DETAILED ACTION
Applicant’s response, filed 17 Feb. 2026 has been fully considered. The following rejections and/or objections are either reiterated or newly applied. They constitute the complete set presently being applied to the instant application.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
Claims 5, 7-10, 14, 19, 22-23, 25, and 38-39 are cancelled.
Claims 1-4, 6, 11-13, 15-18, 20-21, 24, and 26-37 are pending.
Claims 2, 4, 20-21, 24, and 26-34 are withdrawn from further consideration pursuant to 37 CFR 1.142(b) as being drawn to a nonelected invention, there being no allowable generic or linking claim. Election was made without traverse in the reply filed on 16 Feb. 2024.
Claims 1, 3, 6, 11-13, 15-18, and 35-37 are rejected.
Priority
The effective filing date of the claimed invention is 27 Feb. 2018.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 17 Feb. 2026 is compliance with the provisions of 37 CFR 1.97. Accordingly, the list of cited references were considered by the Examiner.
Claim Interpretation
Claims 1 and 3 recite “(G)/(C) generating/generate an output indicative of whether adjuvant therapy is recommended for the subject based on the eTF exceeding the empirical threshold”. The claims previously recite “detecting a residual disease in the subject based on a comparison of the eTF…to an empirical threshold”, but does not require the eTF exceeded the threshold. Therefore, under the broadest reasonable interpretation of the claims, the method of claim 1 does not require generating the recited output, while claim 3 requires the processor is configured to generate the output. See MPEP 2111.04 II.
Claim 12 recites “the at least one error suppression protocol includes correction of artefactual mutations generated…, wherein the correction comprises correcting base calls of the R1 and R2…when a discordance between the base calls of the R1 and R2 at the overlapping position is identified”. Therefore, the correction of artefactual mutations generated in claim 12 is contingent upon there being a discordance between the base calls of the R1 and R2, and thus, under the broadest reasonable interpretation of the claims, the method of claim 12 does not require the correction of artefactual mutations generated by paired-end 150 bp sequencing as claimed. See MPEP 2111.04 II.
Claim Rejections - 35 USC § 112(a)
The rejection of claims 1, 3, 6-7, 10-18, and 35-39 under 35 U.S.C. 112(a) in the Office action mailed 17 Sept. 2025 has been withdrawn in view of claim amendments and cancellations received 17 Feb. 2026.
Claim Rejections - 35 USC § 112(b)
The rejection of claims 1, 3, 6-7, 10-18, and 35-39 under 35 U.S.C. 112(b) in the Office action mailed 17 Sept. 2025 has been withdrawn in view of claim amendments and cancellations received 17 Feb. 2026.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
Claims 1, 3, 6, 11-13, 15-18, and 35-37 are rejected under 35 U.S.C. 112(b) as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, regards as the invention. This rejection is newly recited and necessitated by claim amendment.
Independent claim 1, and claims dependent therefrom, are indefinite for recitation of “(D) filtering artefactual noise, using a trained machine learning model, from the first and the second compendium of reads using at least one error suppression protocol…”. Dependent claim 11, then recites “the at least one error suppression protocol comprises: (a) removing each single nucleotide variation (SNV) identified as an artefactual mutation by calculating…”; and/or (b) removing each SNV identified as an artefactual mutation using discordance testing…”. It is unclear if (1) claim 1 intends to require that each of the at least one error suppression protocols use machine learning model, such that the (a) calculating and/or (b) using discordance testing in claim 11 are required to use machine learning, or (2) if claim 1 intends for the filtering of artefactual noise to use both a trained machine learning model and at least one error suppression protocol, but the error suppression protocol(s) are not necessarily required to use machine learning. A review of Applicant’s specification appears to disclose that error suppression protocols include discordance testing (recited in claim 11) and may be performed separate to a machine-learning based error suppression ([0013]; [0176]-[0178]; FIG. 1D). It is further noted Applicant’s specification does not appear to provide support for the machine learning model performing the discordance testing, and instead only provides support for the machine learning model calculating a score based on the recited features of MQ, MBQ, PIR, and MRBQ in claim 11 (see FIG. 1D, support-vector-machine training). Therefore, in light of Applicant’s specification, the claims will be interpreted to mean that artefactual noise is filtered using a trained machine learning model and using at least one error suppression protocol. Clarification is requested via claim amendment.
Claims 1 and 3, and claims dependent therefrom, are indefinite for recitation of “…(E)…wherein the first filtered read sets provides a subject-specific baseline for the genetic markers used by the one or more integrative mathematical models”. Claims 1 and 3 previously recite “wherein the one or more integrative mathematical models integrate coverage, mutation load, number of detected mutations, and fragment-size shift”, but do not recite a particular set of genetic markers used by the models. As a result, it is unclear if Applicant intends for “the genetic markers” to be referring to the genetic markers from the first biological sample of the subject in step (A), or if this is some other set of genetic markers, such as a subset of the genetic markers in (A), being used in the model(s). Clarification is requested via claim amendment.
Claim 3 is indefinite for recitation of “computing/compute an estimated tumor fraction (eTF) of the second biological sample using the second filtered read sets”. There is insufficient antecedent basis for multiple second filtered read sets in the claim because claim 3 only previously recites one second filtered read set in (D). For purpose of examination, the claim is interpreted to mean “using the second filtered read set[[s]]”.
Claims 3 and 11, and claims dependent therefrom, are indefinite for recitation of “each single nucleotide variation (SNV) identified as an artefactual mutation by calculating an artefact probability for each SNV in the first and the second compendium of reads” in step (D), substeps (a) and (b) of the claims. There is insufficient antecedent basis for “each SNV identified as an artefactual mutation” and “each SNV in the first and second compendium of reads” in the claims because claim 1, from which claim 11 depends, and claim 3 do not require that the first and second compendium of reads comprise SNVs, and furthermore, the claims previously recite “filtering artefactual noise…to produce (i) a first filtered read set…and (ii) a second filtered read set”, but do not disclose any SNVs identified as artefactual mutations or require that the filtered artefactual noise is artefactual SNVs. As a result, it is further unclear if the step of removing each SNV identified as an artefactual mutation is contingent upon there being any SNVs in the first/second compendium of reads, or if the claims intend to require the first and second compendium of reads include SNVs. Clarification is requested Claims 3 and 11 are interpreted to mean that the filtering of artefactual noise using the at least one error suppression protocol comprises removing one or more single nucleotide variations identified as an artefactual mutation by either calculating as in (a) or using discordance testing in (b).
Response to Arguments
Applicant's arguments filed 17 Feb. 2026 regarding 35 U.S.C. 112(b) have been fully considered but they are not persuasive because they do not pertain to the new grounds of rejection set forth above.
Claim Rejections - 35 USC § 101
The rejection of claims 7, 10, 14, and 38-39 under 35 U.S.C. 101 in the Office action mailed 17 Sept. 2025 has been withdrawn in view of claim amendments received 17 Feb. 2026.
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, 3, 6, 11-13, 15-18, and 35-37 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea and law of nature without significantly more. Any newly recited portion is necessitated by claim amendment.
The Supreme Court has established a two-step framework for this analysis, wherein a claim does not satisfy § 101 if (1) it is “directed to” a patent-ineligible concept, i.e., a law of nature, natural phenomenon, or abstract idea, and (2), if so, the particular elements of the claim, considered “both individually and as an ordered combination,” do not add enough to “transform the nature of the claim into a patent-eligible application.” Elec. Power Grp., LLC v. Alstom S.A., 830 F.3d 1350, 1353 (Fed. Cir. 2016) (quoting Alice, 134 S. Ct. at 2355). Applicant is also directed to MPEP 2106.
Step 1: The instantly claimed invention (claims 1 and 3 being representative) is directed to a method and system for detecting residual disease. Therefore, the instantly claimed invention falls into one of the four statutory categories. [Step 1: YES]
Step 2A: First it is determined in Prong One whether a claim recites a judicial exception, and if so, then it is determined in in Prong Two if the recited judicial exception is integrated into a practical application of that exception.
Step 2A, Prong 1: Under the MPEP § 2106.04, the Step 2A (Prong 1) analysis requires determining whether a claim recites an abstract idea, law of nature, or natural phenomenon.
Claims 1 and 3 recite the following steps which fall under the mathematical concepts and/or mental processes groupings of abstract ideas:
filtering/filter artefactual sites from the first compendium of reads by removing recurring sites generated over a cohort of reference healthy samples and/or removing germ line mutations from the genetic markers;
detecting/detect, within a second compendium of reads from a second biological sample of the subject, reads corresponding to the genetic markers associated with the first compendium of reads to generate a tumor-associated genome-wide representation of the genetic markers, wherein the second compendium of reads is subject-specific and genome-wide, and wherein the second biological sample comprises a follow-up sample distinct from the baseline sample and normal sample.
filtering/filter artefactual noise (using a trained machine learning model in claim 1) from the first and second compendium of reads using at least one error suppression protocol to produce (i) a first filtered read set based on and spanning the first compendium of reads and a (ii) second filtered read set based on and spanning the second compendium of reads,
(claim 3 only) wherein the at least one error suppression protocol comprises
(a) removing each SNV identified as an artefactual mutation by calculating an artefact probability for each SNV in the first and second compendium of reads, wherein the artefact probability is calculated as a function of features selected from the group consisting of mapping-quality (MQ), variant base-quality (MBQ), position-in-read (PIR), mean read base quality (MRBQ), and combinations thereof; and/or
(b) removing each SNV identified as an artefactual mutation using discordance testing, wherein the discordance testing groups together independent replicates of a same DNA fragment in PCR or sequencing, identifies a discordant SNV as the artefactual mutation, and removes the discordant SNV;
computing/compute an estimated tumor fraction (eTF) of the second biological sample using the second filtered read sets by applying one or more integrative mathematical models and a background noise model, wherein the one or more integrative mathematical models integrate coverage, mutation load, number of detected mutations, and fragment size-shift, and wherein the background noise model integrates a panel-of-normals (PON) framework and a z-score to determine statistical significance, and wherein the first filtered rest set provides a subject specific baseline for the genetic markers used by the one or more integrative mathematical models; and
detecting/detect a residual disease in the subject based on a comparison of the eTF in the second biological sample to an empirical threshold;
generating/generate an output indicative of whether adjuvant therapy is recommended for the subject based on the eTF exceeding the empirical threshold.
The identified claim limitations falls into one of the groups of abstract ideas of mental processes, for the following reasons. In this case, filtering artefactual sites encompasses identifying markers present in a cohort of reference healthy samples and removing these markers from the first compendium of reads, which amounts to a mere analysis of data. Furthermore, detecting reads from a second-subject-specific genome wide compendium of genetic markers in a second biological sample to generate a tumor-associated genome-wide representation of genetic markers encompasses analyzing and collecting read information corresponding to marker locations of sequencing information generated from a tumor sample of the subject, which is a mental process. Filtering noise from the first and second compendium of reads using an error suppression protocol of calculating a probability that a variant is artefactual and removing the variant can be practically performed in the mind, for example, by inputting quality features into al linear regression model to calculate a probability, and removing variants with a probability above a certain threshold. The other alternative embodiments for filtering noise also amount to a mere analysis of data to identify variants that involve performing data comparisons to identify discordance. Computing an estimated tumor fraction by using one or more integrative mathematical models and a background noise model can be practically performed in the mind by performing addition and multiplication to perform the calculation, as discussed in Applicant’s specification (see FIG. 1D). Detecting a residual disease involves simply comparing the estimated statistical significance framework to a threshold, which is a mental process. Providing a diagnosis of residual disease involves analyzing if the framework is greater than a threshold, which amounts to mental data comparisons. Last, the step of generating an output indicative of whether adjuvant therapy is recommended recites a mental process given it encompasses analyzing the eTF against a threshold and generating a recommendation of adjuvant therapy if the eTF exceeds the threshold. That is other, than reciting the limitations are performed by a processor in claim 3, nothing in the claims precludes the step from being practically performed in the mind. See MPEP 2106.04(a)(2) III.
The steps of using a trained machine learning model and calculating the probability that any single nucleotide variation is an artefactual mutation as a function of features and computing a statistical significance framework using models further recite a mathematical concept. The claim limitations amount to a textual equivalent to performing mathematical calculations. For example, the step of using a trained machine learning model and calculating a probability encompasses inputting numerical values of the features into a linear regression model and performing weighted addition to calculate the probability. Similarly, for estimating a tumor fraction in light of Applicant’s specification (see FIG. 1D), the framework can be estimated by performing addition and multiplication to perform the calculation. Therefore, these limitations recite a mathematical concept.
Last, the claims recite the law of nature of a natural correlation between the presence of genetic markers and residual disease. See MPEP 2106.04(b).
Dependent claims 6, 11-13, 15-18, and 35-37 further recite an abstract idea and/or are part of the abstract idea of claims 1 and 3 above. Dependent claim 6 further recites the mental process of generating a panel of normal (PON) blacklist or mask. Dependent claims 11-13 further limit the filtering noise in claim 1, and thus are part of the judicial exception of filtering noise. Dependent claims 15-16 further limit the mathematical concept and mental process of filtering noise using the background noise model to calculate expected noise distributions and provide an estimated mean and standard-deviation of artefactual mutation detection rate, respectively. Dependent claim 17 further recites the mental process and mathematical concept of integrating fragment size shift into the one or more mathematical model. Dependent claim 18 further recites the mental process and mathematical concept of analyzing intra-subject fragment size shifts in a list of tumor-specific markers and random markers using statistical methods. Dependent claims 35-36 further limit the mental process and mathematical concept of employing a machine learning algorithm to employ one or more support vector machine (SVM). Dependent claim 37 further limits the mental process and mathematical concept of using statistical methods to use one or more tests of significance or a Gaussian mixture model. Therefore, claims 1, 3, 6, 11-13, 15-18, and 35-37 recite an abstract idea. [Step 2A, Prong 1: YES]
Step 2A: Prong 2: Under the MPEP § 2106.04, the Step 2A, Prong 2 analysis requires identifying whether there are any additional elements recited in the claim beyond the judicial exception(s), and evaluating those additional elements to determine whether they integrate the exception into a practical application of the exception. This judicial exception is not integrated into a practical application for the following reasons.
Claims 6, 11-13, 15-18, and 35-37 do not recite any elements in addition to the judicial exception and thus are part of the judicial exception.
The additional elements of claims 1 and 3 include:
a computer (claim 1);
an analyzing unit comprising a processor and a computing unit comprising a processor (claim 3);
receiving/receive a first compendium of reads associated with genetic markers from a first biological sample of a subject, wherein the first compendium of reads is subject-specific and genome-wide, wherein the first biological sample comprising a baseline sample and a normal sample, wherein the baseline sample comprises a tumor sample or a plasma sample, and wherein the normal cell sample comprises peripheral blood mononuclear cells (PBMCs) (claims 1 and 3) (i.e. receiving data); and
The additional elements of a processor/computer and receiving data are generic computer components and/or functions. The courts have found the use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application. See MPEP 2106.05(f).
Furthermore, the steps of receiving reads only serve to collect data for use by the abstract idea, which amounts to insignificant extra-solution that does not integrate the recited judicial exception into a practical application. See MPEP 2106.05(g).
Therefore, the additionally recited elements amount to insignificant extra-solution activity and/more merely uses a computer as a tool, and, as such, the claims as a whole do no integrate the abstract idea into practical application. Thus, claims 1, 3, 6, 11-13, 15-18, and 35-37 are directed to an abstract idea. [Step 2A, Prong 2: NO]
Step 2B: In the second step it is determined whether the claimed subject matter includes additional elements that amount to significantly more than the judicial exception. See MPEP § 2106.05.
The claims do not include any additional steps appended to the judicial exception that are sufficient to amount to significantly more than the judicial exception.
Claims 6, 11-13, 15-18, and 35-37 do not recite any elements in addition to the judicial exception and thus are part of the judicial exception.
The additional elements of claims 1 and 3 are outlined above.
The additional elements of a processor/computer and receiving data are conventional computer components and/or functions. The courts have found the use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not provide significantly more. See Affinity Labs v. DirecTV, 838 F.3d 1253, 1262, 120 USPQ2d 1201, 1207 (Fed. Cir. 2016) (cellular telephone); TLI Communications LLC v. AV Auto, LLC, 823 F.3d 607, 613, 118 USPQ2d 1744, 1748 (Fed. Cir. 2016) (computer server and telephone unit).
Therefore, taken alone, the additional elements do not provide significantly more. Even when viewed as a combination, the additional elements fail to transform the exception into a patent-eligible application of that exception. Thus, the claims as a whole do not amount to significantly more than the exception itself. [Step 2B: NO]
Therefore, the instantly rejected claims are not drawn to eligible subject matter as they are directed to an abstract idea and natural correlation without significantly more. For additional guidance, applicant is directed generally to applicant is directed generally to the MPEP § 2106.
Response to Arguments
Applicant's arguments filed 17 April 2026 regarding 35 U.S.C. 101 have been fully considered but they are not persuasive.
Applicant remarks the claims do not merely analyze existing data in the abstract, but recite a particularized data-gathering architecture that begins with the generation of subject-specific, genome-wide compendia of sequence reads from multiple biologically distinct samples, while prior to 2018, conventional cfDNA-based cancer monitoring relied on targeted panels or limited CNV calling pipelines (Applicant’s remarks at pg. 15, para. 7 to pg. 16, para. 1 and pg. 19, para. 1 and 3).
This argument is not persuasive because it is not commensurate with the scope of the claims. The claims do not require any steps of actually sequencing the various distinct samples using whole genome sequencing to produce the genome-wide compendia of sequence reads, such that these additional elements may potentially amount to significantly more than the judicial exception for being unconventional. Instead, the claims merely use a computer to receive already generated sequence reads. As discussed in the above rejection, the courts have found the use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application or provide significantly more. See Affinity Labs v. DirecTV, 838 F.3d 1253, 1262, 120 USPQ2d 1201, 1207 (Fed. Cir. 2016) (cellular telephone); TLI Communications LLC v. AV Auto, LLC, 823 F.3d 607, 613, 118 USPQ2d 1744, 1748 (Fed. Cir. 2016) (computer server and telephone unit).
Applicant remarks the claims detect reads corresponding to filtered genetic markers within a follow-up dataset, rather than calling variants de novo, which is not a generic mental comparison but a data gathering constraint that requires preservation of genome-wide coverage while interrogating signal at previously defined loci, and such detection is inseparable from the way sequence data is generated and cannot be performed without the physical sequencing outputs (Applicant’s remarks at pg. 15, para. 2).
This argument is not persuasive. It is agreed that analyzing sequencing data does require a physical sequencing output was generated. However, a claim may still recite a mental process of analyzing data, even if the data is required to be generated by a physical process. See MPEP 2106.05(g) providing examples of insignificant extra-solution activity that were not sufficient to integrate a recited judicial exception into a practical application, including performing clinical tests on individuals to obtain input for an equation, In re Grams, 888 F.2d 835, 839-40; 12 USPQ2d 1824, 1827-28 (Fed. Cir. 1989). As discussed in the above rejection, the human mind is practically able to analyze and remove recurring sites generated over a cohort of reference samples and/or germline mutations (e.g. mutations present in both the tumor and PBMC sample).
Applicant remarks the claims not use a trained machine learning model in conjunction with an error suppression protocol that includes calculating an artefact probability, and the use of such trained machine learning models for systematic error suppression was an emerging technology and was not conventional in the field (Applicant’s remarks at pg. 16, para. 3 and pg. 19, para. 2-3). Applicant remarks these techniques exploit physical properties of sequencing reads that were not routinely measured, normalized, or modeled together (Applicant’s remarks at pg. 19, para. 3).
This argument is not persuasive. First, it is noted that the argument on pg. 16 is provided under Applicant’s “Step 2A (Prong 1)” section. However, conventionality of an abstract idea is not a consideration when determining whether a claim recites an abstract idea under Step 2A, Prong 1. See MPEP 2106.04(b), explaining, nor can one patent "a novel and useful mathematical formula," Parker v. Flook, 437 U.S. 584, 585, 198 USPQ 193, 195 (1978).
If Applicant intends to argue the conventionality of additional elements under Step 2B, it is noted that under Step 2B, whether any additional element or combination of additional elements were well-understood, routine, and conventional are evaluated to determine whether a claim amounts to significantly more than the judicial exception. See MPEP 2106.05 II. However, the conventionality of the abstract idea is not a consideration under step 2B. As identified in the above rejection, the step of filtering artefactual noise using a trained machine learning model and an error suppression protocol to calculate an artifact probability are part of the abstract idea, and thus is not evaluated under Step 2B.
Applicant remarks the claims further require computing an estimated tumor fraction (eTF) using mathematical models and a background noise model, which do not operate on abstract data but on physically constrained datasets whose structure is dictated by the sequencing process, and the integration of coverage, mutation load, mutation count, and fragment-size shift were not routinely exploited in minimal residual disease detection pipelines prior to 2018 (Applicant’s remarks at pg. 16, para. 4 to pg. 17, para. 1).
This argument is not persuasive. As identified in the above rejection, the step of estimating a tumor fraction using the various models is part of the abstract idea, and whether the abstract idea is conventional is not evaluated under Step 2B.
Furthermore, as explained above, simply because data being analyzed by the abstract idea was generated (e.g. constrained) by a physical process does not preclude the claim from reciting an abstract idea. See MPEP 2106.05(g) as cited above, and furthermore, MPEP 2106.04(a)(2) I, C., which provides examples of claims reciting mathematical concepts using data generated by physical measurements, including a relationship between reaction rate and temperature, which relationship can be expressed in the form of a formula called the Arrhenius equation, Diamond v. Diehr; 450 U.S. at 178 n. 2, 179 n.5, 191-92, 209 USPQ at 4-5 (1981).
Applicant remarks that accordingly, the claims are not directed to an abstract idea under Step 2A, Prong 1 because they are directed to a specific technical solution for acquiring and structuring biological measurement data that overcomes physical limitations of low tumor fraction cfDNA (Applicant’s remarks at pg. 17, para. 2).
This argument is not persuasive. First, it is noted that while Applicant mentions “Step 2A, Prong 1”, Applicant appears to be arguing the claims are not directed to an abstract idea under Step 2A, Prong 2, and the argument will be considered accordingly.
MPEP 2106.05(a) explains it is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements. See the discussion of Diamond v. Diehr, 450 U.S. 175, 187 and 191-92, 209 USPQ 1, 10 (1981)) in subsection II, below. In addition, the improvement can be provided by the additional element(s) in combination with the recited judicial exception. Furthermore, it is important to keep in mind that an improvement in the abstract idea itself (e.g. a recited fundamental economic concept) is not an improvement in technology.
In the instant case, the “unconventional” limitations discussed above by Applicant are part of the abstract idea, and thus cannot provide the improvement alone. Furthermore, Applicant’s arguments relating to generating sequence reads are not commensurate with the scope of the claims because the claims do not require physical sequencing steps. As presently, recited, the only additional elements in the claims are generic computer components and steps of receiving data for use by the abstract idea, which do not integrate the judicial exception into a practical application. See MPEP 2106.05(f) and 2106.05(g), discussed above.
Applicant remarks the claims as amended cannot be practically performed in the human mind because the claims require filtering artefactual noise using a trained machine learning model, calculating artifact probabilities, performing discordance testing, and applying a background noise model and integrative mathematical models, and these operations require computational processing of millions of sequencing reads with metadata, which cannot be performed in the mind (Applicant’s remarks at pg. 17, para. 3).
This argument is not persuasive. Applicant merely asserts the limitations cannot be practically performed in the mind due to a large amount of data being processed. However, the amount of data, in and of itself is not a limitation which takes a process out of the realm of the human mind. It is the process performed on that data which is the mental step, and mental steps identified in the claims do not have to be fastest, most efficient, or require specialized computing elements. Explanations for why each of these limitations recite a mental process and/or a mathematical concept are provided in the above rejection.
While computations on large amounts of data performed mentally, or with paper and pencil, would take considerable time and effort, the purpose of computers and computer networks is to perform large numbers of calculations, via algorithms, rapidly, and without error (assuming no error in user input). Although a general-purpose computer can perform calculations at a rate and accuracy that can far outstrip the mental performance of a skilled artisan, the nature of the activity is essentially the same, and constitutes an abstract idea. See Bancorp Serves., L.L. C. v. Sun Life Assur. Co. of Canada (U.S.) (holding that “the fact that the required calculations could be performed more efficiently via a computer does not materially alter the patent eligibility of the claimed subject matter”); see also SiRF Tech., Inc. v. Int’l Trade Comm ’n, (Fed. Cir. 2010) (holding that: In order for the addition of a machine to impose a meaningful limit on the scope of a claim, it must play a significant part in permitting the claimed method to be performed, rather than function solely as an obvious mechanism for permitting a solution to be achieved more quickly, i.e., through the utilization of a computer for performing calculations).
Last, it is noted that even if these steps did not recite a mental process, the claims still recite a mathematical concept as identified in the above rejection. Therefore, the analysis would still proceed to Step 2A, Prong 2.
Applicant remarks the methods and systems do not apply mathematics in the abstract, but only after a specific sequence of non-conventional data acquisition and condition steps, including germline subtraction, cohort-based filtering, noise suppression, and normalization, and the models operate on physically constrained datasets whose structure is dictated by the sequencing process (Applicant’s remarks at pg. 18, para. 1).
This argument is not persuasive. The data acquisition and condition steps of germline subtraction, cohort-based filtering, noise suppression, and normalization are all part of the abstract idea, as discussed in the above rejection. Furthermore, for the reasons already discussed above, simply because the data was generated by a physical assay does not preclude the claim from reciting an abstract idea.
It is further noted that this argument is provide under Applicant’s Step 2A, Prong 2 section. However, the above steps are part of the abstract idea, and thus cannot integrate themselves into a practical application.
Applicant remarks the output of the process is not merely a number, but a recommendation for adjuvant therapy, which effects a real-world medical decision that could be reliably made using conventional pre-2018 techniques (Applicant’s remarks at pg. 18, para. 2).
This argument is not persuasive. Generating an output indicative of whether adjuvant therapy is recommended for the subject based on the eTF exceeding the empirical threshold recites a mental process of determining an indication of whether a subject should take adjuvant therapy based on determining the tumor fraction exceeds the threshold, which is a simple data comparison. However, the claim does not recite any additional elements that affect a real-world medical decision, such as a step of treating the subject with an adjuvant therapy. As explained above, an improvement in the abstract idea is not an improvement to technology (see MPEP 2106.05(a)), and furthermore, the abstract idea cannot integrate itself into a practical application.
Applicant remarks the claims improve the technical operation of sequencing-based diagnostics by combining subject-specific genome-wide marker identification with trained machine learning model error suppression, artefact filtering, and mathematical models to enable detection, which is an improvement to a technical field (Applicant’s remarks at pg. 18, para. 3).
This argument is not persuasive. MPEP 2106.05(a) explains it is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements. See the discussion of Diamond v. Diehr, 450 U.S. 175, 187 and 191-92, 209 USPQ 1, 10 (1981)) in subsection II, below. In addition, the improvement can be provided by the additional element(s) in combination with the recited judicial exception. Furthermore, it is important to keep in mind that an improvement in the abstract idea itself (e.g. a recited fundamental economic concept) is not an improvement in technology.
In the instant case, the steps of identifying markers, using a trained machine learning model, filtering, etc., are part of the abstract idea and thus cannot provide integration. Furthermore, an improved in the diagnosis of residual disease based on the claimed analysis of data amounts to an improved abstract idea, which is not a technology.
As presently, recited, the only additional elements in the claims are generic computer components and steps of receiving data for use by the abstract idea, which do not integrate the judicial exception into a practical application. See MPEP 2106.05(f) and 2106.05(g), discussed above.
Applicant remarks that conventional cfDNA cancer detection predominantly relied on targeted sequencing, molecular barcoding, or hotspot mutation tracking rather than using genome-wide detection, and there was no routine practice of combining genome-wide compendium from the various tissues followed by a trained machine learning model for error suppression and the claimed models for estimating tumor fraction (Applicant’s remarks at pg. 19, para. 1).
This argument is not persuasive. As discussed above, the claims do not require any steps of actually sequencing the various distinct samples using whole genome sequencing to produce the genome-wide compendia of sequence reads, such that these additional elements may potentially amount to significantly more than the judicial exception for being unconventional. Instead, the claims merely use a computer to receive already generated sequence reads, which amount to conventional computer components and processes as discussed in the above rejection. In, addition the use of the machine learning model and estimating tumor fraction are also part of the abstract idea, and thus their conventionality is not evaluated under Step 2B.
Applicant remarks the assertion that the steps could be performed mentally ignores the fact that the claimed methods and systems depend on biological measurements that could not be perceived by humans, produced by high-throughput sequencing instruments, and the inventive concept lies in how the data are gathered, conditioned, and constrained so that the arithmetic becomes meaningful, and thus the claims recite a non-conventional diagnostic architecture (Applicant’s remarks at pg. 20, para. 1-2).
This argument is not persuasive. As already discussed above, simply because data being analyzed by the abstract idea was generated (e.g. constrained) by a physical process does not preclude the claim from reciting an abstract idea. See MPEP 2106.05(g) as cited above, and furthermore, MPEP 2106.04(a)(2) I, C., which provides examples of claims reciting mathematical concepts using data generated by physical measurements, including a relationship between reaction rate and temperature, which relationship can be expressed in the form of a formula called the Arrhenius equation, Diamond v. Diehr; 450 U.S. at 178 n. 2, 179 n.5, 191-92, 209 USPQ at 4-5 (1981).
The limitations for how the data is being analyzed to produce a diagnosis of estimated tumor fraction is part of the abstract idea, and thus cannot provide the inventive concept. The only additional elements in the claim include generic computer that is used to receive already generated sequence reads. As explained above, the courts have found the use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application. See MPEP 2106.05(f).
Examiner Comment:
If Applicant intends to argue a specific combination of whole-genome sequencing assays, including sequencing a baseline sample comprising a tumor or plasma sample, in addition to a normal PBMC sample, in combination with whole-genome sequencing of a follow-up sample, the claims should be amended to affirmatively recite sequencing steps that may be evaluated under Step 2B. Furthermore, it is noted that Applicant arguments repeatedly refer to the analysis of cell-free DNA fragments and making a diagnosis from the analysis of such fragments. However, the baseline sample may be a tumor sample and the follow-up sample may also just be a tumor sample, and thus the claims encompass estimating tumor fraction using tissue tumor samples. So any argued improvement relating to diagnoses based on cell-free DNA are not commensurate with the scope of the claims.
Claim Rejections - 35 USC § 103
The previous rejections of claims 1, 3, 6-7, 10-16, and 38-39 under 35 U.S.C. 103 as being unpatentable over Newman (2016) in view of Diehn (2014), as evidenced by Bratman (2014) in the Office action mailed 17 Sept. 2025 has been withdrawn in view of claim amendments and cancellations received 17 Feb. 2026.
The rejection of claims 35-36 under 35 U.S.C. 103 as being unpatentable over Newman (2016) in view of Diehn (2014) and Soo (2017), as applied to claim 1 above, further in view of O’Fallon (2013) in the Office action mailed 17 Sept. 2025 has been withdrawn in view of claim amendments received 17 Feb. 2026.
However, after further consideration, new grounds of rejection are set forth below in view of the claim amendments.
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1, 3, 6, 11-13, 15-18, and 37 are rejected under 35 U.S.C. 103 as being unpatentable over Newman (2016) in view of Diehn (2014) and Soo (2017), as evidenced by Bratman (2014). This rejection is newly recited and necessitated by claim amendment.
Cited reference:
Newman et al., Integrated digital error suppression for improved detection of circulating tumor DNA, 2016, nature biotechnology, 34(5), pg. 547-555 and supplementary (previously cited);
Diehn et al., US 2018/0251848 A1, effectively filed 12 Sept. 2014 (previously cited);
Soo et al., Comparison of Circulating Tumor DNA Recovery from Plasma and Serum, 2017, Blood blood, 130(Suppl 1): 2756, pg. 1-4 (previously cited); and
Newman et al. ( hereinafter Bratman), An ultrasensitive method for quantifying circulating tumor DNA with broad patient coverage, 2014, Nature Medicine, 20(5), pg. 548-554 and Suppl.) (newly cited).
Regarding claim 1, Newman discloses a method for detecting circulating tumor DNA (ctDNA) in a subject (Abstract) comprising the following steps.
Newman discloses (A) performing exome sequencing DNA of a blood sample (i.e. a baseline sample comprising a plasma sample) and paired germline (i.e. normal) sample (i.e. a first biological sample comprising a baseline and normal cell sample) from a patient (i.e. the subject) (pg. 547, col. 1, para. 1, e.g. sequencing of tumor biopsies; FIG. 1, e.g. cell-free DNA first blood draw pg. 548, col. 1, para. 4 to col. 1, para. 1;pg. 551, col. 2, para. 4, e.g. exome sequencing; pg. 554, col. 1, para. 4), to generate (i.e. receive) reads associated with a selector of genetic variants (i.e. markers) (pg. 551, col. 2, para. 4). Newman further discloses the reads are generated using 2x150 paired end sequencing (pg. 556, col. 1, para. 5), such that the reads are of a single base pair length.
Newman further discloses the normal germline cell sample comprises whole blood (ONLINE methods, Biological specimens, e.g. whole blood used for germline DNA isolation), which inherently comprises peripheral blood mononuclear cells, such that germ line mutations from peripheral blood mononuclear cells are removed during the removal of germline mutations in step (B) (ONLINE methods: ctDNA monitoring analysis, para. 1; ONLINE METHODS: Background polishing).
Newman discloses (B) excluding (i.e. filtering) germline SNPs from the sequencing data of the subject if they were present in any patient or control (ONLINE methods: ctDNA monitoring analysis, para. 1, e.g. positions with a germline SNP in any patient or control removed from variant list of patient), wherein the controls are healthy samples (ONLINE METHODS: Background polishing).
Newman discloses (C) sequencing a second follow-up blood sample of the subject (Figure 1, e.g. later blood draw) to identify tumor markers in the second sample (pg. 551, col. 2, para. 4; Figure 4c; pg. 553, col. 2, para. 3, e.g. selector-wide genotyping from exome sequencing of plasma for ctDNA monitoring).
Newman discloses (D) removing library preparation and sequencing errors (i.e. artefactual noise) by training a model of position-specific background distributions (i.e. employing a machine learning algorithm (pg. 549, col. 2, para. 2; ONLINE methods, Background polishing). Newman discloses the errors are removed from the sequencing data of the first and second reads (Fig. 2, e.g. error suppression; Fig. 1, e.g. reads from first and later blood draws) by a protocol comprising analyzing reads of a given barcode family (i.e. independent replicates of the same DNA template) and removing the less abundant mutations from consideration (i.e. discordant SNVs are removed) thereby producing a first and second filtered read set (Fig. 2, e.g. barcodes label reads originating from same DNA template; ONLINE methods, Analysis of molecular barcodes, para. 1-2, step 1, e.g. the frequency of the most abundant nonreference allele is used). Newman discloses the source of the removed errors is library preparation and sequencing (pg. 547, col. 1, para. 2).
Newman discloses (E) computing a circulating tumor DNA (ctDNA) percentage at time 0 (i.e. of the first biological sample) and after one month (i.e. for the second biological sample), using the reads from the corresponding blood draws (Fig. 1, e.g. reads analyzed from blood draws at different time points; FIG. 5(c)-(d), e.g. estimated %ctDNA; eONLINE methods, statistical methods for ctDNA detection).
Newman discloses the ctDNA percent at each time point for each sample is calculated by a mathematical model integrating a relationship between a number of SNVs with an alelle fraction > 0 (i.e. mutation load and detected mutations), a mean SNV fraction (i.e. a coverage, given an SNV fraction is a count of reads with a variant divided by the total coverage at the locus), the probability of observing a single tumor reporter (i.e. a number of detected mutations) (ONLINE methods, statistical methods for ctDNA detection, para. 1-4)
Newman discloses applying a model of the selector-wide background noise (i.e. a background noise model) to the ctDNA level of the mathematical model to determine if the level is statistically significant (i.e. computing a statistical significance framework using the noise and mathematical models) (ONLINE METOHDS, statistical methods for ctDNA detection, para. 3, e.g. tested the null hypothesis that the mean and s are not jointly above selector-wide background).
Newman discloses (F) detecting the ctDNA (i.e. residual disease) if the ctDNA level is above the detection limit at an empirical confidence level of 90% confidence (i.e. the statistical significance framework exceeds an empirical threshold) (ONLINE METOHDS, statistical methods for ctDNA detection, para. 2-3, e.g. detection limit = 0.12% at 90% confidence, statistical significance of ctDNA determined at a given confidence level, e.g. only levels with p value < 0.05 considered detectable).
Newman discloses providing a determination of residual disease when the ctDNA level is above the threshold (i.e. when residual is detected) (ONLINE METOHDS, statistical methods for ctDNA detection, para. 2-3; suppl. Figure 14, e.g. providing graph indicating statistically significant ctDNA detection).
Regarding the dependent claims:
Regarding claim 6, Newman further discloses the (B) excluding (i.e. filtering) germline SNPs from the sequencing data of the subject if they were present in any patient or healthy control samples (ONLINE methods: ctDNA monitoring analysis, para. 1; ONLINE METHODS: Background polishing), comprises generating a background database (i.e. a panel of normal blacklist) from the normal samples (Figure 6; ONLINE methods, Background polishing, para. 3)
Regarding claim 11, Newman further discloses in step (D), the correction of mutation errors generated by library preparation and sequencing (pg. 547, col. 1, para. 2) compares reads (i.e. independent replicates) originating from the same DNA fragment (Fig. 2, e.g. barcodes label reads originating from same DNA template; ONLINE methods, Analysis of molecular barcodes, para. 1-2, step 1, e.g. the frequency of the most abundant nonreference allele is used).
Regarding claim 12, Newman discloses the sequencing is 150 paired-end sequencing (ONLINE METHODS, Library preparation and sequencing), resulting in overlapping read pairs (i.e. R1 and R1 at overlapping positions) (Suppl. Figure 1a, e.g. Read #1 and Read#2 overlap). Newman discloses discordance between the paired-end reads is corrected back to the reference, such that a discordance between base calls of two reads is identified (Figure 2, e.g. pink PCR and sequencing error bars are removed).
Regarding claim 13, Newman further discloses correcting barcode families (i.e. duplication families) generating during sequencing and/or PCR (Fig. 2, e.g. a barcode represents the duplication of a particular nucleic acid fragment), wherein the consensus of reads (i.e. independent replicates) of each barcode family is checked to correct for artefactual mutations not present in a majority of the duplication family (ONLINE METHODS, Analysis of molecular barcodes, para. 2, e.g. at each position in a barcode family, the variant of the position is set to the most abundant non-reference allele, and barcode family consolidated to consensus sequence). Newman further discloses grouping the barcode families having identical barcodes and start/end coordinates after alignment to a reference genome (i.e. the duplication family is recognized by 5’ and 3’ similarity as well as alignment position) (ONLINE METHODS, Analysis of molecular barcodes, para. 2,).
Regarding claim 15¸ Newman further discloses the model of selector-wide background noise (i.e. the background noise model) uses the tumor-associated SNVs for cancer patients (i.e. a patient specific mutation signature) (Fig. 1, e.g. input cfDNA corresponds to the patient) to calculate (1) the selector-wide background μ and s from a cohort of healthy plasma samples (ONLINE METHODS: Background polishing, e.g. gaussian distribution determined from training cohort of healthy cfDNA controls; ONLINE METHODS, Biological specimens e.g. cfDNA is from plasma sample).
Regarding claim 16, Newman further discloses the model of selector-wide background noise (i.e. the background noise model) provides an estimated mean μ and standard deviation (ONLINE METHODS: Background polishing and Statistical analysis for ctDNA detection).
Regarding claims 1, 3, 17-18, and 37¸ Newman does not disclose the following:
First, regarding claim 3¸ Newman does not explicitly disclose a system comprising a processor (i.e. the claimed unit/engines). However, the court held that broadly providing an automatic or mechanical means to replace a manual activity which accomplished the same result is not sufficient to distinguish over the prior art. See MPEP 2144.04.
Further regarding claims 1 and 3, Newman does not disclose the first and second read sets and filtered first and second read sets span the genome, which is interpreted to mean the first and second filtered read sets are also genome-wide after removing errors.
Instead, Newman identifies the genetic markers and tumor-associated genetic markers from whole-exome sequencing data (pg. 551, col. 2, para. 4), but then filtering out regions from the reads less likely to have tumor markers to generate a targeted panel of regions (referred to as a selector) (pg. 550, col. 2, para. 2, e.g. regions that are highly recurrent in NSCLC; FIG. 1, e.g. CAPP-Seq). Newman discloses estimating the percentage of ctDNA and residual disease using tumor markers from the targeted read data (Figure 1), rather than from genome-wide read data.
However, Diehn also discloses the CAPP-Seq selector method of Newman for detecting residual disease (Abstract; Figure 4; FIG. 15A-E) and expands this method to other cancers, stating genomic regions recurrently mutated in a particular cancer can be used in a selector ([0128]; [0129]). Diehn further discloses that the number of genomic regions in a selector may vary depending on the nature of the cancer, and the inclusion of a larger number of regions may increase the likelihood a unique somatic mutation will be identified, but with a cost ([0128]). Diehn further discloses at the extreme, the entire genome of a tumor sample and genomic sample could be sequenced, and resulting sequences could be compared to not any differences with the non-tumor tissue (i.e. genome-wide read sets) ([0128]).
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 method of Newman to have utilized reads across the entire genome in predicting tumor fraction, as shown by Diehn ([0128]; FIG. 15A-E), such that the filtered read sets (i.e. after removing artefactual mutations/noise) are genome-wide. One of ordinary skill in the art would have been motivated to combine the methods of Newman and Diehn in order to increase the likelihood of unique somatic mutations being identified and thus the detection limit of residual disease, as shown by Diehn ([0034]; [0128]). This modification would have had a reasonable expectation of success given both Newman and Diehn disclose the generation and application of a selector for detecting residual disease, and Diehn discloses the number of regions in a selector may vary up to including all regions in a genome ([0128]). Diehn discloses the ctDNA detection limit (i.e. detection of residual disease) increases as the number of available tumor reporters increase ([0034]; FIG. 15A-F).
Furthermore, 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 used all regions in a genome in the selector (i.e. providing genome-wide filtered read sets) through routine experimentation of the number of regions in the selector, as shown by Diehn ([0128]) within the prior art conditions of increasing the number of regions to improve the likelihood of identifying somatic mutations and thus improve residual disease detection, as shown by Diehn ([0034]; [0128]) and decreasing the number of regions to reduce sequencing costs, as shown by Newman (pg. 548, col. 1, para. 4). See MPEP 2144.05 II. A.
Further regarding claims 1, 3, and 17¸ Newman in view of Diehn as applied above, does not disclose the one or more mathematical models are based on a fragment size shift associated with sequence reads
Further regarding claim 18 and 37¸ Newman in view of Diehn as applied above, does not disclose analyzing intra-subject fragment size shifts in a list of tumor-specific markers and random markers using statistical methods, wherein the statistical methods comprise one or more tests for significance.
However, Soo analyzes features upon which levels of ctDNA depend, such as disease burden and the biological source material (Abstract), and discloses that the distribution of fragment lengths of ct-DNA is shifted relative to the fragment lengths of non-tumor derived cell-free DNA (cfDNA) (Figure 1, e.g. mutated DNA in plasma is shorter than non-mutate DNA). Soo further discloses matched serum and plasma samples were analyzed (pg. 2, para. 2), such that the fragment size shifts are an intra-subject fragment size shifts, and discloses testing a statistical significance of size distributions (i.e. a statistical method including a test for significance ) for tumor mutations (i.e. tumor-specific markers) in addition to non-mutated (random) markers were analyzed (Figure 1; pg. 2, para. 4). Soo further discloses that ctDNA is a promising biomarker for noninvasive disease detection (Abstract).
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 method of Newman in view of Diehn to have analyzed intra-subject fragment size shifts using significant tests and used a fragment size shift feature associated with levels of ctDNA , as shown by Soo (Abstract; pg. 2, para. 4 Figure 1) in a mathematical model of Newman. One of ordinary skill in the art would have been motivated to combine the method of Newman in view of Diehn and Soo based on applying the known fragment size shift information of the DNA, determined by significant testing, as shown by Soo (Figure 1) to the known method for predicting levels of ctDNA based on cfDNA features of Newman, given Soo discloses that the levels of ctDNA depends on analytical features such as the distribution of fragment lengths of ct-DNA being different than the lengths of non-tumor derived cfDNA (Figure 1), such that the known technique of Soo is applicable to the method of Newman. Furthermore, one of ordinary skill in the art would have recognized that using fragment size shift features could have predictably been used to predict ctDNA levels, given Soo discloses the size of cfDNA fragments are associated with ctDNA levels (Abstract; Figure 1), and resulted in improved ctDNA level predictions.
Further regarding claims 1 and 3¸ Newman does not explicitly disclose the background noise model uses a z-score to assess statistical significance. However this limitation is inherent in Newman, as evidenced by Bratman.
Newman discloses the statistical significance of ctDNA quantification was performed as previously described, citing Bratman. Bratman discloses the statistical significance of ctDNA detection using indels as reporters was performed by calculating a Z-statistical, converted to a p-value, based on an indels fraction in a given plasma DNA sample against its fraction in every plasma DNA sample in the cohort (supplementary, pg. 14, para. 4). The calculation of and use of a Z-score model based on a distribution of indel fractions from plasma DNA samples is interpreted to read on “a background noise model”. Additionally, Newman discloses applying a model of the selector-wide background noise (i.e. also a background noise model) to the ctDNA level of the mathematical model to determine if the level is statistically significant (ONLINE METOHDS, statistical methods for ctDNA detection, para. 3, e.g. tested the null hypothesis that the mean and s are not jointly above selector-wide background).
Last, regarding claims 1 and 3, Newman further does not explicitly disclose generating an output indicative of whether adjusting a therapy is recommended for the subject based on the eTF exceeding the empirical threshold. It is noted this step is not required under the broadest reasonable interpretation of claim 1, as discussed in claim interpretation. However, in the interest of compact prosecution this limitation is discussed with the required programming of the contingent limitation in claim 3.
Regarding claims 1, and 3, Newman does disclose that the analysis of ctDNA is likely to play a major role in personalized cancer therapy by detecting minimal residual disease that is radiologically occult during disease surveillance (pg. 554, col. 1, para. 1), suggesting that the detection of residual disease should be used to determine a cancer therapy for a person in which residual disease is detected (i.e. provide an adjuvant therapy based on a stratification of whether residual disease is detected or not). Newman discloses providing a determination of residual disease when the ctDNA level is above the threshold (i.e. eTF above the threshold) (ONLINE METOHDS, statistical methods for ctDNA detection, para. 2-3; suppl. Figure 14, e.g. providing graph indicating statistically significant ctDNA detection).
Furthermore, Diehn discloses the method accurately quantifies cell-free tumor DNA from early and advanced tumor stages, and because tumor-derived DNA levels often parallel clinical responses to diverse therapies, the method may be used to detect tumor and facilitate personalized cancer therapy ([0080]). Diehn discloses the method includes recommending or not recommending a therapy (i.e. generating an output indicative of whether adjuvant therapy is recommended) based on the presence or absence of tumor cells ([0046]), and that the results of diagnosing can be used to classify patients into groups for administration of certain therapies ([0047]).
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 system of Newman to have further recommended a therapy adjustment to the subject based on the eTF exceeding a threshold (i.e. the detection of cancer), as suggested by Newman (pg. 554, col. 1, para. 1) and Diehn [0046]-[0047]; [0080]. One of ordinary skill in the art would have been motivated to modify the method of Newman and Diehn in order to treat the subject before the disease can be detected radiologically, as shown by Newman (pg. 554, col. 1, para. 1). This modification would have had a reasonable expectation of success given Newman suggests analyzing ctDNA plays a role in personalized cancer therapy, as discussed above.
Therefore the invention is prima facie obvious.
Claims 35-36 are rejected under 35 U.S.C. 103 as being unpatentable over Newman (2016) in view of Diehn (2014) and Soo (2017), as applied to claim 1 above, further in view of O’Fallon (2013). This rejection is newly recited and necessitated by claim amendment.
Cited reference: O’Fallon et al., A support vector machine for identification of single-nucleotide polymorphisms for next-generation sequencing data, 2013, Bioinformatics, 29(11), pg. 1361-1366 (previously cited).
Regarding claims 35-36, Newman in view of Diehn and Soo disclose the method of claim 1 as applied above.
Further regarding claims 35-36, Newman further discloses removing library preparation and sequencing errors in (D) by training a model of position-specific background distributions (i.e. employing a machine learning algorithm (pg. 549, col. 2, para. 2; ONLINE methods, Background polishing).
Further regarding claims 35-36, Newman in view of Diehn and Soo, as applied to claim 1 above, does not disclose filtering artefactual noise by employing a support vector machine.
However, O’Fallon discloses a method for identifying variants in next-generation sequencing data (Abstract), which uses a trained support vector machine to determine variants of sequence reads, allowing for errors produced by miscalled bases and platform-specific artifacts to be detected in a single statistical procedure (i.e. filtering artifactual noise from true variants) that can be improved as more training data is available (Abstract; pg. 1365, col. 2, para. 1-2).
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 method of Newman in view of Diehn and Soo to have employed a support vector machine to filter artefactual noise, as shown by O’Fallon (Abstract; pg. 1365, para. 1-2). One of ordinary skill in the art would have been motivated to combine the methods of Newman in view of Diehn and Soo with O’Fallon to allow the filtering of artefactual variants using a single statistical procedure that can be improved over time, as shown by O’Fallon (pg. 1365, col. 2, para. 1-2). This modification would have had a reasonable expectation of success given Newman also removes sequencing errors, and thus the method of O’Fallon is applicable to Newman.
Therefore, the invention is prima facie obvious.
Response to Arguments
Applicant's arguments filed 04 April 2025 regarding 35 U.S.C. 103 have been fully considered but they are not persuasive.
Applicant remarks that Newman does not teach genome-wide reads or genome-wide filtered read sets, and instead is built around a targeted selector approach, which is incompatible with genome-wide sequencing since it enriches certain genomic regions (Applicant’s remarks at pg. 21, para. 3 to pg. 22, para. 1). Applicant remarks the proposed combination would effectively replace Newman’s entire disclosure with a different approach, which is not a simple modification or optimization, but a fundamental change in the technical approach (Applicant’s remarks at pg. 23, para. 3).
This argument is not persuasive. Newman is not relied upon for teaching genome-wide reads and filtered read sets. Instead, Diehn makes obvious this limitation as applied in the above rejection. Furthermore, it is not persuasive that Newman is incompatible with genome wide sequencing reads. Diehn discusses the same Capp-seq method disclosed by Newman (Newman is also an author on Diehn). Diehn explicitly states that the number of genomic regions in a selector may vary depending on the nature of the cancer, and the inclusion of a larger number of regions may increase the likelihood a unique somatic mutation will be identified, but with a cost ([0128]). Diehn further discloses at the extreme, the entire genome of a tumor sample and genomic sample could be sequenced, and resulting sequences could be compared to not any differences with the non-tumor tissue (i.e. genome-wide read sets) ([0128]).
Given Diehn states that the selector method of Newman can be expanded to include more regions up to the entire genome, thus increasing the likelihood of detecting a somatic mutation, it is not persuasive that Newman is incompatible with using genome-wide reads. Nor does this change the fundamental approach of Newman, given Diehn explains that the general approach of detecting residual disease can use varying selectors with larger numbers of regions, up to the whole genome, depending on the nature of the cancer. The fundamental approach of using somatic variants to detect cancer would remain the same, and just a larger number of regions are being analyzed.
Applicant remarks Diehn does not cure the deficiencies of Newman because Diehn discloses the reference to whole-genome sequencing as an “extreme” does not constitute a teaching or suggestion that would motivate a skilled person to in the art to modify Newman (Applicant’s remarks at pg. 22, para. 2-5).
This argument is not persuasive. Diehn explicitly provides the motivation of using genome-wide reads of increasing the likelihood of detecting a somatic mutation ([0128]), which would improve the sensitivity of detecting circulating tumor DNA. The use of the word “extreme” in Diehn does not constitute a teaching away from using genome-wide reads.
Applicant remarks that converting Newman to genome-wide sequencing would require either maintaining a same total sequencing depth, resulting in low coverage over mutations, or would require maintaining coverage and increasing total sequencing output, which would be prohibitively expensive and impractical for clinical applications (Applicant’s remarks at pg. 22, para. 5 to pg. 23, para. 1).
This argument is not persuasive. Diehn describes the tradeoff of improving somatic mutation detection with increased sequencing costs, as described above ([0128]), and discloses that selector size impacts the cost and depth of sequencing coverage ([1022]). However, the instant claims do not require any particular sequencing coverage. Nor is there anything in the claims that requires a particular cost of sequencing. Even if the combination of Newman an Diehn utilized higher depth sequencing across a whole genome, this would still read on the claims.
Furthermore, while Applicant discusses that Newman’s high coverage allows for the detection of low variant allele frequencies as a reason for why lower depths of coverage cannot be used, Diehn discloses that more somatic variants can be detected using whole-genome reads ([0128]). Therefore, even if some low frequency variants would be missed due to lower coverage to offset sequencing costs, there would still be a reasonable expectation of success given Diehn discloses the likelihood of detecting a somatic mutation increases (e.g. due to a larger number of variants being screened in larger regions).
Applicant remarks that Diehn discloses expanding the number of regions is not without cost, and the Examiner does not address why a person skilled in the art would abandon a selector-based approach to a genome-wide approach (Applicant’s remarks at pg. 23, para. 2).
This argument is not persuasive. Diehn discloses both the advantages and disadvantages of using genome-wide reads as discussed above. An explicit motivation was provided in the above and previous rejection, of increasing the likelihood of unique somatic mutations being identified and thus the detection limit of residual disease, as shown by Diehn ([0034]; [0128]).
Applicant remarks the motivation to combine is based on hindsight because the reasoning assumes the person skilled in the art would abandon the approach of Newman using a small selector in favor of genome-wide sequencing despite the known tradeoffs, and thus the proposed combination appears to be the product of working backward from the claimed invention (Applicant’s remarks at pg. 23, para. 4 to pg. 24, para. 3).
This argument is not persuasive. In response to applicant's argument that the examiner's conclusion of obviousness is based upon improper hindsight reasoning, it must be recognized that any judgment on obviousness is in a sense necessarily a reconstruction based upon hindsight reasoning. But so long as it takes into account only knowledge which was within the level of ordinary skill at the time the claimed invention was made, and does not include knowledge gleaned only from the applicant's disclosure, such a reconstruction is proper. See In re McLaughlin, 443 F.2d 1392, 170 USPQ 209 (CCPA 1971).
For the reasons discussed above, Diehn does disclose a motivation for using whole-genome regions, and the rejection does not rely on any information gleaned from Applicant’s disclosure.
Conclusion
No claims are allowed.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KAITLYN L MINCHELLA whose telephone number is (571)272-6485. The examiner can normally be reached 7:00 - 4:00 M-Th.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Olivia Wise can be reached on (571) 272-2249. 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.
/KAITLYN L MINCHELLA/Primary Examiner, Art Unit 1685
Prosecution Insights