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 .
Support for the amendments is within the instant application specification.
Applicant’s amendment to the claims filed on 12/16/2025 in response to the Non-Final Rejection mailed on 9/17/2025 is acknowledged. This listing of claims replaces all prior listings of claims in the application.
Claims 13, 20, 22-24, 26, 28, 31-33, 38, 53, 64 are pending and examined on the merits to the extent they read of species elected SEQ ID NO: 4 and Glycan 5402.
Claims 1-12, 14-19, 21, 25, 27, 29-30, 34-37, 39-52, 54-63 are canceled.
Applicant’s remarks filed on 12/16/2025 in response to the Non-Final Rejection mailed on 9/17/2025 have been fully considered and are deemed persuasive to overcome at least one of the rejections and/or objections as previously applied.
The text of those sections of Title 35 U.S. Code not included in the instant action can be found in the prior Office Action.
Withdrawn Rejections
The rejection of claims 13, 20, 22-24, 26, 28, 31 under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph is withdrawn in view of Applicant’s addition of the different glycans into the instant application claims.
Maintained 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.
The rejection of claims 13, 20, 22-24, 28, 31, 33, 53 under 35 U.S.C. 103 as being unpatentable over Miyamoto et al (2018, J. Proteome Res, cited on IDS filed 1/9/2025) {herein Miyamoto} in view of Kourou et al (2014, Computational and Structural Biotechnology Journal, cited on PTO-892 dated 9/17/2025) {herein Kourou} is maintained. The rejection has been modified in view of Applicant’s amendment of claims 13, 33, 53 to recite ‘wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known.’
As amended, claims 13, 20, 22-24, 28, 31 are drawn to a method for identifying a classification for a biological sample from a patient or individual having a disease or condition and monitoring the health status of the patient, the method comprising quantifying by multiple reaction monitoring mass spectroscopy (MRM-MS) one or more glycopeptides in a sample using a Triple QuadrupoleMass Spectrometer (QQQ) and/or qTOF mass spectrometer; wherein the one or more glycopeptides comprises at least one glycopeptide consisting essentially of an amino acid sequence selected from the group consisting of SEQ ID NOs: 1 -262 and wherein the glycopeptide comprises glycan 5402 and at least one of glycans 3200, 3210, 3300, 3310, 3320, 3400, 3410, 3420, 3500, 3510, 3520, 3600, 3610, 3630,3700,3710, 3720,3730, 3740,4200,4210,4300,4301,4310,4311,4320,4400, 4401,4410,4411,4420,4421,4430,4431,4500,4501,4510,4511,4520,4521,4530, 4531,4540,4541,4600,4601,4610,4611,4620,4621,4630,4631,4641,4650,4700, 4701,4710,4711,4720,4730, 5200,5210, 5300,5301, 5310,5311, 5320,5400,5401, 5402,5410,5411, 5412,5420, 5421,5430, 5431,5432, 5500.5501, 5502,5510,5511, 5512,5520, 5521,5522, 5530,5531,5541, 5600,5601, 5602,5610, 5611,5612, 5620, 5621,5631,5650, 5700,5701, 5702,5710, 5711,5712, 5720,5721, 5730,5731,6200, 6210,6300,6301, 6310,6311, 6320,6400, 6401,6402, 6410,6411, 6412,6420,6421, 6432,6500,6501,6502, 6503,6510, 6511,6512,6513, 6520,6521, 6522,6530, 6531, 6532,6540,6541, 6600,6601, 6602,6603, 6610,6611, 6612,6613, 6620,6621,6622, 6623,6630,6640, 6641,6642, 6652,6700, 6701,6703, 6710,6711, 6712,6713,6720, 6721,6730,6731, 6740,7200,7210,7400,7410,7411,7412,7420,7421,7430,7431, 7432,7500,7501,7510,7511,7512,7600,7601,7602,7603,7604,7610,7611,7612, 7613,7614,7620,7621,7622,7623,7640,7700,7701,7702,7703,7710,7711,7712, 7713,7714,7720,7721,7722,7721,7722,7730,7731,7732,7740,7741,7751, 8200,9200, 9210, 10200, 11200, and 12200 in Table 4; wherein the method further comprises detecting the glycosylation site residue where the glycan bonds to the glycopeptide, or detecting a glycosylation site on a glycopeptide; and inputting the quantification into a trained model to generate an output probability, wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known; determining if the output probability is above or below 0.30 for a classification; and identifying a classification for the sample based on whether the output probability is above or below 0.30 wherein the classification is having ovarian cancer or not having ovarian cancer and the classification is used to monitor the onset and progression of disease in the patient.
As amended, claim 33 is drawn to a method for classifying a biological sample, comprising: obtaining a biological sample from a patient, wherein the biological sample comprises one or more glycoproteins; digesting and/or fragmenting one or more glycoproteins in the sample wherein the glycoprotein comprises glycan 5402 and at least one of glycans 3200, 3210, 3300, 3310, 3320, 3400, 3410, 3420, 3500, 3510, 3520, 3600, 3610, 3630, 3700, 3710, 3720, 3730, 3740, 4200, 4210, 4300, 4301, 4310, 4311, 4320, 4400, 4401, 4410, 4411, 4420, 4421, 4430, 4431, 4500, 4501, 4510, 4511, 4520, 4521, 4530, 4531, 4540, 4541, 4600, 4601, 4610, 4611, 4620, 4621, 4630, 4631, 4641, 4650, 4700, 4701, 4710, 4711, 4720, 4730, 5200, 5210, 5300, 5301, 5310, 5311, 5320, 5400, 5401, 5402, 5410, 5411, 5412, 5420, 5421, 5430, 5431, 5432, 5500. 5501, 5502, 5510, 5511, 5512, 5520, 5521, 5522, 5530, 5531, 5541, 5600, 5601, 5602, 5610, 5611, 5612, 5620, 5621, 5631, 5650, 5700, 5701, 5702, 5710, 5711, 5712, 5720, 5721, 5730, 5731, 6200, 6210, 6300, 6301, 6310, 6311, 6320, 6400, 6401, 6402, 6410, 6411, 6412, 6420, 6421, 6432, 6500, 6501, 6502, 6503, 6510, 6511, 6512, 6513, 6520, 6521, 6522, 6530, 6531, 6532, 6540, 6541, 6600, 6601, 6602, 6603, 6610, 6611, 6612, 6613, 6620, 6621, 6622, 6623, 6630, 6640, 6641, 6642, 6652, 6700, 6701, 6703, 6710, 6711, 6712, 6713, 6720, 6721, 6730, 6731, 6740, 7200, 7210, 7400, 7410, 7411, 7412, 7420, 7421, 7430, 7431, 7432, 7500, 7501, 7510, 7511, 7512, 7600, 7601, 7602, 7603, 7604, 7610, 7611, 7612, 7613, 7614, 7620, 7621, 7622, 7623, 7640, 7700, 7701, 7702, 7703, 7710, 7711, 7712, 7713, 7714, 7720, 7721, 7722, 7721, 7722, 7730, 7731, 7732, 7740, 7741, 7751, 8200, 9200, 9210, 10200, 11200, and 12200 in Table 4; detecting and quantifying at least one or more multiple-reaction-monitoring (VIRM) transition using a Triple Quadrupole Mass Spectrometer (QQQ) and/or qTOF mass spectrometer selected from the group consisting of transitions 1-150, wherein the method further comprises detecting the glycosylation site residue where the glycan bonds to the glycopeptide, or detecting a glycosylation site on a glycopeptide; and inputting the quantification into a trained model to generate a output probability, wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known; determining if the output probability is above or below 0.30; and classifying the biological sample based on whether the output probability is above or below 0.30.
As amended, claim 53 is drawn to a method for diagnosing ovarian cancer in an individual comprising: performing mass spectroscopy of a biological sample obtained from the individual using MRM-MS with a QQQ and/or qTOF spectrometer to detect and quantify one or more glycopeptides consisting essentially of an amino acid sequence selected from the group consisting of SEQ ID NOs:1-262; or to detect one or more MRM transitions selected from transitions 1-150 wherein the glycopeptide comprises glycan 5402 and at least one of glycans 3200, 3210, 3300, 3310, 3320, 3400, 3410, 3420, 3500, 3510, 3520, 3600, 3610, 3630, 3700, 3710, 3720, 3730, 3740, 4200, 4210, 4300, 4301, 4310, 4311, 4320, 4400, 4401, 4410, 4411, 4420, 4421, 4430, 4431, 4500, 4501, 4510, 4511, 4520, 4521, 4530, 4531, 4540, 4541, 4600, 4601, 4610, 4611, 4620, 4621, 4630, 4631, 4641, 4650, 4700, 4701, 4710, 4711, 4720, 4730, 5200, 5210, 5300, 5301, 5310, 5311, 5320, 5400, 5401, 5402, 5410, 5411, 5412, 5420, 5421, 5430, 5431, 5432, 5500. 5501, 5502, 5510, 5511, 5512, 5520, 5521, 5522, 5530, 5531, 5541, 5600, 5601, 5602, 5610, 5611, 5612, 5620, 5621, 5631, 5650, 5700, 5701, 5702, 5710, 5711, 5712, 5720, 5721, 5730, 5731, 6200, 6210, 6300, 6301, 6310, 6311, 6320, 6400, 6401, 6402, 6410, 6411, 6412, 6420, 6421, 6432, 6500, 6501, 6502, 6503, 6510, 6511, 6512, 6513, 6520, 6521, 6522, 6530, 6531, 6532, 6540, 6541, 6600, 6601, 6602, 6603, 6610, 6611, 6612, 6613, 6620, 6621, 6622, 6623, 6630, 6640, 6641, 6642, 6652, 6700, 6701, 6703, 6710, 6711, 6712, 6713, 6720, 6721, 6730, 6731, 6740, 7200, 7210, 7400, 7410, 7411, 7412, 7420, 7421, 7430, 7431, 7432, 7500, 7501, 7510, 7511, 7512, 7600, 7601, 7602, 7603, 7604, 7610, 7611, 7612, 7613, 7614, 7620, 7621, 7622, 7623, 7640, 7700, 7701, 7702, 7703, 7710, 7711, 7712, 7713, 7714, 7720, 7721, 7722, 7721, 7722, 7730, 7731, 7732, 7740, 7741, 7751, 8200, 9200, 9210, 10200, 11200, and 12200 in Table 4; wherein the method further comprises detecting the glycosylation site residue where the glycan bonds to the glycopeptide, or detecting a glycosylation site on a glycopeptide; inputting the quantification of the detected glycopeptides or the MRM transitions into a trained model to generate an output probability, wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known, determining if the output probability is above or below 0.30; and identifying a diagnostic classification for the individual based on whether the output probability is above or below 0.30 a threshold for a classification; and diagnosing the individual as having ovarian cancer or not having ovarian cancer based on the diagnostic classification.
With respect to claims 13, 28, 31, 33, 53, Miyamoto teaches a multiple reaction monitoring (MRM)- based method for the protein and site-specific quantitation of serum glycoproteins for the identification and classification of ovarian cancer in subjects (abstract). The method comprises enzymatic digestion/fragmentation and analysis of glycopeptides using MRM-MS on triple quadruploe (QQQ) spectrometers for the accurate quantitation of proteins from biological specimens (page 223, column 1, para 2); page 227, column 2, para 1). Eighteen transitions were monitored by QQQ for protein and protein subclass quantitation (table 2). The glycoprotein A1AT which is comprised of the peptide sequence AVLTIDEK was identified as a glycoprotein within the biological sample from patient with ovarian cancer (table 2). Said peptide sequence is 100% identical to SEQ ID NO: 4 of the instant application claims 13, 53 (appendix A). Said glycoprotein is comprised of glycan 5402, which is H5N4S2 and glycan 6503, which is also represented as H6N5S3 (page 223, column 2, para 3). Said glycans are attached to site N107 of the glycoprotein (page 223, column 2, para 3). Absent evidence otherwise, it is the Examiner’s position that said site is the site where the glycan bonds to the glycoprotein in light of the teaching by Miyamoto that the glycan is attached to the N107 site in the glycoprotein (page 223, column 2, para 3). Miyamoto further teaches sites of glycosylation were identified on the glycoprotein A1AT (table 3). Absent evidence otherwise, it is the Examiner’s position that said teaching is detecting a glycosylation site on a glycoprotein. The Global Protein Machine was used to analyze peptide profiles and identify unique tryptic peptides (page 223, column 2, para 3; page 225, column 1, para 1). Data was input into an in-house built software tool GPFinder (page 226, column 2, para 1). Miyamoto further teaches a method wherein statistical analysis is utilized as a predictive model for determining the probability of a subject having ovarian cancer, based on diagnostic biomarkers (page 230, column 2, para 2). A training set consisting of 40 cases and 40 controls were analyzed, and differential analyses was performed to identify aberrant glycopeptide levels (abstract). The p-value for subjects with A1AT glycoprotein is 0.00001 (Table S3).
With respect to claim 22, since the art teaches the structure of a multiple reaction monitoring (MRM)- based method for the protein and site-specific quantitation of serum glycoproteins for the identification and classification of ovarian cancer in subjects (abstract), it is the Examiner’s position that the classification of serum glycoproteins from subjects with ovarian cancer would necessarily be identified with greater than 80% confidence.
With respect to claims 23-24, Miyamoto teaches a method wherein 18 samples were quantitated by MRM-MS at different time points and compared based on their peptide sequence, precursorion, production, collision energy and retention time (table 2). Absent evidence otherwise, it is the Examiner’s position that the retention times in table 2 are different time points, based on the glycoprotein.
However, Miyamoto does not teach inputting the quantification into a trained model to generate an output probability, wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known (claims 13, 33, 53); determining if the output probability is above or below 0.30 for a classification; and identifying a classification for the sample based on whether the output probability is above or below 0.30 (claim 13). Miyamoto does not teach a method wherein the trained model was trained using a machine learning algorithm selected from the group consisting of a deep learning algorithm, a neural network algorithm, an artificial neural network algorithm, a supervised machine learning algorithm, a linear discriminant analysis algorithm, a quadratic discriminant analysis algorithm, a support vector machine algorithm, a linear basis function kernel support vector algorithm, a radial basis function kernel support vector algorithm, a random forest algorithm, a genetic algorithm, a nearest neighbor algorithm, k-nearest neighbors, a naive Bayes classifier algorithm, a logistic regression algorithm, or a combination thereof (claim 20). Miyamoto does not teach inputting the quantification of the detected glycopeptides or the MRM transitions into a trained model to generate an output probability, wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known, determining if the output probability is above or below 0.30 (claim 53).
With respect to claims 13, 20, 33, 53, Kourou teaches a method wherein machine learning software is used as a model for studying cancer risk and outcomes in patients (abstract). Kourou further teaches there are two main common types of ML methods known as (i) supervised learning and (ii) unsupervised learning (page 6, column 1, para 6). In supervised learning a labeled set of training data is used to estimate or map the input data to the desired output (page 6, column 1, para 6). ML also includes ANNs (Artificial Neural Networks) that handle a variety of classification or pattern recognition problems (page 10, column 2, para 4) by detecting specific biomarkers (page 160, column 1, para 2). Absent evidence otherwise, it is the Examiner’s position that biomarkers and glycoproteins are the same as both are comprised of protein that allows for identification of samples. ANNs is trained to generate an output as a combination between the input variables (page 10, column 2, para 4). Furthermore, ANNs is the gold standard for classification of cancer samples (page 10, column 2, para 4). The classifiers performance is assessed based on accuracy and are under the curve (page 10, column 1, para 4). The quantitative metrics of accuracy and AUC are used for assessing the overall performance of a classifier (page 10, column 1, para 4; column 2, para 1). Kourou further teaches 48.774 mammographic findings as well as demographic risks factors and tumor characteristics were considered (page 13, column 2, para 2). Absent evidence otherwise, it is the Examiner’s position that ‘48.774 mammographic findings as well as demographic risks factors and tumor characteristics’ are ‘multiple factors in combination’ upon which Kourou relied upon for the identification, classification and prediction of cancers. In addition, Kourou teaches this dataset was then fed as input to the ANN model (page 13, column 2, para 1). The calculated AUC of their model was 0.965 following training and testing by means of ten-fold cross validation (page 13, column 2, para 1). The model can accurately estimate the risk assessment of breast cancer patients by integrating a large data sample (page 13, column 2, para 1). The model is unique among others if we consider that the most important factors they used to train the ANN model are the mammography findings with tumor registry outcomes (page 13, column 2, para 1). Regarding the recitation ‘determining if the output probability is above or below 0.30 for a classification’ within the instant application claims 13, 33, 53, MPEP 2106.04(a)(2)C states ‘A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. There is no particular word or set of words that indicates a claim recites a mathematical calculation. That is, a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.’ As such, the recitation of ‘‘determining if the output probability is above or below 0.30 for a classification’ within the instant application claims 13, 33, 53 is an abstract idea. According to MPEP 2106.04(a), ‘The abstract idea exception has deep roots in the Supreme Court’s jurisprudence. See Bilski v. Kappos, 561 U.S. 593, 601-602, 95 USPQ2d 1001, 1006 (2010) (citing Le Roy v. Tatham, 55 U.S. (14 How.) 156, 174–175 (1853)). Despite this long history, the courts have declined to define abstract ideas. However, it is clear from the body of judicial precedent that software and business methods are not excluded categories of subject matter. For example, the Supreme Court concluded that business methods are not "categorically outside of § 101's scope," stating that "a business method is simply one kind of ‘method’ that is, at least in some circumstances, eligible for patenting under § 101." Bilski, 561 U.S. at 607, 95 USPQ2d at 1008 (2010). See also Content Extraction and Transmission, LLC v. Wells Fargo Bank, 776 F.3d 1343, 1347, 113 USPQ2d 1354, 1357 (Fed. Cir. 2014) ("there is no categorical business-method exception"). Likewise, software is not automatically an abstract idea, even if performance of a software task involves an underlying mathematical calculation or relationship. See, e.g., Thales Visionix, Inc. v. United States, 850 F.3d 1343, 121 USPQ2d 1898, 1902 (Fed. Cir. 2017) ("That a mathematical equation is required to complete the claimed method and system does not doom the claims to abstraction."); McRO, Inc. v. Bandai Namco Games Am. Inc., 837 F.3d 1299, 1316, 120 USPQ2d 1091, 1103 (Fed. Cir. 2016) (methods of automatic lip synchronization and facial expression animation using computer-implemented rules were not directed to an abstract idea); Enfish, 822 F.3d 1327, 1336, 118 USPQ2d 1684, 1689 (Fed. Cir. 2016) (claims to self-referential table for a computer database were not directed to an abstract idea).
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to apply the teachings of Miyamoto of a method wherein a multiple reaction monitoring (MRM)- based method for the protein and site-specific quantitation of serum glycoproteins for the identification and classification of ovarian cancer in subjects (abstract) or combine the teachings of Kourou because Kourou teaches a method wherein machine learning software is used as a model for studying cancer risk and outcomes in patients (abstract).
One of ordinary skill in the art would be motivated to either use the teachings of Miyamoto et al. by itself or combine the teachings of Kourou because Kourou provides the motivation for Miyamoto to input the quantification of glycoproteins analyzed by MRM-MS and QQQ into a trained model/machine learning model to generate an output probability because it would allow for better classification of patients at risk of and diagnosed with cancer, allow for more appropriate selection of therapeutic agents and better patient outcomes (Kourou: abstract). Furthermore, utilizing a trained model would improve the understanding of cancer progression in subjects (Kourou: abstract). One of ordinary skill in the art knowing the benefit of identifying a classification and being able to accurately quantitate the different types of glycoprotein patterns based on cancer types of a biological sample from a patient having or at risk of developing ovarian cancer based on the teachings of Miyamoto and Kourou would have a reasonable expectation of success that inputting the quantification taught by Miyamoto into the trained model taught by Kourou would result in the classification of cancer and the ability to monitor the onset and progression of disease in patients as Kourou teaches various supervised machine learning techniques that have resulted in an improved understanding of cancer progression in subjects and the ability to better understand the molecular basis of cancer (page 12, column 2, para 2).
One of skill in the art would have a reasonable expectation of success to make and use the claimed method for identifying a classification for a biological sample from a patient or individual having a disease or condition and monitoring the health status of the patient because Miyamoto provides a multiple reaction monitoring (MRM)- based method for the protein and site-specific quantitation of serum glycoproteins and the identification and classification of ovarian cancer in subjects (abstract). Whereas Kourou provides the teachings of a method wherein machine learning software is utilized as a model for studying cancer risk and outcomes in patients (abstract). Therefore there would be a reasonable expectation of success to arrive at the above invention. Therefore, the above invention would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention.
RESPONSE TO REMARKS: Applicant's arguments filed 12/16/2025 have been fully considered but they are not persuasive. Beginning on p. 12 of Applicants’ remarks, Applicants in summary contends that the references fail to disclose or suggest generating an output probability as recited in the present claims. Applicant contends that Miyamoto does not disclose or otherwise suqqest the trained model, nor does it suqqest the generation of an output probability.
The arguments are not persuasive. Examiner contends that Kourou was dependent upon to provide Miyamoto with the motivation to utilize machine learning techniques, including ANNs, as a mechanism for prognostic detection and disease prognosis (Kourou: page 11, column 2, para 4). These techniques can discover and identify patterns and relationships between them, from complex datasets, while they are able to effectively predict future outcomes of a cancer type (Kourou: page 9, Column 1, para 2). Examiner contends that Kourou teaches there are two main common types of ML methods known as (i) supervised learning and (ii) unsupervised learning (Kourou: page 9, column 1, para 5). In supervised learning a labeled set of training data is used to estimate or map the input data to the desired output (Kourou: page 9, column 1, para 5).
Applicant contends that the trained model recited in the present claims compares the quantification to multiple factors in combination, rather than the differential analysis described in Miyamoto.
The arguments are not persuasive. Examiner contends that the recitation of ‘multiple factors in combination’ is indefinite as it is unclear which ‘multiple factors in combination’ Applicant is referencing by that statement. Nevertheless, Kourou teaches 48.774 mammographic findings as well as demographic risks factors and tumor characteristics were considered (page 13, column 2, para 2). Absent evidence otherwise, it is the Examiner’s position that ‘48.774 mammographic findings as well as demographic risks factors and tumor characteristics’ are ‘multiple factors in combination’ upon which Kourou relied upon for the identification, classification and prediction of cancers. Kourou further teaches this dataset was then fed as input to the ANN model (Kourou: page 13, column 2, para 1). The calculated AUC of their model was 0.965 following training and testing by means of ten-fold cross validation (page 13, column 2, para 1). The authors claimed that their model can accurately estimate the risk assessment of breast cancer patients by integrating a large data sample (Kourou: page 13, column 2, para 1). They also declared that their model is unique among others if we consider that the most important factors they used to train the ANN model are the mammography findings with tumor registry outcomes (Kourou: page 13, column 2, para 1).
Applicant contends that none of the cited references disclose or otherwise describe the output probability threshold of 0.3 as recited in the present claims.
The arguments are not persuasive. Examiner contends that MPEP 2106.04(a)(2)C states ‘A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. There is no particular word or set of words that indicates a claim recites a mathematical calculation. That is, a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.’ As such, the recitation of ‘‘determining if the output probability is above or below 0.30 for a classification’ within the instant application claims 13, 33, 53 is an abstract idea. According to MPEP 2106.04(a), ‘The abstract idea exception has deep roots in the Supreme Court’s jurisprudence. See Bilski v. Kappos, 561 U.S. 593, 601-602, 95 USPQ2d 1001, 1006 (2010) (citing Le Roy v. Tatham, 55 U.S. (14 How.) 156, 174–175 (1853)). Despite this long history, the courts have declined to define abstract ideas. However, it is clear from the body of judicial precedent that software and business methods are not excluded categories of subject matter. For example, the Supreme Court concluded that business methods are not "categorically outside of § 101's scope," stating that "a business method is simply one kind of ‘method’ that is, at least in some circumstances, eligible for patenting under § 101." Bilski, 561 U.S. at 607, 95 USPQ2d at 1008 (2010). See also Content Extraction and Transmission, LLC v. Wells Fargo Bank, 776 F.3d 1343, 1347, 113 USPQ2d 1354, 1357 (Fed. Cir. 2014) ("there is no categorical business-method exception"). Likewise, software is not automatically an abstract idea, even if performance of a software task involves an underlying mathematical calculation or relationship. See, e.g., Thales Visionix, Inc. v. United States, 850 F.3d 1343, 121 USPQ2d 1898, 1902 (Fed. Cir. 2017) ("That a mathematical equation is required to complete the claimed method and system does not doom the claims to abstraction."); McRO, Inc. v. Bandai Namco Games Am. Inc., 837 F.3d 1299, 1316, 120 USPQ2d 1091, 1103 (Fed. Cir. 2016) (methods of automatic lip synchronization and facial expression animation using computer-implemented rules were not directed to an abstract idea); Enfish, 822 F.3d 1327, 1336, 118 USPQ2d 1684, 1689 (Fed. Cir. 2016) (claims to self-referential table for a computer database were not directed to an abstract idea).
Applicant contends that Kourou presents the use of supervised machine learning models in cancer prognosis and prediction. Applicant contends that Kourou suggests that decisions regarding cancer prognosis can be improved using data such as family history, age, diet, weight, high-risk habits, and exposure to environmental carcinogens. Applicant contends that the present claims feature methods comprising the quantification of one or more glycopeptides in a sample and utilizing a trained model to generate an output probability. One of skill in the art would understand that glycosylation profiles and the factors described in Kourou are significantly different factors in cancer prognosis and prediction. Applicant contends that Kourou teaches away from classifying a biological sample using a quantification of glycopeptides present in the sample. Applicant contends that Kourou refers to variables that are far different from the quantification recited in the present claims and implies that applying a machine learning model to a large set of features that are independent would result in better validation. Applicant contends that the claimed invention relies on identifying at least one glycopeptide that comprises one of SEQ ID Nos: 1 -262, and detecting glycosylation site residues, which is a far more specific and dependent metric in classifying a sample than the models described in Kourou.
The arguments are not persuasive. Examiner contends that Kourou teaches a method wherein machine learning software is used as a model for studying cancer risk and outcomes in patients (abstract). ML also includes ANNs (Artificial Neural Networks) that handles a variety of classification or pattern recognition problems (page 10, column 2, para 4) by detecting specific biomarkers (page 160, column 1, para 2). Absent evidence otherwise, it is the Examiner’s position that biomarkers and glycoproteins are the same as both are comprised of protein that allows for identification of samples. ANNs is trained to generate an output as a combination between the input variables (page 10, column 2, para 4). Furthermore, ANNs is the gold standard for classification of cancer samples (page 10, column 2, para 4). The classifiers performance is assessed based on accuracy and are under the curve (page 10, column 1, para 4). The quantitative metrics of accuracy and AUC are used for assessing the overall performance of a classifier (page 10, column 1, para 4; column 2, para 1).
The rejection of claims 26, 32 under 35 U.S.C. 103 as being unpatentable over Miyamoto et al (2018, J. Proteome Res, cited on IDS filed 1/9/2025) {herein Miyamoto} in view of Kourou et al (2014, Computational and Structural Biotechnology Journal, cited on PTO-892 dated 9/17/2025) {herein Kourou} as applied to claims 13, 20, 22-24, 28, 31, 33, 53 and in further view of Sandberg et al (WO 2009/075883 A2, Publication Date: 18 June 2009, cited in IDS filed 4/25/2022) {herein Sandberg} is maintained. The rejection has been modified in view of Applicant’s amendment of claims 13, 33, 53 to recite ‘wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known.’
Previously presented claim 26 is drawn to the method of claim 13, wherein monitoring the health status of the patient comprises monitoring the onset and progression of disease in the patient with risk factors including genetic mutations or detecting cancer recurrence.
Previously presented claim 32 is drawn to the method of claim 13, further comprising treating the individual with a therapeutically effective amount of a therapeutic agent selected from the group consisting of a chemotherapeutic, an immunotherapy, a hormone therapy, a targeted therapy, and combinations thereof.
The teachings of Miyamoto and Kourou as applied to claims 13, 20, 22-24, 28, 31, 33, 53 are set forth in the 103 rejection above.
However, Miyamoto and Kourou do not teach the method of claim 26, wherein monitoring the health status of the patient comprises monitoring the onset and progression of disease in the patient with risk factors including genetic mutations or detecting cancer recurrence (claim 26). The method of claim 32, further comprising treating the individual with a therapeutically effective amount of a therapeutic agent selected from the group consisting of a chemotherapeutic, an immunotherapy, a hormone therapy, a targeted therapy, and combinations thereof (claim 32).
With respect to claim 26, Sandberg teaches genetic polymorphisms within glycosylated enzymes play a role in ovarian cancer in patients, suggesting that genetic factors can affect glycosylation (page 68, lines 29-31).
With respect to claim 32, Sandberg teaches a method of identifying glycoproteins possessing a cancer-specific glycoform, as well as diagnostic and therapeutic methods and compositions related to glycoprotein cancer biomarkers (page 2, lines 9-11). Sandberg further teaches treating a subject having cancer or a precancerous condition by administering a therapeutic agent based on the diagnosis (page 29, lines 22-23). Such agents include chemotherapeutic agents (page 31, lines 11-12).
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to apply the teachings of Miyamoto of a method wherein multiple reaction monitoring (MRM)- based method is used for the protein and site-specific quantitation of serum glycoproteins for the identification and classification of ovarian cancer in subjects (abstract). Or combine the teachings of Kourou and Sandberg because Sandberg teaches a method of identifying glycoproteins possessing a cancer-specific glycoform, as well as diagnostic and therapeutic methods and compositions related to glycoprotein cancer biomarkers (page 2, lines 9-11). Whereas, Kourou teaches a method of using machine learning software as a model for studying cancer risk and outcomes in patients (abstract).
One of ordinary skill in the art would be motivated to either use the teachings of Miyamoto et al. by itself or combine the teachings of Kourou and Sandberg because Sandberg provides the motivation for one of ordinary skill in the art to administer an effective amount of a therapeutic agent (antibody) that specifically binds the cancer-specific glycoform if the classification indicates that the individual has ovarian cancer as said therapeutic agent could be conjugated to a cytotoxic agent, further enhancing toxicity to targeted cells (cancer cells) (page 31, lines 6-13). One of ordinary skill in the art knowing the benefit of identifying an individual with ovarian cancer based on the teachings of Miyamoto, Kourou and Sandberg would have a reasonable expectation of success that inputting the quantification taught by Miyamoto into the trained model taught by Kourou would result in the classification of whether a subject has cancer. There would be a reasonable expectation of success that utilizing the teaching of Miyamoto to identify the target glycoproteins and the method of Sandberg to target identified glycoforms by conjugating a cytotoxic agent, such as a chemotherapeutic agent, to a glycan-specific antibody to treat the subject (Sandberg: page 31, lines 6-13) would result in an effective method for targeting cancer cells. Said method would allow for an improved understanding of cancer progression in subjects (Kourou: page 12, column 2, para 2) and a better targeting of cancer cells in subjects.
One of skill in the art would have a reasonable expectation of success to make and use the claimed method for identifying a classification for a biological sample from a patient or individual having a disease or condition and monitoring the health status of the patient because Miyamoto provides multiple reaction monitoring (MRM)- based method for the protein and site-specific quantitation of serum glycoproteins for the identification and classification of ovarian cancer in subjects (abstract). Kourou provides the teachings of a method utilizing machine learning software as a model for studying cancer risk and outcomes in patients (abstract). Whereas, Sandberg provides the method of identifying glycoproteins possessing a cancer-specific glycoform, as well as diagnostic and therapeutic methods and compositions related to glycoprotein cancer biomarkers (page 2, lines 9-11). Therefore there would be a reasonable expectation of success to arrive at the above invention. Therefore, the above invention would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention.
RESPONSE TO REMARKS: Applicant's arguments filed 12/16/2025 have been fully considered but they are not persuasive. Beginning on p. 15 of Applicants’ remarks, Applicants in summary contends that claims 26 and 32 rely on the method of claim 13, which features generating an output probability using a trained model that compares the quantification to types, absolute amounts, or relative amounts of glycopeptides in at least one previously classified sample or to a reference value obtained from a population of individuals whose disease state is known and determining if the output probability is above or below 0.3. Applicant contends that those features are not mentioned or suggested in the cited references - as outlined above - and therefore one of skill in the art would have not have found it obvious to utilize a method with the features above to arrive at claims 26 and 32. Kourou teaches a method wherein machine learning software is used as a model for studying cancer risk and outcomes in patients (abstract).
The arguments are not persuasive. Examiner contends that Kourou teaches in supervised learning a labeled set of training data is used to estimate or map the input data to the desired output (Kourou: page 6, column 1, para 6). ML also includes ANNs (Artificial Neural Networks) that handle a variety of classification or pattern recognition problems (Kourou: page 10, column 2, para 4) by detecting specific biomarkers (Kourou: page 160, column 1, para 2). Absent evidence otherwise, it is the Examiner’s position that biomarkers and glycoproteins are the same as both are comprised of protein that allows for identification of samples. ANNs is trained to generate an output as a combination between the input variables (Kourou: page 10, column 2, para 4). The quantitative metrics of accuracy and AUC are used for assessing the overall performance of a classifier (Kourou: page 10, column 1, para 4; column 2, para 1). Examiner contends that MPEP 2106.04(a)(2)C states ‘A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. There is no particular word or set of words that indicates a claim recites a mathematical calculation. That is, a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.’ As such, the recitation of ‘‘determining if the output probability is above or below 0.30 for a classification’ within the instant application claims 13, 33, 53 is an abstract idea. According to MPEP 2106.04(a), ‘The abstract idea exception has deep roots in the Supreme Court’s jurisprudence. See Bilski v. Kappos, 561 U.S. 593, 601-602, 95 USPQ2d 1001, 1006 (2010) (citing Le Roy v. Tatham, 55 U.S. (14 How.) 156, 174–175 (1853)). Despite this long history, the courts have declined to define abstract ideas. However, it is clear from the body of judicial precedent that software and business methods are not excluded categories of subject matter. For example, the Supreme Court concluded that business methods are not "categorically outside of § 101's scope," stating that "a business method is simply one kind of ‘method’ that is, at least in some circumstances, eligible for patenting under § 101." Bilski, 561 U.S. at 607, 95 USPQ2d at 1008 (2010). See also Content Extraction and Transmission, LLC v. Wells Fargo Bank, 776 F.3d 1343, 1347, 113 USPQ2d 1354, 1357 (Fed. Cir. 2014) ("there is no categorical business-method exception"). Likewise, software is not automatically an abstract idea, even if performance of a software task involves an underlying mathematical calculation or relationship. See, e.g., Thales Visionix, Inc. v. United States, 850 F.3d 1343, 121 USPQ2d 1898, 1902 (Fed. Cir. 2017) ("That a mathematical equation is required to complete the claimed method and system does not doom the claims to abstraction."); McRO, Inc. v. Bandai Namco Games Am. Inc., 837 F.3d 1299, 1316, 120 USPQ2d 1091, 1103 (Fed. Cir. 2016) (methods of automatic lip synchronization and facial expression animation using computer-implemented rules were not directed to an abstract idea); Enfish, 822 F.3d 1327, 1336, 118 USPQ2d 1684, 1689 (Fed. Cir. 2016) (claims to self-referential table for a computer database were not directed to an abstract idea).
Applicant contends that the claimed invention achieves an improved method for classifying biological samples with respect to ovarian cancer. Examples 3 and 4 highlight the accuracy, sensitivity, and specificity of the model utilized in this invention. The study in Example 4 showed an accuracy of 91.9%, a sensitivity of 91.4%, and a specificity of 92.4%. Those results are a significant improvement from those calculated in Miyamoto for the A1T1 glycoprotein.
The arguments are not persuasive. Examiner contends that since the art teaches the structure of a method for identifying a classification for a biological sample from a patient or individual having a disease or condition and monitoring the health status of the patient, it is the Examiners position that the trained model taught by Kourou would necessarily show an accuracy of 91.9%, a sensitivity of 91.4%, and a specificity of 92.4%.
Applicant contends that the glycosylation sites reported in Miyamoto are different from those in the present application for the above-mentioned glycans. For example, paras. [0066- 0263] disclose multiple sites in which the 5402 glycan attaches to the glycopeptide. While Miyamoto presents that the glycan corresponding to glycan 5402 is attached at site 107, the method disclosed in the present application identified residue 271 as the site in which glycan 5402 attaches to the glycopeptide comprising SEQ ID NO. 4. See para. [0066] of the as-filed Specification.
The arguments are not persuasive. Examiner contends that residue 271 as the site in which glycan 5402 attaches to the glycopeptide comprising SEQ ID NO. 4 is not recited within the instant application claims. Examiner reminds Applicant that the specification is read in-light of the instant application. If Applicant intends for said limitations to be reequipments of the claims, it is recommended that they be incorporated into the instant application claims, as long as there is recitation of the limitations within the instant application specification.
The rejection of claims 38 and 64 under 35 U.S.C. 103 as being unpatentable over Miyamoto et al (2018, J. Proteome Res, cited on IDS filed 1/9/2025) {herein Miyamoto} in view of Kourou et al (2014, Computational and Structural Biotechnology Journal, cited on PTO-892 dated 9/17/2025) {herein Kourou} and in further view of Sandberg et al (WO 2009/075883 A2, Publication Date: 18 June 2009, cited in IDS filed 4/25/2022) {herein Sandberg} is maintained. The rejection has been modified in view of Applicant’s rejection of claim 38 and 64 to recite ‘wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known.’
As amended, claim 38 is drawn to a method for treating a patient having ovarian cancer; the method comprising: obtaining, or having obtained, a biological sample from the patient; digesting and/or fragmenting, or having digested or having fragmented, one or more glycoproteins in the sample wherein the glycoprotein comprises glycan 5402 and at least one of glycans 3200, 3210, 3300, 3310, 3320, 3400, 3410, 3420, 3500, 3510, 3520, 3600, 3610, 3630, 3700, 3710, 3720, 3730, 3740, 4200, 4210, 4300, 4301, 4310, 4311, 4320, 4400, 4401, 4410, 4411, 4420, 4421, 4430, 4431, 4500, 4501, 4510, 4511, 4520, 4521, 4530, 4531, 4540, 4541, 4600, 4601, 4610, 4611, 4620, 4621, 4630, 4631, 4641, 4650, 4700, 4701, 4710, 4711, 4720, 4730, 5200, 5210, 5300, 5301, 5310, 5311, 5320, 5400, 5401, 5402, 5410, 5411, 5412, 5420, 5421, 5430, 5431, 5432, 5500, 5501, 5502, 5510, 5511, 5512, 5520, 5521, 5522, 5530, 5531, 5541, 5600, 5601, 5602, 5610, 5611, 5612, 5620, 5621, 5631, 5650, 5700, 5701, 5702, 5710, 5711, 5712, 5720, 5721, 5730, 5731, 6200, 6210, 6300, 6301, 6310, 6311, 6320, 6400, 6401, 6402, 6410, 6411, 6412, 6420, 6421, 6432, 6500, 6501, 6502, 6503, 6510, 6511, 6512, 6513, 6520, 6521, 6522, 6530, 6531, 6532, 6540, 6541, 6600, 6601, 6602, 6603, 6610, 6611, 6612, 6613, 6620, 6621, 6622, 6623, 6630, 6640, 6641, 6642, 6652, 6700, 6701, 6703, 6710, 6711, 6712, 6713, 6720, 6721, 6730, 6731, 6740, 7200, 7210, 7400, 7410, 7411, 7412, 7420, 7421, 7430, 7431, 7432, 7500, 7501, 7510, 7511, 7512, 7600, 7601, 7602, 7603, 7604, 7610, 7611, 7612, 7613, 7614, 7620, 7621, 7622, 7623, 7640, 7700, 7701, 7702, 7703, 7710, 7711, 7712, 7713, 7714, 7720, 7721, 7722, 7721, 7722, 7730, 7731, 7732, 7740, 7741, 7751, 8200, 9200, 9210, 10200, 11200, and 12200 in Table 4; and detecting and quantifying one or more multiple-reaction-monitoring (MRM) transitions selected from the group consisting of transitions 1-150 using a Triple Quadrupole Mass Spectrometer (QQQ) and/or qTOF mass spectrometer, wherein the method further comprises detecting the glycosylation site residue where the glycan bonds to the glycopeptide, or detecting a glycosylation site on a glycopeptide; inputting the quantification into a trained model to generate an output probability, wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known; determining if the output probability is above or below a threshold for a classification; and classifying the patient based on whether the output probability is above or below a threshold for a classification, wherein the classification is selected from the group consisting of. (A) a patient in need of a chemotherapeutic agent;(B) a patient in need of a immunotherapeutic agent; (C) a patient in need of hormone therapy; (D) a patient in need of a targeted therapeutic agent; (E) a patient in need of surgery; (F) a patient in need of neoadjuvant therapy;(G) a patient in need of chemotherapeutic agent, immunotherapeutic agent, hormone therapy, targeted therapeutic agent, neoadjuvant therapy, or a combination thereof, before surgery;(H) a patient in need of chemotherapeutic agent, immunotherapeutic agent, hormone therapy, targeted therapeutic agent, neoadjuvant therapy, or a combination thereof, after surgery;(I) or a combination thereof; administering a therapeutically effective amount of a therapeutic agent to the patient: wherein the therapeutic agent is selected from chemotherapy if classification A or I is determined; wherein the therapeutic agent is selected from immunotherapy if classification B or I is determined; or wherein the therapeutic agent is selected from hormone therapy if classification C or I is determined; or wherein the therapeutic agent is selected from targeted therapy if classification D or I is determined wherein the therapeutic agent is selected from neoadjuvant therapy if classification F or I is determined; wherein the therapeutic agent is selected from chemotherapeutic agent, immunotherapeutic agent, hormone therapy, targeted therapeutic agent, neoadjuvant therapy, or a combination thereof if classification G or I is determined; and wherein the therapeutic agent is selected from chemotherapeutic agent, immunotherapeutic agent, hormone therapy, targeted therapeutic agent, neoadjuvant therapy, or a combination thereof if classification H or I is determined.
As amended, claim 64 is drawn to a method of treating an individual with ovarian cancer, the method comprising quantifying by mass spectroscopy (MS) one or more glycopeptides in a sample obtained from the individual using MRM-MS with a QQQ and/or qTOF spectrometer wherein the one or more glycopeptides comprise at least one glycopeptide consisting essentially of an amino acid sequence selected from the group consisting of SEQ ID NOs: 1 - 262 and wherein the glycopeptide comprises glycan 5402 and at least one of glycans 3200, 3210, 3300, 3310, 3320, 3400, 3410, 3420, 3500, 3510, 3520, 3600, 3610, 3630, 3700, 3710, 3720, 3730, 3740, 4200, 4210, 4300, 4301, 4310, 4311, 4320, 4400, 4401, 4410, 4411, 4420, 4421, 4430, 4431, 4500, 4501, 4510, 4511, 4520, 4521, 4530, 4531, 4540, 4541, 4600, 4601, 4610, 4611, 4620, 4621, 4630, 4631, 4641, 4650, 4700, 4701, 4710, 4711, 4720, 4730, 5200, 5210, 5300, 5301, 5310, 5311, 5320, 5400, 5401, 5402, 5410, 5411, 5412, 5420, 5421, 5430, 5431, 5432, 5500. 5501, 5502, 5510, 5511, 5512, 5520, 5521, 5522, 5530, 5531, 5541, 5600, 5601, 5602, 5610, 5611, 5612, 5620, 5621, 5631, 5650, 5700, 5701, 5702, 5710, 5711, 5712, 5720, 5721, 5730, 5731, 6200, 6210, 6300, 6301, 6310, 6311, 6320, 6400, 6401, 6402, 6410, 6411, 6412, 6420, 6421, 6432, 6500, 6501, 6502, 6503, 6510, 6511, 6512, 6513, 6520, 6521, 6522, 6530, 6531, 6532, 6540, 6541, 6600, 6601, 6602, 6603, 6610, 6611, 6612, 6613, 6620, 6621, 6622, 6623, 6630, 6640, 6641, 6642, 6652, 6700, 6701, 6703, 6710, 6711, 6712, 6713, 6720, 6721, 6730, 6731, 6740, 7200, 7210, 7400, 7410, 7411, 7412, 7420, 7421, 7430, 7431, 7432, 7500, 7501, 7510, 7511, 7512, 7600, 7601, 7602, 7603, 7604, 7610, 7611, 7612, 7613, 7614, 7620, 7621, 7622, 7623, 7640, 7700, 7701, 7702, 7703, 7710, 7711, 7712, 7713, 7714, 7720, 7721, 7722, 7721, 7722, 7730, 7731, 7732, 7740, 7741, 7751, 8200, 9200, 9210, 10200, 11200, and 12200 in Table 4; wherein the method further comprises detecting the glycosylation site residue where the glycan bonds to the glycopeptide, or detecting a glycosylation site on a glycopeptide; and inputting the quantification into a trained model to generate an output probability, wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known; determining if the output probability is above or below 0.30; and identifying a classification for the sample based on whether the output probability is above or below 0.30, wherein the classification is having ovarian cancer or not having ovarian cancer, and administering an effective amount of a therapeutic agent if the classification indicates that the individual has ovarian cancer.
With respect to claims 38, Miyamoto teaches a multiple reaction monitoring (MRM)- based method for the protein and site-specific quantitation of serum glycoproteins for the identification and classification of ovarian cancer in subjects (abstract). The subjects are preoperative, thereby, absent evidence otherwise, it is the Examiner’s position that they are in-need of surgery (page 223, column 2 , para 1).The method comprises enzymatic digestion/fragmentation and analysis of glycopeptides using MRM-MS on triple quadruploe (QQQ) spectrometers for the accurate quantitation of proteins from biological specimens (page 223, column 1, para 2); page 227, column 2, para 1). Eighteen transitions were monitored by QQQ for protein and protein subclass quantitation (table 2). The glycoprotein A1AT which is comprised of the peptide sequence AVLTIDEK was identified as a glycoprotein within the biological sample from patient with ovarian cancer (table 2). Said peptide sequence is 100% identical to SEQ ID NO: 4 of the instant application claims 38, 64 (appendix A). Said glycoprotein is comprised of glycan 5402, which is H5N4S2and glycan 6503, which is also represented as H6N5S3 (page 223, column 2, para 3). Said glycans are attached to site N107 of the glycoprotein (page 223, column 2, para 3). Absent evidence otherwise, it is the Examiner’s position that said site is the site where the glycan bonds to the glycoprotein in light of the teaching by Miyamoto that the glycan is attached to the N107 site in the glycoprotein (page 223, column 2, para 3). Miyamoto further teaches sites of glycosylation were identified on the glycoprotein A1AT (table 3). Absent evidence otherwise, it is the Examiner’s position that said teaching is detecting a glycosylation site on a glycoprotein. The Global Protein Machine was used to analyze peptide profiles and identify unique tryptic peptides (page 223, column 2, para 3; page 225, column 1, para 1). Data was input into an in-house built software tool GPFinder (page 226, column 2, para 1). Miyamoto further teaches a method wherein statistical analysis is utilized as a predictive model for determining the probability of a subject having ovarian cancer, based on diagnostic biomarkers (page 230, column 2, para 2). A training set consisting of 40 cases and 40 controls was analyzed, and differential analyses were performed to identify aberrant glycopeptide levels (abstract). The p-value for subjects with A1AT glycoprotein is 0.00001 (Table S3). Overall, the study shows the feasibility of monitoring patients at different stages ovarian cancer based on differential glycosylation patterns (page 230, column 1, para 1; column 2, para 3). Examiner is interpreting the different stages to range from stages III to IV as said stages are characterized by different stages of the disease.
However, Miyamoto does not teach inputting the quantification into a trained model to generate an output probability, wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known (claims 38; 64); determining if the output probability is above or below a threshold for a classification; and classifying the patient based on whether the output probability is above or below a threshold for a classification, wherein the classification is selected from the group consisting of. (A) a patient in need of a chemotherapeutic agent;(B) a patient in need of a immunotherapeutic agent; (C) a patient in need of hormone therapy; (D) a patient in need of a targeted therapeutic agent; (E) a patient in need of surgery; (F) a patient in need of neoadjuvant therapy;(G) a patient in need of chemotherapeutic agent, immunotherapeutic agent, hormone therapy, targeted therapeutic agent, neoadjuvant therapy, or a combination thereof, before surgery;(H) a patient in need of chemotherapeutic agent, immunotherapeutic agent, hormone therapy, targeted therapeutic agent, neoadjuvant therapy, or a combination thereof, after surgery;(I) or a combination thereof; administering a therapeutically effective amount of a therapeutic agent to the patient: wherein the therapeutic agent is selected from chemotherapy if classification A or I is determined; wherein the therapeutic agent is selected from immunotherapy if classification B or I is determined; or wherein the therapeutic agent is selected from hormone therapy if classification C or I is determined; or wherein the therapeutic agent is selected from targeted therapy if classification D or I is determined wherein the therapeutic agent is selected from neoadjuvant therapy if classification F or I is determined; wherein the therapeutic agent is selected from chemotherapeutic agent, immunotherapeutic agent, hormone therapy, targeted therapeutic agent, neoadjuvant therapy, or a combination thereof if classification G or I is determined (claim 38). inputting the quantification into a trained model to generate an output probability, wherein the trained model compares the quantification to types, absolute amounts, and relative amounts of glycopeptides in at least one previously classified sample, or to a reference value obtained from a population of individuals whose disease state is known; determining if the output probability is above or below 0.30; and identifying a classification for the sample based on whether the output probability is above or below 0.30, administering an effective amount of a therapeutic agent (claims 38 and 64).
With respect to claims 38, 64, Kourou teaches a method wherein machine learning software is used as a model for studying cancer risk and outcomes in patients (abstract). Kourou further teaches there are two main common types of ML methods known as (i) supervised learning and (ii) unsupervised learning (page 6, column 1, para 6). In supervised learning a labeled set of training data is used to estimate or map the input data to the desired output (page 6, column 1, para 6). ML also includes ANNs (Artificial Neural Networks) that handle a variety of classification or pattern recognition problems (page 10, column 2, para 4). ANNs is trained to generate an output as a combination between the input variables (page 10, column 2, para 4) by detecting specific biomarkers (page 160, column 1, para 2). Absent evidence otherwise, it is the Examiner’s position that biomarkers and glycoproteins are the same as both are comprised of protein that allows for identification of samples. Furthermore, ANNs is the gold standard for classification of cancer samples (page 10, column 2, para 4). The classifiers performance is assessed based on accuracy and are under the curve (page 10, column 1, para 4). The quantitative metrics of accuracy and AUC are used for assessing the overall performance of a classifier (page 10, column 1, para 4; column 2, para 1). Regarding the recitation ‘determining if the output probability is above or below 0.30 for a classification’ within the instant application claim 13, MPEP 2106.04(a)(2)C states ‘A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. There is no particular word or set of words that indicates a claim recites a mathematical calculation. That is, a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.’ As such, the recitation of ‘‘determining if the output probability is above or below 0.30 for a classification’ within the instant application claim 38, 64 is an abstract idea. According to MPEP 2106.04(a), ‘The abstract idea exception has deep roots in the Supreme Court’s jurisprudence. See Bilski v. Kappos, 561 U.S. 593, 601-602, 95 USPQ2d 1001, 1006 (2010) (citing Le Roy v. Tatham, 55 U.S. (14 How.) 156, 174–175 (1853)). Despite this long history, the courts have declined to define abstract ideas. However, it is clear from the body of judicial precedent that software and business methods are not excluded categories of subject matter. For example, the Supreme Court concluded that business methods are not "categorically outside of § 101's scope," stating that "a business method is simply one kind of ‘method’ that is, at least in some circumstances, eligible for patenting under § 101." Bilski, 561 U.S. at 607, 95 USPQ2d at 1008 (2010). See also Content Extraction and Transmission, LLC v. Wells Fargo Bank, 776 F.3d 1343, 1347, 113 USPQ2d 1354, 1357 (Fed. Cir. 2014) ("there is no categorical business-method exception"). Likewise, software is not automatically an abstract idea, even if performance of a software task involves an underlying mathematical calculation or relationship. See, e.g., Thales Visionix, Inc. v. United States, 850 F.3d 1343, 121 USPQ2d 1898, 1902 (Fed. Cir. 2017) ("That a mathematical equation is required to complete the claimed method and system does not doom the claims to abstraction."); McRO, Inc. v. Bandai Namco Games Am. Inc., 837 F.3d 1299, 1316, 120 USPQ2d 1091, 1103 (Fed. Cir. 2016) (methods of automatic lip synchronization and facial expression animation using computer-implemented rules were not directed to an abstract idea); Enfish, 822 F.3d 1327, 1336, 118 USPQ2d 1684, 1689 (Fed. Cir. 2016) (claims to self-referential table for a computer database were not directed to an abstract idea). Regarding the limitation “wherein the therapeutic agent is selected from chemotherapy if classification A or I is determined; wherein the therapeutic agent is selected from immunotherapy if classification B or I is determined; or wherein the therapeutic agent is selected from hormone therapy if classification C or I is determined; or wherein the therapeutic agent is selected from targeted therapy if classification D or I is determined wherein the therapeutic agent is selected from neoadjuvant therapy if classification F or I is determined; wherein the therapeutic agent is selected from chemotherapeutic agent, immunotherapeutic agent, hormone therapy, targeted therapeutic agent, neoadjuvant therapy, or a combination thereof if classification G or I is determined; and wherein the therapeutic agent is selected from chemotherapeutic agent, immunotherapeutic agent, hormone therapy, targeted therapeutic agent, neoadjuvant therapy, or a combination thereof if classification H or I is determined’ (instant application claim 38), this language does not require steps to be performed or limit the claim to a particular structure and does not limit the scope of the claim. See MPEP 2106.C and 2111.04. Instead, the “wherein” clause merely recites a correlation between a the classification of the glycoproteins and its usefulness in selections an appropriate therapeutic agent.
However, Kourou does not teach the method of claims 38 and 64 of administering an effective amount of a therapeutic agent (claims 38 and 64).
With respect to claims 38 and 64, Sandberg teaches a method of identifying glycoprotein possessing a cancer-specific glycoform, as well as diagnostic and therapeutic methods and compositions related to glycoprotein cancer biomarkers (page 2, lines 9-11). Sandberg further teaches treating a subject having cancer or a precancerous condition by administering a therapeutic agent based on the diagnosis (page 29, lines 22-23). Such agents include chemotherapeutic agents (page 31, lines 11-12).
Before the effective filing date of the claimed invention, it would have been obvious to one of ordinary skill in the art to apply the teachings of Miyamoto of a multiple reaction monitoring (MRM)- based method for the protein and site-specific quantitation of serum glycoproteins for the identification and classification of ovarian cancer in subjects (abstract). Or combine the teachings of Kourou and Sandberg because Sandberg teaches a method of identifying glycoproteins possessing a cancer-specific glycoform, as well as diagnostic and therapeutic methods and compositions related to glycoprotein cancer biomarkers (page 2, lines 9-11). Whereas, Kourou teaches a method wherein machine learning software is used as a model for studying cancer risk and outcomes in patients (abstract).
One of ordinary skill in the art would be motivated to either use the teachings of Miyamoto et al. by itself or combine the teachings of Kourou and Sandberg because Sandberg provides the motivation for one of ordinary skill in the art to administer an effective amount of a therapeutic agent (antibody) that specifically binds the cancer-specific glycoform if the classification indicates that the individual has ovarian cancer as said therapeutic agent could be conjugated to a cytotoxic agent, further enhancing toxicity to targeted cells (page 31, lines 6-13). Kourou provides the motivation for Miyamoto to input the quantification of glycoproteins analyzed by MRM-MS and QQQ into a trained model/machine learning model to generate an output probability because it would allow for better classification of patients at risk of and diagnosed with cancer, allow for more appropriate selection of therapeutic agents and better patient outcomes (Kourou: abstract). Furthermore, utilizing a trained model would improve the understanding of cancer progression in subjects (Kourou: abstract). One of ordinary skill in the art knowing the benefit of treating an individual with ovarian cancer based on the teachings of Miyamoto, Kourou and Sandberg would have a reasonable expectation of success that inputting the quantification taught by Miyamoto into the trained model taught by Kourou would result in the classification of whether a subject has cancer. There would be a reasonable expectation of success that utilizing the teaching of Miyamoto to identify the target glycoproteins and the method of Sandberg to target identified glycoforms by conjugating a cytotoxic agent, such as a chemotherapeutic agent, to a glycan-specific antibody to treat the subject (Sandberg: page 31, lines 6-13) would result in an effective method for targeting cancer cells. Said method would allow for an improved understanding of cancer progression in subjects (Kourou: page 12, column 2, para 2) and a better targeting of cancer cells in subjects.
One of skill in the art would have a reasonable expectation of success to make and use the claimed method for identifying a classification for a biological sample from a patient or individual having a disease or condition and monitoring the health status of the patient because Miyamoto provides multiple reaction monitoring (MRM)- based method for the protein and site-specific quantitation of serum glycoproteins for the identification and classification of ovarian cancer in subjects (abstract). Kourou provides the teachings of machine learning software as a model for studying cancer risk and outcomes in patients (abstract). Whereas, Sandberg provides the method of identifying glycoproteins possessing a cancer-specific glycoform, as well as diagnostic and therapeutic methods and compositions related to glycoprotein cancer biomarkers (page 2, lines 9-11). Therefore there would be a reasonable expectation of success to arrive at the above invention. Therefore, the above invention would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention.
RESPONSE TO REMARKS: Examiner contends that Miyamoto in view of Kourou and Sandberg teach the instant application claim of a method for treating a patient having ovarian cancer by utilizing predictive methods of machine learning software to quantitate aberrant biomarkers (glycoproteins) associated with cancers.
Conclusion
Status of claims
Claims 13, 20, 22-24, 26, 28, 31-33, 38, 53, 64 are pending and examined on the merits to the extent they read of species elected SEQ ID NO: 4 and Glycan 5402.
Claims 1-12, 14-19, 21, 25, 27, 29-30, 34-37, 39-52, 54-63 are canceled.
Claims 13, 20, 22-24, 26, 28, 31-33, 38, 53, 64 are rejected.
No claims are in condition for allowance.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ERICA NICOLE JONES-FOSTER whose telephone number is (571)270-0360. The examiner can normally be reached mf 7:30a - 4:30p.
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, Manjunath Rao can be reached at 571-272-0939. 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.
/ERICA NICOLE JONES-FOSTER/ /MANJUNATH N RAO/Examiner, Art Unit 1656 Supervisory Patent Examiner, Art Unit 1656
Appendix A
Miyamoto et al alignment with SEQ ID NO: 4
Query Match 100.0%; Score 37; Length 8;
Best Local Similarity 100.0%;
Matches 8; Conservative 0; Mismatches 0; Indels 0; Gaps 0;
Qy 1 AVLTIDEK 8
||||||||
Db 1 AVLTIDEK 8