DETAILED ACTION
The Applicant’s filing, received 21 April 2026, has been fully considered. This application is a CONTINUATION of Application # 16/826,042, now abandoned. The following rejections and/or objections are either reiterated or newly applied. They constitute the complete set presently being applied to the instant application.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of the Claims
Claims 1-20 are pending.
Claims 1-20 are rejected.
Priority
This application is a CON of 16/826,042, filed 20 March 2020,
which claims benefit of 62/822,730, filed 22 March 2019.
Therefore, the effective filing date of the claimed invention is 22 March 2019.
Drawings
The objections to the drawings for failing to comply with 37 CFR 1.84(p)(5) in the Office action mailed 23 October 2025 have been withdrawn in view of the amendment received 21 April 2026.
The replacement drawings received 21 April 2026 have been accepted.
Specification
The objection to the disclosure in the Office action mailed 23 October 2025 has been withdrawn in view of the amendment received 21 April 2026.
The substitute specification received 21 April 2026 has been entered.
Claim Interpretation
The amendment received 21 April 2026 has been fully considered, however after further consideration, a new grounds of interpretation is raised in view of the amendment.
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation is:
module in claim 1.
Because this claim limitation is being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it is being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
The written description discloses a corresponding structure for the generic placeholder:
module in claim 1, in the specification at paragraph [0058] (e.g., a grouping of features (i.e., data)) and at paragraph [0059] (Table 1); and at Figure 1B (e.g., reference # 152).
If applicant does not intend to have this limitation interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Objections
The objections to claims 1, 15, 16, and 17 in the Office action mailed 23 October 2025 have been withdrawn in view of the amendment received 21 April 2026.
Claim Rejections - 35 USC § 112
The rejection of claims 1-20 under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, in the Office action mailed 23 October 2025 has been withdrawn in view of the amendment received 21 April 2026.
Claim Rejections - 35 USC § 101
The amendment received 21 April 2026 has been fully considered, however after further consideration, the rejection of claims 1-20 under 35 U.S.C. 101 in the Office action mailed 23 October 2025 has been maintained with modification in view of the amendment, as noted below.
The eligibility analysis of claims 1-20 has been modified to incorporate the amended claim limitations.
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claims recite both: (a) mathematical concepts, (e.g., mathematical relationships, formulas or equations, mathematical calculations); and (b) mental processes, i.e., concepts performed in the human mind, (e.g., observation, evaluation, judgement, opinion).
Claim Interpretations
Claim 1 recites the limitation “processing a test sample, comprising blood from the test patient, to generate a test dataset.” This limitation is interpreted to require one or more active steps of determining the molecular information within the test sample, and encompassing steps of performing genomic sequencing to obtain the sequence reads from the nucleic acid obtained from the patient’s biological sample (e.g., see Specification, paras. [0076] & [0077]).
Claim 1 further recites the limitation “processing the test dataset with a main classifier structured as a feed-forward neural network comprising an input layer and an output layer, wherein the main classifier is configured to (i) receive a plurality of feature values, each feature value corresponding to a respective module comprising a subset of features comprising genes whose expression levels correspond to a phenotype associated with the condition, and (ii) generate probabilities corresponding to a plurality of classes of the condition, and wherein the main classifier is generated from a heterogenous repository of input data, the heterogenous repository comprising a set of datasets comprising: a first dataset comprising…, and a second dataset acquired…corresponding to the status of the condition.” This limitation is interpreted to recite a product-by-process limitation with the product being the main classifier structured as a feed-forward neural network comprising an input layer and an output layer, and further interpreted to not require the process of performing the active steps of producing the product (e.g., accessing a heterogeneous repository of input data, and training the main classifier).
Response to Arguments
The Applicant’s arguments/remarks received 21 April 2026 have been fully considered, but are not persuasive.
The Applicant states on pages 10-11 of the Remarks that the Examiner’s interpretation of the limitation “processing a test sample…to generate a test dataset” as requiring an active genomic sequencing step is inconsistent with the plain language of the claim and the specification, and therefore should not be limited to any particular laboratory technique, and moreover, Figure 1A illustrates that the system operates on stored “biological sample sequence data” and “data constructs” associated with samples, indicating that the claimed “processing” step encompasses handling and transforming existing data constructs, not necessarily generating raw sequence reads de novo.
These arguments are persuasive in part, to the extent that the claim interpretation has been modified to not be limited to any particular laboratory technique, and not persuasive in part, to the extent that it is noted that the claim limitation actually recites “processing a test sample, comprising blood from the test patient, to generate a test dataset” (emphasis added) and therefore the claim is interpreted to require an active step of determining the molecular information within the test sample.
The Applicant states on page 11 of the Remarks that the Examiner’s characterization of the “main classifier…generated from a heterogeneous repository” limitation as a product-by-process limitation is likewise improper. The Applicant further states that the claim does not merely recite a product defined by how it is made; rather, it affirmatively requires specific structural and functional characteristics of the classifier, including that it is “structured as a feed-forward neural network comprising an input layer and an output layer” and that it is “generated from a heterogeneous repository of input data” comprising multiple independently acquired datasets, and that these limitations define the architecture and data provenance of the classifier itself, not merely the steps used to create it.
The arguments are not persuasive, to the extent that claim 1 does not actually recite steps of generating, i.e., producing, the main classifier, thus necessitating the claim interpretation provided above.
Subject matter eligibility evaluation in accordance with MPEP 2106.
Eligibility Step 1: Step 1 of the eligibility analysis asks: Is the claim to a process, machine, manufacture or composition of matter?
Claims 1-20 are directed to a method (i.e., process) for evaluating a condition of a test patient.
Therefore, these claims are encompassed by the categories of statutory subject matter, and thus satisfy the subject matter eligibility requirements under step 1.
[Step 1: YES]
Eligibility Step 2A: First it is determined in Prong One whether a claim recites a judicial exception, and if so, then it is determined in Prong Two whether the recited judicial exception is integrated into a practical application of that exception.
Eligibility Step 2A Prong One: In determining whether a claim is directed to a judicial exception, examination is performed that analyzes whether the claim recites a judicial exception, i.e., whether a law of nature, natural phenomenon, or abstract idea is set forth or described in the claim.
Independent claim 1 recites the following steps which fall within the mental processes and/or mathematical concepts groupings of abstract ideas:
processing the test dataset with a main classifier structured as a feed-forward neural network comprising an input layer and an output layer (i.e., mathematical concepts), wherein the main classifier is configured to
(i) receive a plurality of feature values, each feature value corresponding to a respective module comprising a subset of features comprising genes whose expression levels correspond to a phenotype associated with the condition, and
(ii) generate probabilities corresponding to a plurality of classes of the condition, and wherein the main classifier is generated from a heterogenous repository of input data, the heterogenous repository comprising a set of datasets comprising:
a first dataset comprising a first plurality of features comprising values corresponding to a status of the condition, and
a second dataset acquired independently of the first dataset and comprising a second plurality of features different from the first plurality of features and comprising values corresponding to the status of the condition;
wherein the first plurality of features and the second plurality of features are acquired through different technical backgrounds and are mapped to a common set of variables and co-normalized across datasets prior to training the main classifier;
generating an output, comprising a characterization of a set of expression values of a set of features, from the output layer (i.e., mathematical concepts); and
selecting a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier (i.e., mental processes).
Dependent claims 2-10, 13, 15, and 17-20 recite the following steps which fall within the mental processes and/or mathematical concepts groupings of abstract ideas, as noted below.
Dependent claim 2 further recites:
wherein the condition comprises an infection or sterile inflammation (i.e., mental processes).
Dependent claim 3 further recites:
wherein the input layer of the main classifier is configured to receive the plurality of feature values from each of a set of feeder neural network input layers (i.e., mathematical concepts).
Dependent claim 4 further recites:
wherein the set of feeder neural network input layers comprises a first feeder input layer for processing mRNA abundance values for a first set of genes represented in the test dataset, and a second feeder input layer for processing mRNA abundance values for a second set of genes represented in the test dataset (i.e., mathematical concepts).
Dependent claim 5 further recites:
wherein the set of datasets comprises at least three datasets corresponding to three cohorts of patients, wherein each of the at least three datasets is generated using a different measurement technique (i.e., mental processes).
Dependent claim 6 further recites:
generating and training the main classifier, wherein the set of datasets used to train the main classifier comprises non-overlapping, independent datasets (i.e., mathematical concepts).
Dependent claim 7 further recites:
(C) co-normalizing values for features present in the first dataset and the second dataset to remove an inter-dataset batch effect, wherein co-normalizing comprises implementing a co-normalization function that requires healthy controls for derivation of correction factors, thereby calculating, for each respective training subject in the first plurality of training subjects and for each respective training subject in the second plurality of training subjects, a set of co-normalized feature values (i.e., mental processes and mathematical concepts); and
(D) training the main classifier against a composite training set, (i.e., mathematical concepts), the composite training set comprising, for each respective training subject in the first plurality of training subjects and for each respective training subject in the second plurality of training subjects:
(i) a summarization of the set of co-normalized feature values (i.e., mental processes), and
(ii) an indication of the absence, presence or stage of the condition (i.e., mental processes).
Dependent claim 8 further recites:
wherein co-normalizing feature values comprises determining an expected expression value of each of the first plurality of features and the second plurality of features and adjusting the expected expression value of each of the first plurality of features and the second plurality of features for modifications of mean and standard deviation attributed to execution of the first measurement technique and the second measurement technique (i.e., mental processes and mathematical concepts).
Dependent claim 9 further recites:
wherein co-normalizing feature values is performed iteratively, with acceptance of a third dataset generated using a third measurement technique, for training the main classifier (i.e., mental processes and mathematical concepts).
Dependent claim 10 further recites:
wherein the inter-dataset batch effect includes an additive component and a multiplicative component and wherein co-normalizing shrinks resulting parameters representing the additive component and a multiplicative component (i.e., mental processes and mathematical concepts).
Dependent claim 13 further recites:
wherein values of the set of features comprises expression values of a first set of genes comprising: IFI27, JUP, and LAX1 (i.e., mental processes).
Dependent claim 15 further recites:
wherein values of the set of features comprises expression values of a first set of genes comprising: CEACAM1, ZDHHC19, C9orf95, GNA15, BATF, and C3AR1 (i.e., mental processes).
Dependent claim 17 further recites:
wherein values of the set of features comprises expression values of a first set of genes comprising: DEFA4, CD163, RGS1, PER1, HIF1A, SEPP1, C11orf74 and CIT (i.e., mental processes).
Dependent claim 18 further recites:
classifying the test patient as at risk for death within 30 days of hospital admission (i.e., mental processes).
Dependent claim 19 further recites:
wherein the first plurality of features comprises nucleic acid expression features, and wherein the second plurality of features comprises protein expression features (i.e., mental processes).
Dependent claim 20 further recites:
wherein the status of the condition is a diseased condition, and wherein the first dataset comprises data from a first subportion of subjects that are free of the diseased condition, and a second subportion of subjects that exhibit the diseased condition (i.e., mental processes).
The abstract ideas recited in the claims are evaluated under the broadest reasonable interpretation (BRI) of the claim limitations when read in light of and consistent with the specification. As noted in the foregoing section, the claims are determined to contain limitations that can practically be performed in the human mind with the aid of a pen and paper (e.g., selecting a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier, is a limitation that comprises a mental process with regard to data that may involve observation, evaluation, judgement, and opinion), and therefore recite judicial exceptions from the mental process grouping of abstract ideas. Additionally, the recited limitations that are identified as judicial exceptions from the mathematical concepts grouping of abstract ideas (e.g., co-normalizing values in different datasets to remove an inter-dataset batch effect involves performing mathematical calculations) are abstract ideas irrespective of whether or not the limitations are practical to perform in the human mind.
Therefore, claims 1-20 recite an abstract idea.
[Step 2A Prong One: YES]
Eligibility Step 2A Prong Two: In determining whether a claim is directed to a judicial exception, further examination is performed that analyzes if the claim recites additional elements that when examined as a whole integrates the judicial exception(s) into a practical application (MPEP 2106.04(d)). A claim that integrates a judicial exception into a practical application will apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception. The claimed additional elements are analyzed to determine if the abstract idea is integrated into a practical application (MPEP 2106.04(d)(I); MPEP 2106.05(a-h)). If the claim contains no additional elements beyond the abstract idea, the claim fails to integrate the abstract idea into a practical application (MPEP 2106.04(d)(III)).
The judicial exceptions identified in Eligibility Step 2A Prong One are not integrated into a practical application because of the reasons noted below.
Dependent claims 2-6, 8-10, 13, 15, and 17-20 do not recite any elements in addition to the judicial exception, and thus are part of the judicial exception.
The additional elements in independent claim 1 include:
processing a test sample, comprising blood from the test patient, to generate a test dataset (i.e., gathering data for use in the claimed process); and
based upon the output, treating the test patient by administering a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier.
The additional elements in dependent claims 7, 11, 12, 14, and 16 include:
(A) for each respective training subject in a first plurality of training subjects: acquiring values of the first plurality of features represented in the first dataset by processing a sample of the respective training subject with a first measurement technique, wherein the first dataset comprises, for each respective training subject in the first plurality of training subjects an indication of the absence, presence or stage of the condition (i.e., acquiring data for use in the claimed process) (claim 7);
(B) for each respective training subject in a second plurality of training subjects independent of the first plurality of training subjects: acquiring values of the second plurality of features represented in the second dataset by processing a sample of the respective training subject with a second measurement technique different from the first measurement technique, wherein the second dataset comprises, for each respective training subject in the second plurality of training subjects an indication of the absence, presence or stage of the condition (i.e., acquiring data for use in the claimed process) (claim 7);
generating the first dataset using a first measurement technique, and generating the second dataset using a second measurement technique (i.e., further defining the step of gathering data for use in the claimed process) (claim 11);
the first measurement technique comprises RNAseq for a first cohort, and wherein the second measurement technique comprises using DNA microarrays for a second cohort different from the first cohort (i.e., further defining the step of gathering data for use in the claimed process) (claim 12);
treating the test patient comprises treating the test patient for a viral infection (claim 14); and
treating the test patient comprises treating the test patient for sepsis (claim 16).
The additional elements of processing a test sample, comprising blood from the test patient, to generate a test dataset (claim 1); generating the first dataset using a first measurement technique, and generating the second dataset using a second measurement technique (claim 11); and the first measurement technique comprises RNAseq for a first cohort, and wherein the second measurement technique comprises using DNA microarrays for a second cohort different from the first cohort (claim 12); are steps used in the process of gathering data for use in the claimed process, and therefore do not add more than insignificant extra-solution activities to the judicial exceptions (MPEP 2106.05(g)).
The additional elements in claim 7 at step (A) for each respective training subject in a first plurality of training subjects: acquiring values of the first plurality of features represented in the first dataset, upon processing a sample of the respective training subject with a first measurement technique, wherein the first dataset comprises, for each respective training subject in the first plurality of training subjects an indication of the absence, presence or stage of the condition; and at step (B) for each respective training subject in a second plurality of training subjects independent of the first plurality of training subjects: acquiring values of the second plurality of features represented in the second dataset upon processing a sample of the respective training subject with a second measurement technique different from the first measurement technique, wherein the second dataset comprises, for each respective training subject in the second plurality of training subjects an indication of the absence, presence or stage of the condition; are steps in the process of gathering data for use in the claimed process, and therefore do not add more than insignificant extra-solution activities to the judicial exceptions (MPEP 2106.05(g)).
The additional elements of based upon the output, treating the test patient by administering a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier (claim 1); treating the test patient comprises treating the test patient for a viral infection (claim 14); and treating the test patient comprises treating the test patient for sepsis (claim 16); do not amount to more than a recitation of the words “apply it” and therefore do not amount to more than mere instructions to implement an abstract idea, because the claims fail to recite details of how a solution (i.e., the treatment) to a problem is accomplished, and also because the generality of the application of the judicial exception does not provide more than a broad applicability to the treatment step (MPEP 2106.05(f)).
Furthermore, the additional elements of based upon the output, treating the test patient by administering a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier (claim 1); treating the test patient comprises treating the test patient for a viral infection (claim 14); and treating the test patient comprises treating the test patient for sepsis (claim 16); do not affirmatively recite an action that effects a particular treatment or prophylaxis for a disease or medical condition, because the treatment step must be “particular,” i.e., specifically identified (MPEP 2106.05(d)(2)).
Further still, the additional element of based upon the output, treating the test patient by administering a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier (claim 1) is a contingent limitation (i.e., non-obligatory) that comprises an embodiment of the claim that does not require a step of performing an action that effects a particular treatment or prophylaxis for a disease or medical condition, because the predicted class of the condition indicated by the output of the main classifier could indicate the absence of the condition in the test patient.
Thus, the additionally recited elements merely amount to insignificant extra-solution activity, and/or do not amount to more than mere instructions to apply an exception, and/or do not affirmatively recite a particular treatment, and as such, when all limitations in claims 1-20 have been considered as a whole (i.e., the analysis takes into consideration all the claim limitations and how those limitations interact and impact each other when evaluating whether the exception is integrated into a practical application), the claims are deemed to not recite any additional elements that would integrate a judicial exception into a practical application, and therefore claims 1-20 are directed to an abstract idea (MPEP 2106.04(d)).
[Step 2A Prong Two: NO]
Eligibility Step 2B: Because the claims recite an abstract idea, and do not integrate that abstract idea into a practical application, the claims are probed for a specific inventive concept. The judicial exception alone cannot provide that inventive concept or practical application (MPEP 2106.05). Identifying whether the additional elements beyond the abstract idea amount to such an inventive concept requires considering the additional elements individually and in combination to determine if they amount to significantly more than the judicial exception (MPEP 2106.05A i-vi).
The claims do not include any additional elements that are sufficient to amount to significantly more than the judicial exception(s) because of the reasons noted below.
Dependent claims 2-6, 8-10, 13, 15, and 17-20 do not recite any elements in addition to the judicial exception(s).
The additional elements recited in independent claim 1 and dependent claims 7, 11, 12, 14, and 16 are identified above, and carried over from Step 2A: Prong Two along with their conclusions for analysis at Step 2B. Any additional element or combination of elements that was considered to be insignificant extra-solution activity at Step 2A: Prong Two was re-evaluated at Step 2B, because if such re-evaluation finds that the element is unconventional or otherwise more than what is well-understood, routine, conventional activity in the field, this finding may indicate that the additional element is no longer considered to be insignificant; and all additional elements and combination of elements were evaluated to determine whether any additional elements or combination of elements are other than what is well-understood, routine, conventional activity in the field, or simply append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception, per MPEP 2106.05(d).
The additional elements of using a main classifier structured as a feed-forward neural network comprising an input layer and an output layer (claim 1) (i.e., computers and/or computer components); and acquiring data (claim 7); are conventional computer components and/or functions (see MPEP at 2106.05(b) and 2106.05(d)(II) regarding conventionality of computer components and computer processes).
The additional elements of processing a test sample, comprising blood from the test patient, to generate a test dataset (claim 1); generating the first dataset using a first measurement technique, and generating the second dataset using a second measurement technique (claim 11); and wherein the first measurement technique comprises RNAseq for a first cohort, and wherein the second measurement technique comprises using DNA microarrays for a second cohort different from the first cohort (claim 12); are conventional. Evidence for the conventionality is shown by:
Ong et al. (Journal of Developmental Origins of Health and Disease, 2015, Vol. 6(1), pp. 10-16, as cited in the Office action mailed 23 October 2025).
Ong et al. reviews computational and statistical methods in DNA methylation analyses and interpretation, and the challenges of gene-environment interaction analyses, and multiple genome-wide molecular data integration (Title; Abstract; and page 10, col. 2, para. 2). Ong et al. shows using Infinium450K data from blood of individuals of different ages (page 12, col. 2, para. 3); and further shows integration of multiple genome-wide molecular data sets allowing one to explore the causal relationships between the different layers of biological control, e.g., exploring the causal relationship between genotype, DNA methylation and gene expression by combining data from RNA-seq, SNP genotyping and the Infinium450K array performed on umbilical cords of newborn infants (page 13, col. 1, paras. 2-3).
The additional elements of treating the test patient by administering a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier (claim 1); treating the test patient comprises treating the test patient for a viral infection (claim 14) and treating the test patient comprises treating the test patient for sepsis (claim 16); are conventional. Evidence for the conventionality is shown by:
Plunkett et al. (BMJ, 2015, Vol. 350:h3017, pp. 1-12, as cited in the Information Disclosure Statement (IDS) received 09 January 2024, and as cited in the Office action mailed 23 October 2025).
Plunkett et al. reviews sepsis in children (Title; and Abstract) and shows that there is a growing interest in the use of biomarkers for the diagnosis and monitoring of sepsis and septic shock, and further shows using classification and regression tree analysis to identify a series of biomarkers in order to risk stratify patients for the purposes of clinical decision making (page 5, col. 1, para. 7). Plunkett et al. further discusses antibiotic therapy (page 8, col. 1, paras. 8-9 through col. 2, paras. 1-7; and Box 6), and antiviral therapy (page 8, col. 2, para. 10 through page 9, col. 1, para. 1).
Therefore, when taken alone (i.e., individually), all additional elements in claims 1-20 do not amount to significantly more than the above-identified judicial exception(s). Even when evaluated as an ordered combination, the additional elements fail to transform the exception(s) into a patent-eligible application of that exception. Thus, claims 1-20 are deemed to not contribute an inventive concept, i.e., amount to significantly more than the judicial exception(s) (MPEP 2106.05(II)).
[Step 2B: NO]
Response to Arguments
The Applicant’s arguments/remarks received 21 April 2026 have been fully considered, but are not persuasive.
The Applicant states on page 12 of the Remarks that the Applicant disagrees with the rejection and the characterizations made in the Office action (mailed 23 October 2025). The Applicant further states on page 13 (top) that the claims do not recite a judicial exception, and are not directed to an abstract idea, either as identified in the 2019 Revised Patent Subject Matter Eligibility Guidance (“2019 Revised PEG”), the July 2024 Subject Matter Eligibility Examples (“July 2024 SME”) and as further clarified in the USPTO’s August 2025 Memorandum on Subject Matter Eligibility (“August 2025 Memo”) and in view of the PTAB’s precedential reasoning in Ex parte Desjardins, (“Desjardins”). The Applicant further states on page 13 (bottom) that the Office asserts that the claims can be characterized as reciting mental process and/or mathematical concepts based on a high-level description of generating a characterization, processing the test data, and generating an output, however the Applicant disagrees, and submits that the August 2025 Memo expressly cautions examiners against characterizing claims as mental processes where the recited operations cannot practically be performed in the human mind, even if the operations involve data analysis. The Applicant further states that the Memo emphasizes that the correct inquiry is not whether a step is conceptually analyzable by a human, but whether it is practically performable without computer implementation. The Applicant further states on page 14 (top) that when properly considered, claim 1 recites a specific machine-learning-based processing pipeline operating on biologically-derived data generated from a test sample, including limitations that require computational operations involving high-dimensional biological data, neural network architectures, and cross-dataset integration that cannot practically be performed in the human mind or with pen and paper. The Applicant further states that a human cannot mentally implement or simulate a feed-forward neural network trained on heterogeneous biological datasets, nor can a human perform the claimed processing of expression values and generation of classifier outputs in a meaningful or reproducible manner. The Applicant further states on page 14 (middle) that accordingly, the claimed operations cannot practically be performed in the human mind, either mentally or manually, and therefore fall outside the mental processes grouping identified in the 2019 Revised PEG.
These arguments are not persuasive, because first, the August 4, 2025 memorandum regarding reminders on evaluating subject matter eligibility of claims under 35 U.S.C. 101 clearly states (page 1) that the memorandum is not intended to announce any new USPTO practice or procedure and is meant to be consistent with existing USPTO guidance, and further states that Examiners should consult the specific MPEP sections for more thorough information. Second, the rejection above and the rejections of record have been raised in accordance with the eligibility analysis framework provided at MPEP 2106, which incorporates the 2019 Revised PEG, and the Advanced Notice of Change (“ANC”) revising the November 2024 publication of the MPEP to include Ex parte Desjardins explicitly states that it is not intended to announce any new USPTO practice or procedure and is meant to be consistent with the existing guidance, and the examiners are expected to consider the ANC, particularly when evaluating subject matter eligibility of claims related to machine learning or AI. Third, at Eligibility Step 2A Prong One, examiners evaluate whether the claim recites a judicial exception, i.e., whether a law of nature, natural phenomenon, or abstract idea is set forth or described in the claim, and as noted in the above rejection, amended claim 1 recites at least the mental process of selecting a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier. Fourth, the Applicant’s argument that the claims recite limitations that cannot be practically performed in the human mind includes limitations that are not identified as mental processes in the above rejection, e.g., simulating a feed-forward neural network.
The Applicant states on page 14 (bottom) of the Remarks that the Office also characterizes claim 1 as reciting a mathematical operation, and that the Office’s analysis broadly characterizes the limitations reciting processing the test data and also generating an output as mathematical concepts, but the claim does not set forth any mathematical formula, equation, or calculation. The Applicant further states that instead, the claim recites concrete operations, and as explained in the August 2025 Memorandum, limitations that merely involve data processing or machine learning do not recite a mathematical concept unless a specific mathematical relationship or calculation is explicitly set forth. The Applicant further states on page 15 (top) that here, the claim does not recite any such mathematical relationships, but instead recites a specific technological implementation of a classifier operating on biological data, and accordingly, independent claim 1 does not recite a judicial exception and is therefore not directed to an abstract idea under Prong One.
These arguments are not persuasive, because first, as noted in the foregoing response to arguments, the August 4, 2025 memorandum regarding reminders on evaluating subject matter eligibility of claims under 35 U.S.C. 101 clearly states (page 1) that the memorandum is not intended to announce any new USPTO practice or procedure and is meant to be consistent with existing USPTO guidance, and further states that Examiners should consult the specific MPEP sections for more thorough information. Second, the rejection above and the rejections of record have been raised in accordance with the eligibility analysis framework provided at MPEP 2106, which incorporates the 2019 Revised PEG. Third, regarding distinguishing claims that recite a judicial exception from claims that merely involve a judicial exception at Step 2A Prong One, the August 2025 Memo asks the examiner to consider for example, the published USPTO examples 39 and 47. Those Examples, and others (e.g., Examples 48 and 49 in the July 2024 SME) have been considered in the eligibility analysis of the instant claims in the above rejection. However, it is noted that the Eligibility Examples are hypothetical and only intended to be illustrative of the claim analysis performed using MPEP 2106, and of the particular issues noted in the Example, and therefore, the Examples should be interpreted based on the fact patterns set forth in a particular Example, as other fact patterns may have different eligibility outcomes, as evidenced in the rejection of the instant claims above. Fourth, regarding whether or not a claim limitation recites a judicial exception, e.g., a mathematical concept, it is noted that the MPEP at 2106.04 II.A.1. is instructive, describing an example of a claim that merely involves, or is based on, an exception is a claim to "A teeter-totter comprising an elongated member pivotably attached to a base member, having seats and handles attached at opposing sides of the elongated member." This claim is based on the concept of a lever pivoting on a fulcrum, which involves the natural principles of mechanical advantage and the law of the lever. However, this claim does not recite (i.e., set forth or describe) these natural principles and therefore is not directed to a judicial exception. In contrast, a claim that recites (i.e., sets forth or describes) processing a dataset with a classifier structured as a feed-forward neural network to generate an output comprising a characterization of a set of expression values of a set of features, recites a mathematical concept.
The Applicant states on page 15 (middle) that at Prong Two, the claims integrate the judicial exception into a practical application. The Applicant summarizes aspects of the eligibility analysis at MPEP 2106.04(d)(1), and further states that the Federal Circuit and the MPEP agree that one must rely on the specification for determining whether the claims are directed to an improvement in computer technology, and that specifically, the Federal Circuit has expressly found that a claimed invention was not directed to an abstract idea because it was directed to an improvement to an existing technology as bolstered by the specification’s teaching that the claimed invention achieves other benefits over conventional technologies. The Applicant further states on page 16 (top) that the August 2025 Memo reiterates this point with particular clarity, instructing examiners that the additional limitations should not be evaluated in a vacuum, completely separate from the recited judicial exception, and instead, the analysis must be conducted by considering the claim as a whole, including how the additional elements use or interact with the exception to achieve a meaningful technical result. The Applicant further states on page 16 (middle) that here, the Office concludes that the additional elements do not integrate the alleged abstract idea into a practical application, characterizing the recited steps as merely data gathering or instructions to implement an abstract idea on a generic computer, and further states that this analysis improperly ignores the specific technological framework recited in the claims and the corresponding improvements described in the specification. The Applicant further states on page 16 (bottom) that as an initial matter, the specification at paragraphs [0003] – [0006] identifies a specific technical problem in the field of bioinformatics and machine learning, namely the difficulty of training and applying classifiers using heterogeneous biological datasets generated from different measurement techniques, which exhibit differing feature spaces, distributions, and batch effects. The Applicant further states that the specification explains that conventional approaches cannot directly combine such datasets due to inter-dataset variability and platform specific bias and therefore fail to fully leverage available biological data. The Applicant further submits that the claimed embodiments provide a technical solution by generating the classifier from a heterogeneous repository of input data comprising datasets with different feature sets, and by integrating these datasets through mapping and co-normalization to remove inter-dataset batch effects prior to training and use of the classifier. The Applicant points to the specification at paragraphs [0007], [0008], [0031] – [0033], and [0217] – [0222], and further states on page 17 (top) that these operations are not generic data processing, but instead constitute a specific computational framework that enables the system to operate on biologically heterogeneous data in a manner that was not previously feasible, and as a result, the invention improves the functioning of machine learning systems in bioinformatics by enabling robust cross-platform training and more reliable classification of patient conditions based on integrated biological datasets. The Applicant further states on page 17 (middle) that the claims themselves reflect these improvements through specific limitations, and that these limitations require the system to construct and operate a neural network classifier using heterogeneous datasets with different feature spaces, which is a non-trivial technical task that requires specialized data integration and preprocessing, and further states that these limitations collectively define a particular way of processing biological data and training a classifier, rather than merely reciting the result of classification. The Applicant further states that the additional step of based on the output, treating the test patient, further ties the claimed processing to a concrete real-world application in clinical decision-making.
These arguments are not persuasive, because first, a claim reciting a judicial exception is not directed to the judicial exception if it also recites additional elements demonstrating that the claim as a whole integrates the exception into a practical application, and one way to demonstrate such integration is when the claimed invention improves the functioning of a computer or improves another technology or technical field. With regard to evaluating improvements in the functioning of a computer, or an improvement to any other technology or technical field at Step 2A Prong Two, the courts have not provided an explicit test for this consideration, but have instead illustrated how it is evaluated in numerous decisions. In short, first the specification should be evaluated to determine if the disclosure provides sufficient details such that one of ordinary skill in the art would recognize the claimed invention as providing an improvement, and second, if the specification sets forth an improvement in technology, the claim must be evaluated to ensure that the claim itself reflects the disclosed improvement, i.e., the claim includes the components or steps of the invention that provide the improvement described in the specification. However, as noted in the above rejection at Step 2A Prong Two, when all limitations in claims 1-20 have been considered as a whole (i.e., the analysis takes into consideration all the claim limitations and how those limitations interact and impact each other when evaluating whether the exception is integrated into a practical application), the claims are deemed to not recite any additional elements that would integrate a judicial exception into a practical application, and therefore claims 1-20 are directed to an abstract idea, and not an improvement to computer functionality itself, or an improvement to another technology or technical field. Second, regarding the Applicant’s argument that the specification identifies a specific technical problem in the field of bioinformatics and machine learning, namely the difficulty of training and applying classifiers using heterogeneous biological datasets generated from different measurement techniques, which exhibit differing feature spaces, distributions, and batch effects, and the Applicant’s argument that the claims themselves reflect these improvements through specific limitations that recite a particular way of processing biological data and training a classifier, it is noted in response that the claimed limitations being argued as reflecting the improvement are limitations that comprise the judicial exceptions identified at Step 2A Prong One (e.g., processing data and training a classifier are judicial exceptions), and therefore, the eligibility analysis in the above rejection has determined that the instant claimed improvement to the functioning of machine learning systems in bioinformatics by enabling robust cross-platform training and more reliable classification of patient conditions based on integrated biological datasets, is a purported improvement to the abstract idea (data analysis), and not an improvement to computer functionality itself, or an improvement to another technology or technical field. Third, regarding the Applicant’s argument that the additional step of treating the test patient based on the output further ties the claimed processing to a concrete real-world application in clinical decision-making, it is noted that the eligibility analysis in the above rejection has determined that this limitation does not recite a particular treatment or prophylaxis for a disease or medical condition.
The Applicant states on page 17 (bottom) and page 18 (top) that consistent with Desjardins, claims that recite specific improvements to how a machine learning model operates or is trained are not directed to an abstract idea when the claims reflect those improvements, and that in Desjardins, the Appeals Review Panel (“ARP”) found eligibility where the claims improved how a model learned across tasks and preserved prior knowledge, emphasizing that examiners must not evaluate claims at a high level of generality or dismiss meaningful technical limitations. The Applicant further states that similarly, the claimed embodiments improve how a classifier is generated and applied using heterogeneous biological datasets, enabling integration of data from different measurement platforms and improving the robustness and applicability of the resulting model, and further states that the Office’s position that the neural network and heterogeneous dataset limitations are merely instructions to implement an abstract idea fails to account for this specific technological improvement and instead improperly reduces the claims to a generalized concept. The Applicant further states on page 18 (bottom) that the Office also asserts that the recited steps of acquiring and processing biological samples constitute insignificant extra-solution activity, however, the claim requires processing a test sample… to generate a test dataset comprising biological expression data, which forms the basis for the subsequent machine learning process. The Applicant further states that this is not incidental activity, but rather a necessary part of the claimed technological process that enables the classifier to operate on real-world biological inputs, and moreover, the combination of sample processing, heterogeneous dataset integration, neural network-based classification, and treatment forms an ordered sequence of steps that collectively implement a specific technological solution, and when considered as a whole, these limitations impose meaningful constraints on how the alleged abstract idea is applied and therefore integrate any such idea into a practical application. The Applicant further states on page 19 (top) that accordingly, the claims are directed to a specific technological solution in the field of bioinformatics and machine learning, and not to an abstract idea.
These arguments are not persuasive, because, with regard to the Applicant’s attempt to analogize the instant claims to the claims in Desjardins, these arguments are not persuasive at least because the fact patterns differ between the claims at issue in Desjardins and the instant claims, not least in that the “ARP” in Desjardins notes that the Federal Circuit held that the eligibility determination should turn on whether “the claims are directed to an improvement to computer functionality versus being directed to an abstract idea” (citing Enfish), prior to the “ARP” finding that the improvement to how the machine learning model operates allows artificial intelligence (AI) systems to use less of their storage capacity and enables reduced system complexity, as supported by the Specification – i.e., the improvement to the model provided an improvement to computer functionality itself. This fact pattern contrasts with the fact pattern recited in the instant claims, which broadly recite steps of training and operating a neural network classifier on heterogeneous data, and therefore the eligibility analysis in the above rejection has determined that the claimed improvement to the functioning of machine learning systems in bioinformatics by enabling robust cross-platform training and more reliable classification of patient conditions based on integrated biological datasets, is a purported improvement to the abstract idea (data analysis), and not an improvement to computer functionality itself, or an improvement to another technology or technical field. Therefore, claims should be interpreted based on the fact patterns set forth in a particular claim, as other claims with different fact patterns may have different eligibility outcomes, as evidenced in the rejection of the instant claims above.
The Applicant states on page 19 (middle) that even if the Office were to conclude that claim 1 is directed to an abstract idea, the claim nonetheless recites significantly more than any alleged exception under Step 2B. The Applicant further states that under MPEP 2106.05, the relevant inquiry is whether the additional claim elements, individually and as an ordered combination, amount to well-understood, routine, and conventional activity. The Applicant further states that the Applicant disagrees with the Office’s conclusion that the additional elements are generic and merely implement an abstract idea using a generic computer and conventional data gathering techniques, and further states that the present claims do not merely collect and analyze data, but instead, the claims recite a specific ordered combination of operations that changes how a machine learning system processes heterogeneous biological data, and more particularly, the collective limitations require integration of datasets having different feature spaces and the use of a neural network classifier trained on such heterogeneous data, which is a specific technical arrangement rather than a generic data analysis step. The Applicant further states on page 19 (bottom) and page 20 (top) that the Office asserts that the recited computer components, such as a neural network classifier, are generic and therefore insufficient, however, the proper Step 2B inquiry is not whether individual components are known, but whether the claimed arrangement and interaction of those components is conventional. The Applicant further states that even if neural networks and biological data processing were known in isolation, the Office has not provided factual support demonstrating that it was well-understood, routine, or conventional to generate and use a classifier trained from a heterogeneous repository comprising datasets with different feature pluralities and acquired using different measurement techniques, not to integrate such datasets through mapping and co-normalization prior to training. The Applicant point to the specification at paragraph [0004], and further states on page 20 (middle) that additionally, conventional diagnostic systems typically rely on a single dataset or a uniform feature space and do not train classifiers using datasets with differing feature pluralities acquired from different measurement techniques, nor do they integrate such datasets through co-normalization to enable unified model training, and further states that the claimed embodiments depart from such approaches by providing a system that can operate across heterogeneous biological data sources, thereby enabling improved robustness and applicability of the resulting classifier, and that this represents more than routine or conventional activity and instead reflects a specific technological advancement, and accordingly, even if the claims were considered to recite an abstract idea, they include an inventive concept and recite significantly more than the alleged exception under Step 2B.
These arguments are not persuasive, because first, the Applicant is reminded that a conclusion of whether a claim is eligible at Step 2B requires that all relevant considerations be evaluated, which comprises steps of: (1) carrying over the identification of any additional element(s) in the claim from Step 2A Prong Two; (2) carrying over the conclusions from Step 2A Prong Two on the considerations discussed in MPEP §§ 2106.05(a) - (c), (e) (f) and (h); (3) re-evaluating any additional element or combination of elements that was considered to be insignificant extra-solution activity per MPEP § 2106.05(g), because if such re-evaluation finds that the element is unconventional or otherwise more than what is well-understood, routine, conventional activity in the field, this finding may indicate that the additional element is no longer considered to be insignificant; and (4) evaluating whether any additional element or combination of elements are other than what is well-understood, routine, conventional activity in the field, or simply append well-understood, routine, conventional activities previously known to the industry, specified at a high level of generality, to the judicial exception, per MPEP § 2106.05(d). Second, and as noted in the above rejection, when taken alone (i.e., individually), all additional elements in claims 1-20 do not amount to significantly more than the above-identified judicial exception(s). Even when evaluated as an ordered combination, the additional elements fail to transform the exception(s) into a patent-eligible application of that exception. Thus, claims 1-20 are deemed to not contribute an inventive concept, i.e., amount to significantly more than the judicial exception(s). Second, the Applicant’s argument appears to present multiple limitations as being relevant to the Step 2B analysis, but that are actually identified as judicial exceptions at Step 2A Prong One in the above rejection, and therefore are not evaluated at Step 2B (e.g., integrating datasets through co-normalization to enable unified model training comprises the limitations identified as judicial exceptions in the above rejection, and therefore comprise limitations that are not relevant to the Step 2B analysis).
The Applicant states on page 21 (middle) of the Remarks that as described in the specification, features are not processed as undifferentiated data, but are instead biologically structured into modules of genes associated with specific phenotypes of the clinical condition, and are summarized prior to input into the classifier, and that this reflects a particular technical implementation of a machine learning model that is specifically adapted to operate on heterogeneous biological data, rather than a generic application of data analysis. The Applicant further states that the claimed operations therefore improve how a machine learning system processes biologically-derived data by imposing a structured, phenotype-driven representation of gene expression results, and that such limitations are not mental processes and cannot be practically performed in the human mind, nor do they merely recite a mathematical concept.
These arguments are not persuasive, because first, regarding the Applicant’s argument that “features are not processed as undifferentiated data, but are instead biologically structured into modules of genes associated with specific phenotypes of the clinical condition,” it is noted in response that this argument appears to be arguing that a particular type of data is not actually just information in the form of data, i.e., a judicial exception, thus, this argument appears to be inconsistent with the eligibility analysis at MPEP 2106. Second, the above rejection identifies claim interpretations reciting a neural network operating on data as mathematical concepts, for example, but in other limitations where a neural network is not operating on the data, the claim limitations are identified as mental processes and/or mathematical concepts, as noted and discussed in the above rejection.
The Applicant states on page 22 (top) of the Remarks that the current amendments specify that, based upon the output, treating the test patient by selecting and administering a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier. The Applicant further states that in view of the current amendments, the classifier output is not merely informational, but is used to directly control treatment decisions, such as administering different therapies depending on the predicted class of the condition, e.g., bacterial versus viral infection. The Applicant further states that this type of limitation is expressly recognized by the USPTO as indicative of eligibility, e.g., in Example 49 (Fibrosis Treatment), claims that apply a model output to select and administer a particular treatment are found to integrate any abstract idea into a practical application because they affect a real-world medical intervention rather than merely generating or displaying data, and similarly, the amended claims do not merely analyze biological data, but instead use the output of a machine learning classifier to drive a specific therapeutic action applied to a patient, and further states that this represents a concrete application of the claimed technology in a clinical setting and goes well beyond insignificant extra-solution activity or a mere instruction to apply it.
These arguments are not persuasive, because first, claim 1 in the instant application is not analogous to Example 49 (claim 2 in particular), not least because claim 2 of Example 49 is directed to a post-surgical fibrosis treatment that recites administering a particular treatment (i.e., Compound X eye drops) to a glaucoma patient, i.e., the hypothetical claim is determined to apply or use a recited judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition (MPEP 2106.04(d)(2)), whereas in contrast, instant claim 1 only recites administering a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier, which does not recite a particular treatment or prophylaxis for a disease or medical condition. Second, the Applicant is reminded that the Eligibility Examples are hypothetical and only intended to be illustrative of the claim analysis performed using MPEP 2106, and of the particular issues noted in the Example, and therefore, the Examples should be interpreted based on the fact patterns set forth in a particular Example, as other fact patterns may have different eligibility outcomes, as evidenced in the rejection of the instant claims above.
Claim Rejections - 35 USC § 103
The amendment received 21 April 2026 has been fully considered, however after further consideration, all rejections under 35 U.S.C. 103 in the Office action mailed 23 October 2025 are maintained in view of the amendment.
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
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 1-14, and 16-20 are rejected under 35 U.S.C. 103 as being unpatentable over Sweeney et al. (“Robust classification of bacterial and viral infections via integrated host gene expression diagnostics.” Science Translational Medicine, 2016, vol. 8, no. 346, pp. 1-12, as cited in the Information Disclosure Statement (IDS) received 09 January 2024, and as cited in the Office action mailed 23 October 2025) in view of Ahn et al. (“Deep Learning-based Identification of Cancer of Normal Tissue using Gene Expression Data.” IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2018, Madrid, Spain, pp. 1748-1752, doi:10.1109/BIBM.2018.8621108, as cited in the Office action mailed 23 October 2025) in view of Khatri et al. (“Biomarkers for use in prognosis of mortality in critically ill patients.” WO 2018/004806), as cited in the Office action mailed 23 October 2025.
Independent claim 1 and dependent claims 2-14 and 16-20 are broadly directed to a classifier that is optimized for use on multiple heterogeneous gene expression datasets.
Sweeney et al. is broadly directed to an integrated multicohort analysis framework to analyze multiple gene expression data sets to identify a biomarker that can classify patients with bacterial or viral infections.
Ahn et al. is broadly directed to training and using a feed-forward neural network architecture to identify cancer or normal tissue using multiple different sources of gene expression data.
Khatri et al. is broadly directed to the use of biomarkers from gene expression panels for the prognosis of mortality, and administering treatment based on the prognosis.
Regarding claim 1, Sweeney et al. shows a robust classification decision model for discrimination of bacterial and viral infections using a multicohort analysis and independent cohorts (Title; and Abstract); and datasets generated from test samples that are whole blood (Table 1 and Table 2). The purpose of the Sweeney et al. study was to use an integrated multicohort analysis framework to analyze multiple gene expression datasets to identify a biomarker that can classify patients with bacterial or viral infections (page 8, col. 1, para. 5); the cohorts of gene expression datasets originated from Affymetrix arrays, Illumina arrays, Agilent, arrays, GE arrays, and other commercial arrays (page 8, col. 2, para. 3). Sweeney et al. shows that there were dozens of public microarray cohorts that profiled patients with either bacterial or viral infections, but not both, and that it would be advantageous to be able to compare a gene score across these cohorts, but that this previously not possible because each different microarray has widely different background measurements for each gene, and among studies using the same types of microarrays, there are large batch effects (page 9, col. 1, para. 4), and as part of an effort to make use of these data, Sweeney et al. developed a method called COCONUT (i.e., Combat CO-Normalization Using conTrols) which normalizes control samples from different cohorts to allow for direct comparison of diseased samples from those same cohorts (page 9, col. 1, para. 5). Sweeney et al. further shows these datasets represent a wide variety of clinical conditions, including a range of infection types (Gram-positive, Gram-negative, atypical bacterial, common respiratory viruses, and dengue) and severities (mild infections to septic shock) (page 4, col. 2, para. 1).
Regarding claim 1, Sweeney et al. does not show processing the test dataset with a main classifier structured as a feed-forward neural network comprising an input layer and an output layer or steps for training the classifier; and based upon the output, treating the test patient.
Regarding claim 1, Ahn et al. shows a deep neural network-based classification model using a feed-forward network architecture for the identification of cancer and normal samples (page 1750, col. 1, para. 1); a heterogenous data set for training the classification model comprising statuses of the condition (Table II.); and generating an output of expression profiles for classification of cancer (page 1751, col. 2, paras. 2-3; and Fig. 5).
Regarding claim 1, Sweeney et al. in view of Ahn et al. does not show based upon the output, treating the test patient by selecting and administering a therapy corresponding to a predicted class of the condition indicated by the output of the main classifier.
Regarding claim 1, Khatri et al. shows biomarkers and methods of using them for aiding diagnosis, prognosis, and treatment of critically ill patients with sepsis, severe trauma, or burns (Title; and Abstract). Khatri et al. further shows diagnosing and treating a patient having an infection, the method comprising: a) obtaining a biological sample from the patient; b) measuring levels of expression biomarkers; and c) administering a sepsis treatment if the patient is diagnosed with sepsis; and in some embodiments, the method further comprises calculating a mortality gene score for the patient based on the levels of the biomarkers, wherein a higher mortality gene score for the patient compared to a control subject indicates that the patient is at high risk of mortality within 30 days, and administering intensive care unit treatment to the patient if the patient is at high risk of mortality within 30 days (page 22, lines 10-23).
Regarding claim 2, Sweeney et al. further shows an integrated antibiotics decision model that can discriminate between patients with severe acute infections and those with inflammation (page 10, col. 1, para. 3).
Regarding claim 7, Sweeney et al. further shows the limitations recited in step (C) of the claim, i.e., the COCONUT method which co-normalizes control samples from different cohorts to allow for direct comparison of diseased samples from those same cohorts and to obtain batch-corrected data (page 9, col.1, paras. 4-5 through col. 2, para. 7).
Regarding claim 9, Sweeney et al. further shows a co-normalizing method that is solved iteratively (page 9, col. 2, para. 2).
Regarding claim 10, Sweeney et al. further shows correcting for location and scale of each gene by first solving an ordinary least-squares model for gene expression and then shrinking the resulting parameters using an empirical Bayes estimator, solved iteratively, and with additive and multiplicative batch effects (page 9, col. 2, paras. 2-7).
Regarding claim 13, Sweeney et al. further shows a gene set optimized for diagnosis [higher in viral infections (IFI27, JUP, and LAX1)] (page 2, col. 2, para. 1).
Regarding claim 20, Sweeney et al. further shows data sets that comprise subsets of data representing diseased and healthy conditions (page 2, Table 1).
Regarding claims 3-9, 11, 12, 14, and 16-19, Sweeney et al. does not show the input layer of the main classifier is configured to receive the plurality of feature values from each of a set of feeder neural network input layers (claim 3); the set of feeder neural network input layers comprises a first feeder input layer for processing mRNA abundance values for a first set of genes represented in the test dataset, and a second feeder input layer for processing mRNA abundance values for a second set of genes represented in the test dataset (claim 4); the set of datasets comprises at least three datasets corresponding to three cohorts of patients, wherein each of the at least three datasets is generated using a different measurement technique (claim 5); generating and training the main classifier, wherein the set of datasets used to train the main classifier comprises non-overlapping, independent datasets (claim 6); steps (A), (B), and (D) (claim 7); co-normalizing feature values comprises determining an expected expression value of each of the first plurality of features and the second plurality of features and adjusting the expected expression value of each of the first plurality of features and the second plurality of features for modifications of mean and standard deviation attributed to execution of the first measurement technique and the second measurement technique (claim 8); a third dataset generated using a third measurement technique, for training the main classifier (claim 9); generating the first dataset using a first measurement technique, and generating the second dataset using a second measurement technique (claim 11); or the first measurement technique comprises RNAseq for a first cohort, and wherein the second measurement technique comprises using DNA microarrays for a second cohort different from the first cohort (claim 12); treating the test patient comprises treating the test patient for a viral infection (claim 14); treating the test patient comprises treating the test patient for sepsis (claim 16); values of the set of features comprises expression values of a first set of genes comprising: DEFA4, CD163, RGS1, PER1, HIF1A, SEPP1, C11orf74 and CIT (claim 17); classifying the test patient as at risk for death within 30 days of hospital admission (claim 18); and the first plurality of features comprises nucleic acid expression features, and wherein the second plurality of features comprises protein expression features (claim 19).
Regarding claim 3, Ahn et al. shows manually curated data from multiple datasets (page 1749, Methods: Sections A & B).
Regarding claim 4, Ahn et al. shows different data sources used in the study from different platforms comprising different sets of gene expression data (page 1749, Table I; and Fig. 1).
Regarding claim 6, Ahn et al. shows training data from different databases comprising different platforms (page 1749, Tables I, II, and III).
Regarding claim 7, steps (A), (B), and (D), Ahn et al. shows training data from different datasets generated from different gene expression platforms and comprising normal and diseased conditions (page 1749, Tables I, II, and III; and Fig. 1).
Regarding claim 8, Ahn et al. shows performing within-sample standardization using the mean and standard deviation of gene expression values (page 1749, col. 2, para. 1).
Regarding claim 9, Ahn et al. shows training a classifier model (page 1749, col. 2, Section B.).
Regarding claims 11 and 12, Ahn et al. shows different datasets generated from different gene expression measuring platforms, i.e., microarray and RNA sequencing (page 1749, Table I).
Regarding claims 5, 9, 14, and 16-19, Sweeney et al. in view of Ahn et al. does not show the set of datasets comprises at least three datasets corresponding to three cohorts of patients, wherein each of the at least three datasets is generated using a different measurement technique (claim 5); a third dataset generated using a third measurement technique (claim 9); treating the test patient comprises treating the test patient for a viral infection (claim 14); treating the test patient comprises treating the test patient for sepsis (claim 16); values of the set of features comprises expression values of a first set of genes comprising: DEFA4, CD163, RGS1, PER1, HIF1A, SEPP1, C11orf74 and CIT (claim 17); classifying the test patient as at risk for death within 30 days of hospital admission (claim 18); and the first plurality of features comprises nucleic acid expression features, and wherein the second plurality of features comprises protein expression features (claim 19).
Regarding claim 5, Khatri et al. further shows an integrated multi-cohort meta-analysis framework to analyze multiple gene expression datasets to identify a set of genes that can predict mortality in patients with sepsis using data from two different gene expression microarray public repositories in addition to datasets from Glue Grant (page 43, lines 10-30 through page 44, lines 1-23).
Regarding claim 9, Khatri et al. further shows that expression levels of biomarkers are determined by measuring polynucleotide levels, and that the levels of transcripts of specific biomarker genes can be determined from the amount of mRNA or polynucleotides derived therefrom, and that polynucleotides can be detected and quantitated by a variety of methods including microarray analysis, polymerase chain reaction (PCR), reverse transcriptase polymerase chain reaction (RT-PCR), Northern blot, and serial analysis of gene expression (SAGE) (page 23, lines 30-31 through page 24, lines 1-5).
Regarding claim 14, Khatri et al. further shows a critically ill patient diagnosed with a viral infection is further administered a therapeutically effective dose of an antiviral agent (page 19, lines 27-31).
Regarding claim 16, Khatri et al. further shows administering a sepsis treatment comprising antimicrobial therapy, supportive care, or an immune-modulating therapy if the patient is diagnosed with sepsis (page 5, lines 27-28).
Regarding claims 17 and 18, Khatri et al. further shows a method for determining mortality risk and treating a patient suspected of having a life-threatening condition by first analyzing the levels of expression and wherein increased levels of expression of the DEFA4, CD163, PER1, RGS1, HIF1A, SEPP1, C11orf74, and CIT biomarkers compared to the reference value ranges for the biomarkers for a control subject indicate that the patient is at high risk of mortality within 30 days (page 3, lines 12-24).
Regarding claim 19, Khatri et al. further shows that patient data is analyzed by one or more methods including multi-dimensional protein identification technology (MUDPIT) technology (page 6, lines 8-12).
Therefore, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the methods shown by Sweeney et al. by incorporating a DNN-based classification model (i.e., a feed-forward neural network) for the identification of cancer and normal samples using gene expression data as shown by Ahn et al. and discussed above. One of ordinary skill in the art would have been motivated to combine the method of Sweeney et al. with the method of Ahn et al., because Ahn et al. explains that deep learning has proven to show outstanding performance in resolving recognition and classification problems (Abstract), and shows a method for building a universal classifier by deep learning of gene expression data stored in large public databases. This modification would have had a reasonable expectation of success given that both Sweeney et al. and Ahn et al. disclose using gene expression data from multiple gene expression repositories to generate multicohort analysis frameworks to analyze multiple gene expression datasets to identify biomarkers that can classify patients according to a condition, e.g., patients that are healthy versus patients with cancer or with bacterial or viral infections.
It would have been further prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the methods shown by Sweeney et al. in view of Ahn et al. by incorporating biomarkers and methods of using them for prognosis of mortality in critically ill patients with sepsis, as shown by Khatri et al. and discussed above. One of ordinary skill in the art would have been motivated to combine the method of Sweeney et al. in view of Ahn et al. with the method of Khatri et al. because Khatri et al. shows using biomarkers identified from gene expression analyses for determining mortality risk and treating a patient suspected of having a life-threatening condition such as sepsis. This modification would have had a reasonable expectation of success given that both Sweeney et al. in view of Ahn et al. and Khatri et al. disclose using gene expression levels of biomarkers to classify a patient according to a risk or a condition.
Claim 15 is rejected under 35 U.S.C. 103 as being unpatentable over Sweeney et al. (hereinafter Sweeney et al. (2016)) in view of Ahn et al. in view of Khatri et al. as applied to claims 1-14 and 16-20 above, and further in view of Sweeney et al. (hereinafter Sweeney et al. (2015)) (“A comprehensive time-course-based multicohort analysis of sepsis and sterile inflammation reveals a robust diagnostic gene set.” Science Translational Medicine, 2015, vol. 7, no. 287, pp. 1-15, as cited in the Information Disclosure Statement (IDS) received 09 January 2024, and as cited in the Office action mailed 23 October 2025).
Dependent claim 15 is directed to the expression values of a particular group of genes.
Sweeney et al. (2015) is broadly directed to a comprehensive time-course-based multicohort analysis of sepsis and sterile inflammation.
Sweeney et al. (2016) in view of Ahn et al. in view of Khatri et al. as applied to claims 1-14 and 16-20 above, do not show values of the set of features comprises expression values of a first set of genes comprising: CEACAM1, ZDHHC19, C9orf95, GNA15, BATF, and C3AR1 (claim 15).
Regarding claim 15, Sweeney et al. (2015) shows an 11-gene set that separates SIRS/trauma from sepsis that includes the genes CEACAM1, ZDHHC19, C9ORF95, GNA15, BATF, and C3AR1 (page 5, Table 3).
Therefore, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the methods shown by Sweeney et al. (2016) in view of Ahn et al. in view of Khatri et al. as applied to claims 1-14 and 16-20 above, by incorporating a diagnostic gene set that distinguishes sterile inflammation from infectious inflammation as shown by Sweeney et al. (2015) and discussed above. One of ordinary skill in the art would have been motivated to combine the method of Sweeney et al. (2016) in view of Ahn et al. in view of Khatri et al. as applied to claims 1-14 and 16-20 above, with the methods of Sweeney et al. (2015) because Sweeney et al. (2015) explains that although many studies of gene expression in sepsis had been published, distinguishing sepsis form a sterile systemic inflammatory response syndrome (SIRS) was still largely up to clinical suspicion, and therefore the diagnostic 11-gene set of Sweeney et al. (2015) combined with the machine learning classifier shown by Sweeney et al. (2016) in view of Ahn et al. in view of Khatri et al. as applied to claims 1-14 and 16-20 above, would have had a reasonable expectation of success given that they both disclose classification of conditions using expression diagnostics.
Response to Arguments
The Applicant’s arguments/remarks received 21 April 2026 have been fully considered, but are not persuasive.
The Applicant points to Ex parte H. Garrett Wada et al. and In re Suitco Surface, Inc., on page 23 of the Remarks, and states (para. 2) that the combined references fail to teach or suggest all claimed limitations, as is required, and that this is particularly noticeable when the claims are interpreted in light of the underlying specification, as is always required. The Applicant further summarizes aspects of the rejection under 35 U.S.C. in the Office action mailed 23 October 2025 that are not shown by the primary reference, and further states on page 24 (para. 1) that the Applicant disagrees with the Office’s position that the secondary references collectively teach those limitations. The Applicant further states (para. 2) that the Applicant disagrees with the Office’s position that the Ahn reference discloses a heterogeneous data set for training the classification model, and further states (para. 3) that Ahn does not disclose performing inter-sample normalization, and further states that Ahn does not disclose training a neural network on a heterogeneous repository of input data having different features. The Applicant further states on page 25 (para. 1) that the Khatri reference does not disclose any type of machine learning classifier or treatment decisions based on the output of such a model. The Applicant further states (para. 2) that the claims have been amended solely in an effort to facilitate compact prosecution, and further states on page 26 (para. 1) that the teachings of the combined references are readily distinguishable from the claimed limitations, either as previously presented or as currently amended. The Applicant further summarizes aspects of the references used in the rejection under 35 U.S.C. 103 in the Office action mailed 23 October 2025 and how the amended claims recite limitations that are not disclosed by those references, e.g., the Applicant states that as described in the specification, features are not treated as undifferentiated inputs, but are instead organized into biologically meaningful modules of genes associated with particular phenotypes of the clinical condition and are summarized prior to input into the classifier (Remarks: page 26, para. 2, and page 27 para. 1).
These arguments are not persuasive, because first, the Ahn reference discloses at least in the Abstract that Ahn first trained the deep neural network (DNN) to discriminate between cancer and normal samples using various gene selection strategies and therapeutic target genes from commercial cancer panels and genes in NCI-curated cancer pathways. Second, the Ahn reference is not relied on to show performing inter-sample normalization (see Sweeney et al. for inter-sample normalization in the above rejection). Third, the Khatri reference is not relied on to show using a main classifier structured as a feed-forward neural network (see the Ahn reference in the above rejection for the use of a feed-forward neural network), but rather, the Khatri reference is relied on to show that it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified Sweeney et al. in view of Ahn et al. to incorporate methods using biomarkers identified from gene expression analyses for determining mortality risk and treating a patient suspected of having a life-threatening condition such as sepsis. Fourth, each of the references used in the above rejection show or at least suggest using gene expression datasets that organized by phenotype, e.g., the Sweeney reference shows using datasets of multiple clinically heterogeneous cohorts representing a wide range of clinical conditions (i.e., phenotypes), as noted in the above rejection.
Conclusion
No claims are allowed.
THIS ACTION IS MADE FINAL. 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.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to STEVEN W. BAILEY whose telephone number is (571)272-8170. The examiner can normally be reached Mon - Fri. 1000 - 1800.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, KARLHEINZ SKOWRONEK can be reached at (571) 272-9047. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/S.W.B./Examiner, Art Unit 1687
/Joseph Woitach/Primary Examiner, Art Unit 1687
Prosecution Insights