antitumour immunity. We also provide a framework for personalized cancer immunotherapy through the identification and selective targeting of individual tumour neoantigens, and present the potential benefits and obstacles to this approach of targeted immunotherapy.” (Abstract)
Regarding the recited in claim 22, selecting neoantigens that have a decreased likelihood of being capable of inducing an autoimmune response to normal tissue in the subject relative to unselected neoantigens based on the machine-learned presentation model.
Yarchoan teaches “These neoantigens are an attractive immune target because their selective expression on tumours may minimize immune tolerance as well as the risk of autoimmunity.” (Abstract), In addition, Zhuting teaches “By selecting suitable antigen targets, a potent and tumour-specific immune response can be induced while minimizing autoimmunity” (main, 4th paragraph, Lines 4-6)
It would have been obvious to a person of ordinary skill in the art to combine Jurtz, Zhuting and Yarchoan. The combination would improve the model’s specificity by the selection of truly tumor specific neoantigens and lower the risk of autoimmune toxicity. There is a likelihood of success since each step in both teachings is well established, validated and supported by literature and clinical results.
Claims 27-30 are rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Burden (Journal of Molecular Graphics and Modelling. 2005)
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Jurtz is applied to claim 15 above.
Jurtz is applied to claim 17 above.
Jurtz is applied to claim 19 above.
Jurtz is applied to claim 20 above.
Jurtz is applied to claims 21 and 22 above.
Regarding the recited in claim 27
Jurtz did not teach the use of more than one neural network model
Regarding the recited in claim 27, the neural network model includes a plurality of network models for the class II MHC alleles, each network model assigned to a corresponding class II MHC allele of the class II MHC alleles and including a series of nodes arranged in one or more layers.
Jurtz did not teach plurality of networks.
Regarding the use plurality of network models, Burden teaches “We have employed Bayesian neural network methods to build QSAR models” (Introduction, 5th paragraph), and used 2 neural models to represent two HLA protein “We have employed Bayesian neural network methods to build QSAR models explaining the more complex MHC class II-binding activity of peptides to two HLA protein alleles “(Introduction, 5th paragraph), in addition “We derived four models for each training data set (101 and 301)” (3.1. Modelling of the training data). Regarding the use of nodes or neurons “We used a three-layer neural network architecture containing a single hidden layer with one or two neurodes to determine how well the Bayesian net could train and generalize. (2.5. Protein-peptide QSAR modelling, Lines 2-4)
Regarding claim 28
Regarding the recited in claim 28, each of the one or more convolutional neural networks including a series of nodes arranged in one or more layers.
Burden teaches “We used a three-layer neural network architecture containing a single hidden layer with one or two neurodes to determine how well the Bayesian net could train and generalize.” (2.5. Protein-peptide QSAR modelling, Lines 2-4)
Regarding the recited in claim 28, having a filter of a different size, the filter of each of the one or more convolutional neural networks sized to identify the positions of the amino acids in the peptide sequence of each neoantigen that comprise a binding core or a binding anchor of the peptide sequence.
Regarding the use of amino acid positions, Burden teaches “They assumed that binding affinity was an additive function of the contributions of amino acids in each position of the peptide, essentially a type of Free-Wilson approach, with additional allowance for interactions between a given amino acid and its neighbors” (Introduction, 2nd paragraph, Lines 7-8). Regarding the use of filter or kernel “At each amino acid position one of 20 indicator variables is set to 1 to denote the presence of that amino acid. As the peptide motifs used in building the models were nonamers, and there are 20 possible amino acids at each position” (2.4.1. Binary (Bin20) descriptors). this approach teaches a filter, also known as a binding site preference for a specific amino acid sequence. By representing each amino acid position with 20 indicator variables, the model learns to recognize and prioritize nonamer motifs with particular amino acid compositions, essentially creating a filter that favors those patterns. Regarding using a binding anchor or motif, Burden teaches “Peptides that bind to these MHCs have recognition motifs consisting of nine amino acids.” (2.1. Training data sets, Lines 6-8), in addition, Burden teaches “Buus described how privileged binding motifs exist in peptide binders and how QSAR methods could be used to build predictive models of human immune reactivities (Introduction, 2nd paragraph, Lines 4-5)
Regarding the recited in claim 29, the neural network model is trained by updating the parameters of the neural network model, and wherein the parameters of at least two network models are jointly updated for at least one training iteration.
Burden teaches “Pareff is the number of effective parameters used by the neural net (3.1. Modelling of the training data, 2nd paragraph). Regarding updating the neural network parameters, neural networks automatically update their parameters (weights and biases) with each iteration during training, as evidence by Jagadeesha “This iterative process of updating weights based on gradients is repeated until the model converges to an optimal configuration, improving its ability to make accurate predictions.” (Backpropagation).
Regarding claim 30
Regarding the recited n claim 30, the machine-learned presentation model is a deep learning model that includes one or more layers of nodes.
Deep learning is a type of machine learning that uses neural networks, Brenden teaches “We used a three-layer neural network architecture containing a single hidden layer with one or two neurodes to determine how well the Bayesian net could train and generalize.” (2.5. Protein-peptide QSAR modelling, Lines 2-4)
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Burden. Th combination would enhance the predictive power by using neural network with multiple layers and nodes, having the ability to combine and process diverse data types. There is a likelihood of success since all methods and techniques used are well known in the art, before the filling date of this application.
Claims 31 are rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Youssef (Journal of Molecular Graphics and Modelling. 2005)
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Jurtz is applied to claim 15 above.
Jurtz is applied to claim 17 above.
Jurtz is applied to claim 19 above.
Jurtz is applied to claim 20 above.
Jurtz is applied to claims 21 and 22 above.
Jurtz is applied to claims 27-30.
Regarding the recited in claim 31
Regarding the recited in claim 31, identifying the one or more T-cells comprises co-culturing the one or more T-cells with one or more of the neoantigens in the subset under conditions that expand the one or more T-cells.
Jurtz does not teach co- culturing of T-cells
Yossef teaches “Illustration of the new high-throughput approach for enrichment, culturing, and screening strategy of TILs.” (Figure 1), which is equivalent to culturing individual T-cells with specific neoantigens. Regarding expand one or more T-cells Yossef teaches “TIL cultures that contain reactive T cells are further expanded and reinfused to the patient” (Introduction, 2nd paragraph, Line 3).
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Yossef teachings. The combination would provide an efficient and powerful computational tool to predict neoantigens and the ability to apply functional validation via applying T cell co -cultures. The combination is a unique platform for therapy and research. There is a likelihood of success, since it integrates well established, complementary techniques from immunoinformatic, peptide immunology and T cell biology, each of which is individually validated and well known in the art.
Claims 32 and 35 are rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16, 18 and 26 above, in view of Davis (JCI Insight. 2018).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Jurtz is applied to claim 15 above.
Jurtz is applied to claim 17 above.
Jurtz is applied to claim 19 above.
Jurtz is applied to claim 20 above.
Jurtz is applied to claims 21 and 22 above.
Jurtz is applied to claims 27-30.
Jurtz is applied to claim 31 above.
Regarding claim 32
Jurtz did not teach multimer.
Regarding the recited in claim 32, contacting the one or more T-cells with an MHC multimer.
Davis teaches “Labelling antigen-specific T cells with peptide–MHC multimers has provided an invaluable way to monitor T cell-mediated immune responses.” (Abstract)
Regarding the recited in claim 32, one or more of the neoantigens in the subset under conditions that allow binding between the T-cells and the MHC multimer.
Regarding allow binding between the T-cells and the MHC multimer, David teaches “This led some of us (J.D.A. and M.M.D.) and colleagues to try different ways of multimerizing peptide–MHC complexes to improve their binding characteristics” (Main, Lines 13-15).
Regarding the recited in claim 35
Jurtz did not teach isolation of T-cells
Regarding the recited in claim 35, An isolated T-cell that is antigen-specific for at least one
selected neoantigen.
Davis was able to able to purify or select T-cells with particular specificity “All of these forms contain multiple peptide–MHC complexes or other T cell ligands that form multiple bonds with TCRs to achieve stable binding and, therefore, can be used to label and purify T cells of a particular specificity.” (Main, 2nd paragraph, Lines 7-9)
It would have been obvious to a person or ordinary skill in the art to combine Jurtz and Davis, the combination would ensure fast and broad coverage of neoantigen and T cells discovery workflow, the use of multimer and functional assays will provide a definitive proof of immunogenicity. There is a likelihood of success since all methods uses are well known in the art, using well-known elements and techniques in a predictable manner to achieve a known result.
Claims 36 and 38 are rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16 and 18 above, in view of Sharpe (Dis Model Mech. 2015 Ap) and in view of Zhuting Hu (Blood: 2018).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Jurtz is applied to claim 15 above.
Jurtz is applied to claim 17 above.
Jurtz is applied to claim 19 above.
Jurtz is applied to claim 20 above.
Jurtz is applied to claims 21 and 22 above.
Jurtz is applied to claims 27-30.
Jurtz is applied to claim 31 above.
Jurtz is applied to claims 32 and 35 above.
Regarding the recited in claim 36
Jurtz did not include genetic engineered or expanded T-cells.
Regarding the recited in claim 36, genetically engineering a plurality of T-cells to express at least one of the one or more identified T-cell receptors.
Regarding the use of genetically engineered T-cells, Sharpe teaches “The potential of these approaches has been demonstrated in particular by the successful use of genetically modified T cells to treat B cell haematological malignancies in clinical trials.” (Abstract), which reads on the use of modified T-cells with identified receptor, here can bind and treat B cells haematological malignancies.
Regarding the recited in claim 36, culturing the plurality of T-cells under conditions that expand the plurality of T-cells; and infusing the expanded T-cells into the subject.
Regarding expanding the T-cells, Zhuting Hu teaches “Generating TCR-expressing cell lines has several advantages over transduction of primary T cells, including the ability to expand and maintain a fixed repertoire of TCR-expressing cells indefinitely” (Discussion, 2nd paragraph, Lines 4-6).
Regarding infusion of cells, Sharpe teaches “Studies have indicated that T-cell survival and proliferation in vivo might be dependent on the differentiation status of the infused T cells” (Manufacturing challenges, 3rd paragraph, Lines 4-6).
Regarding the recited in claim 38
Jurtz did not include genetic expanded T-cells and infusing them.
Regarding the recited in claim 38, culturing the one or more identified T-cells under conditions that expand the one or more identified T-cells; and infusing the expanded T-cells into the subject.
Regarding expanding the T-cells, Zhuting Hu teaches “Generating TCR-expressing cell lines has several advantages over transduction of primary T cells, including the ability to expand and maintain a fixed repertoire of TCR-expressing cells indefinitely” (Discussion, 2nd paragraph, Lines 4-6).
Regarding infusion of cells, Sharpe teaches “Studies have indicated that T-cell survival and proliferation in vivo might be dependent on the differentiation status of the infused T cells” (Manufacturing challenges, 3rd paragraph, Lines 4-6)
It would have been obvious to a person of ordinary skill in the art to combine Jurtz, Sharpe and Hu. The combination would provide a workflow from which predicted peptides identified by Jurtz can be cloned and expanded. In addition, genetic engineering allows the introduction of a high affinity T cell receptor (TCR) into abundant, healthy T cells overcoming natural scarcity. The combination would satisfy the goal of identifying clinically actionable tumor antigens for therapeutic intervention. There is a likelihood of success since each step is well known in the art and their combination merely involves the application of known techniques to achieve predictable results.
Claim 37 are rejected under 35 U.S.C. 103 as being unpatentable over Jurtz (J Immunol. 2017 Oct 4), as applied to claims 1, 13, 16 and 18 above, in view of Zhuting Hu (Blood: 2018).
Jurtz is applied to claims 1, 13, 16, 18 and 26 above.
Jurtz is applied to claim 2 above.
Jurtz is applied to claim 3 above.
Jurtz is applied to claim 4 above.
Jurtz is applied to claim 5 above.
Jurtz is applied to claims 6 and 8 above.
Jurtz is applied to claim 7 above.
Jurtz is applied to claim 12 above.
Jurtz is applied to claim 14 above.
Jurtz is applied to claim 15 above.
Jurtz is applied to claim 17 above.
Jurtz is applied to claim 19 above.
Jurtz is applied to claim 20 above.
Jurtz is applied to claims 21 and 22 above.
Jurtz is applied to claims 27-30.
Jurtz is applied to claim 31 above.
Jurtz is applied to claims 32 and 35 above.
Jurtz is applied to claims 36 and 38 above.
Regarding the recited in claim 37
Jurtz did not teach cloning.
Regarding the recited in claim 37, cloning the T-cell receptor sequences of the one or more identified T-cells into an expression vector.
Regarding the use of expression vector which are used in cloning, Zhuting Hu teaches the use of PEW lentiviral vector “Jurkat∆αβ cells were transduced to express a nuclear factor of activated T cells (NFAT)-luciferase construct9,10 in PEW lentiviral vector (Jurkat∆αβ reporter cells)” (Methods, Human peripheral blood mononuclear cell samples and cell lines, Lines 4-5).
Regarding the recited in claim 37, transfecting each of the plurality of T-cells with the expression vector.
Zhuting Hu teaches "The TCR construct was assembled and transferred to lentiviral vector PEW and used to transduce Jurkat∆αβ reporter cells" implies that transfection with an expression vector (the TCR construct) occurred. Transduction, in this context, refers to the process of delivering genetic material (in this case, the TCR construct) into the Jurkat∆αβ cells using a virus (lentiviral vector). This process is functionally equivalent to transfection. (TCR cloning and expression system, Lines 7-8).
It would have been obvious to a person of ordinary skill in the art to combine Jurtz and Hu teachings. The combination would provide a workflow from which predicted peptides identified by Jurtz can be cloned and expanded. In addition, the neoantigens-specific TCRs are transduced into T-cells for potential use in immunology. The combination would satisfy the goal of identifying clinically actionable tumor antigens for therapeutic intervention. There is a likelihood of success since each step is well known in the art and their combination merely involves the application of known techniques to achieve predictable results.
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine
grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or
improper timewise extension of the "right to exclude" granted by a patent and to prevent
possible harassment by multiple assignees. A nonstatutory double patenting rejection is
Page 19
appropriate where the conflicting claims are not identical, but at least one examined application
claim is not patentably distinct from the reference claim(s) because the examined application
claim is either anticipated by, or would have been obvious over, the reference claim(s). See,
e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d
1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Langi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir.
1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d
438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA
1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321 (c) or 1.321 (d) may be
used to overcome an actual or provisional rejection based on nonstatutory double patenting
provided the reference application or patent either is shown to be commonly owned with the
examined application, or claims an invention made as a result of activities undertaken within the
scope of a joint research agreement. See MPEP § 717.02 for applications subject to
examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159.
See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to
file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR
1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory
double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be
accompanied by a reply requesting reconsideration of the prior Office action. Even where the
NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1.
For a reply to a non-final Office action, see 37 CFR 1.111 (a). For a reply to final Office action,
see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c)
may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
Application/Control Number: 18/150,390
Art Unit: 1672
Page 20
The USPTO Internet website contains terminal disclaimer forms which may be used.
Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in
which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or
PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely
online using web-screens. An eTerminal Disclaimer that meets all requirements is autoprocessed
and approved immediately upon submission. For more information about eTerminal
Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
12. Claims 1-4, 8, 12, 14 and 15 are provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1, 2, 4, 7-8, 12 and 14-15 of copending Application No. 16403331 (reference application).
Regarding the recited in the following claims:
Instant application claim 1 “training peptide sequences encoded as numerical vectors including information regarding a plurality of amino acids that make up the peptides and a set of positions of the amino acids in the peptides”.
Reference application claim 1” training peptide sequences encoded as numerical vectors including information regarding a plurality of amino acids that make up the peptide sequences and a set of positions of the amino acids in the peptide sequences”.
Instant application claim 1” each presentation likelihood in the set representing the likelihood that a corresponding neoantigen is presented by the one or more class II MHC alleles on the surface of the tumor cells of the subject”
Reference application claim 1 “each presentation likelihood in the set representing the likelihood that a corresponding antigen is presented by one or more class IMHC alleles on the surface of the tumor cells of the subject”
Instant application claim 2” generate a dependency score for each of the one or more class II MHC alleles indicating whether the class II MHC allele will present the neoantigen based on the particular amino acids at the particular positions of the peptide sequence.”
Reference application claim 2 “generating a dependency score for each of the one or more MHC alleles indicating whether the MHC allele will present the neoantigen based on the particular amino acids at the particular positions of the peptide sequence.”
Instant application claim 3“transforming the dependency scores to generate a corresponding per-allele likelihood for each class II MHC allele indicating a likelihood that the corresponding class II MHC allele will present the corresponding neoantigen; and combining the per-allele likelihoods to generate the presentation likelihood of the neoantigen..”
Reference application claim 7 “transforming the dependency scores to generate a corresponding per-allele likelihood for each MHC allele indicating a likelihood that the corresponding MHC allele will present the corresponding neoantigen”
Referred application claim 4” transforming the dependency scores models the presentation of the neoantigen as mutually exclusive across the one or more MHC alleles.”
Reference application claim 4 “transforming the combination of the dependency scores models the presentation of the neoantigen as interfering between the one or more class II MHC alleles.
Instant application claim 8 “combining the dependency scores for each of the class II MHC alleles and the dependency score for the allele noninteracting features; and transforming the combined dependency scores to generate the presentation likelihood.”
Reference application claim 8” combining the dependency scores for each of the MHC alleles and the dependency score for the allele noninteracting features; and (b) transforming the combined dependency scores to generate the presentation likelihood.”
Instant application claim 12 “encoding the peptide sequence comprises encoding the peptide sequence using a one-hot encoding scheme.”
Reference application claim 12 “encoding the peptide sequence comprises encoding the peptide sequence using a one-hot encoding scheme.”
Instant application claim 14 “(a) data associated with peptide-MHC binding affinity measurements for at least one of the peptides and (b) data associated with peptide-MHC binding stability measurements for at least one of the peptides.
Reference application claim 14 “(a) data associated with peptide-MHC binding affinity measurements for at least one of the training peptide sequences; (b) data associated with peptide-MHC binding stability measurements for at least one of the training peptide sequences.”
Instant application claim 15 “the set of presentation likelihoods are further identified by at least expression levels of the one or more class IIMHC alleles in the subject, as measured by RNA-seq or mass spectrometry.”
Reference application claim 15 “the set of presentation likelihoods are further identified by at least expression levels of the one or more MHC alleles in the subject, as measured by RNA-seq or mass spectrometry.
This is a provisional nonstatutory double patenting rejection because the patentably
indistinct claims have not in fact been patented.
Conclusion
No claims are allowed.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MARWY ELATTAR whose telephone number is (571)272-1182. The examiner can normally be reached full time.
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, Olivia Wise can be reached on 5712722248. 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.E./ Examiner, Art Unit 1685
/LARRY D RIGGS II/ Supervisory Patent Examiner, Art Unit 1686
Prosecution Insights