DETAILED ACTION
Applicant’s response, filed 25 Feb. 2026 and entered 26 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.
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 26 Feb. 2026 has been entered.
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 4-5, 7-8, 10, 14, 16-17, 19-21, 24, 26-27, 29, 31-32, 34, 36, and 38 are cancelled.
Claims 1-3, 6, 9, 11-13, 15, 18, 22-23, 25, 28, 30, 33, 35, and 37 are pending.
Claims 1-3, 6, 9, 11-13, 15, 18, 22-23, 25, 28, 30, 33, 35, and 37 are rejected.
Claims 1, 35, and 37 are objected to.
Priority
The effective filing date of the claimed invention is 30 Aug. 2019.
Drawings
The drawings filed 11 Nov. 2020 are objected to for the following reasons. This objection is previously stated.
The drawings are objected to as failing to comply with 37 CFR 1.84(p)(5) because they include the following reference character(s) not mentioned in the description:
#310-1 and 310-2 in FIG. 3
The drawings fail to comply with 37 CFR 1.84(u)(1), which states View numbers must be preceded by the abbreviation "FIG.". Therefore each figure label of “Figure 1”, “Figure 2A”, etc. should be relabeled “FIG. 1”, “FIG. 2A”, etc.
Corrected drawing sheets in compliance with 37 CFR 1.121(d), and/or amendment to the specification to add the reference character(s) in the description in compliance with 37 CFR 1.121(b) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
Response to Arguments
Applicant's arguments filed 25 Feb. 2026 regarding the drawings have been fully considered but they are not persuasive because they do not present any arguments.
Claim Objections
The objection to claims 1, 35, and 37 in the Office action mailed 27 Aug. 2025 has been withdrawn in view of claim amendments received 25 Feb. 2026.
Claims 1, 35, and 37 are objected to because of the following informalities. This objection is newly recited and necessitated by claim amendment:
Claims 1, 35, and 37 recite “…wherein transforming the sequencing dataset…into the feature tensor comprises: applying…function, encoding….position;”, which is a grammatical error and is missing an “and” between the applying and encoding steps. The claims should recite “applying…function, and encoding…”.
Appropriate correction is required.
Response to Arguments
Applicant's arguments filed 25 Feb. 2026 regarding the claim objections have been fully considered but they do not pertain to the newly recited objection set forth above.
Claim Interpretation
Claims 1, 35, and 37 recite “…causing sequencing of a biological sample of a subject, the sequencing being the parallel sequencing…”. Claims 35 and 37 recite “A non-transitory computer readable storage medium storing instructions…, when executed, causing the computer to perform steps…” and “A computer system, comprising: one or more processors; memory…causing the one or more processors to perform steps…”. Thus the claims encompass a genetic computer capable of performing the “causing sequencing….”. Therefore, the claims are interpreted to only require causing sequencing (e.g. transmitting data with instructions to sequence to a sequencer), rather than requiring the method comprises sequencing or the computer performs the step of sequencing, given this is within what the metes and bounds of what a computer can perform.
Claim Rejections - 35 USC § 112(b)
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, 9, 11-13, 15, 18, 22-23, 25, 28, 30, 33, 35, and 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.
Claims 1, 35, and 37, and claims dependent therefrom, are indefinite for recitation of “applying a discounted value by assigning each case read a discounted value based on its quality score rank, such that each additional base read for a given score is weighted according to a discount function”. First, it is unclear what set of base reads “each additional base read [for a given score]” is referring to, and in what way this is intended to be different than “each base read”. For example, it is not clear if “each additional base read” is referring to some subset, or all, of the base reads, a different set of base reads. It is further unclear what score “a given score” is referring to (e.g. the confidence score, quality score, discounted value, or the quality score rank?). Last, it is unclear if the phrase “such that each additional base read…is weighted according to a discount function” is simply reciting an intended use or result of applying a discounted value by assigning each read a discounted value, or if the claims intend to require weighting “each addition base read for a given score” according to a discount function (e.g. after assigning each base read a discounted value). It is further unclear if the “discounted value” after applying is intended to be the same “discounted value” assigned to each case read, or if the claims intend to apply a discounted value determined by assigning each read some other discounted value. Clarification is requested.
Claims 1, 35, and 37, and claims dependent therefrom, are indefinite for recitation of “encoding…the discounted value, the bag depth and the low bag depth flag for the base position”. The claims previously recite “1,000 or more base positions” and “a respective plurality of features for each respective base read of the plurality of base reads for the respective base position”. Therefore, it is unclear which base position, “the base position” is referring to, and as a result, which bag depth and low bag depth flag are being referenced. With further respect to “the discounted value”, the claims previously recite “applying a discounted value by assigning each base read a discounted value”. Therefore, it is further unclear which base read, “the discounted value “ is referring to given each base read was assigned a discounted value.
Claim 12 is indefinite for recitation of “…wherein the feature tensor represents a quantified distribution…and the quantified distribution comprises a discounted distribution that is determined by: ranking…”. Claim 1, as amended, now recites “transforming the sequencing dataset…into the feature tensor comprises: applying a discounted value…such that each additional base for a given score is weighted…, encoding, as elements of the feature tensor, the discounted value….”, which could be considered to be a “quantified distribution compris[ing] a discounted distribution”. It is unclear if: (1) claim 12 is intending to recite a separate process for getting a quantified distribution comprising a discounted distribution, than the process in claim 1, such that the limitations for how the ”discounted distribution” of claim 12 are still interpreted as a product by process limitation (see para. [0090] of previous Office action); or (2) claim 12 is further limiting how the feature tensor of claim 1 is determined, such that the claims actively require determining the quantified distribution. Clarification is requested. If Applicant intends (1), then the claim is further indefinite for “the discounted value” in line 11 because it is not clear if this is referring to the discounted value recited in claim 12 or the discounted value recited in claim 1. If Applicant intends (2), the claims should be amended to clearly link the wherein clause of claim 12 to the step of transforming the sequencing dataset into a feature tensor.
Claim Rejections - 35 USC § 112(d)
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claims 3 and 18 are rejected under 35 U.S.C. 112(d) as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends. This rejection is newly recited.
Claim 3 recites “…wherein the plurality of base positions is at least 10 base positions, at least 100 base positions, at least 1000 base positions…or at least 1 million base positions”. However, claim 1, from which claim 3 ultimately depends, already recites there are “1,000 or more base positions” (i.e. at least 10, at least 100, or at least 1000 base positions). Therefore, claim 3 fails to further limit the subject matter of claim 1, from which it depends.
Claim 18 recites “wherein the sequencing dataset further comprises one or more additional features from the group consisting of: a bag depth count of a total number of base reads at the respective base position, a first bag depth indicating when the total number of base reads is below a minimum threshold…”. However, claim 1, from which claim 18 depends, already recites “the respective plurality of features comprise:… a bag depth count of a total number of base reads at the respective base position, a low bag depth flag indicating when the total number of base reads is below a minimum threshold”. Claim 1 also already requires the sequencing data is transformed “into a feature tensor that represents a distribution of the plurality of features”, such that the transforming represents values of the above one or more additional features. Thus, claim 1 already requires the dataset comprises “one” of the features in the recited group of claim 18 (e.g. a bag depth count) and representing values of this feature elements of the feature tensor.
Applicant may cancel the claim(s), amend the claim(s) to place the claim(s) in proper dependent form, rewrite the claim(s) in independent form, or present a sufficient showing that the dependent claim(s) complies with the statutory requirements.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-3, 6, 9, 11-13, 15, 18, 22-23, 25, 28, 30, 33, 35, and 37 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Any newly recited portion herein 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, 35, and 37 being representative) is directed to a method, product, and system. 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, 35, and 37 recite the following steps which fall under the mathematical concepts and/or mental processes groupings of abstract ideas:
applying a machine-learning based (ML-based) base calling processing to determine consensus base calls for 1,000 or more base positions, wherein the ML-based base calling process comprises:
b) transforming the sequencing dataset of 1,000 or more sequence reads into a feature tensor that represents a distribution of the plurality of features for each respective base read in the sequencing dataset, wherein the feature tensor comprises a plurality of dimensions each corresponding to a feature of the plurality of features, wherein transforming the sequencing dataset of 1,000 or more sequence reads into a feature tensor comprises:
applying a discounted value by assigning each base read a discounted value based on its quality score rank, such that each additional base read for a given score is weighted according to a discount function,
encoding, as elements of the feature tensor, the discounted value, the bag depth and the low bag depth flag for the base position.
c) training a neural network for determining consensus base calls in the sequencing dataset of 1,000 or more sequence reads, wherein training the neural network comprises:
initiating weighting factors of the neural network;
generating training feature vectors from training data comprising a plurality of base reads for base positions,
selecting features from the training feature vectors based on a misclassification analysis,
analyzing distribution of the training feature vectors in a principal component analysis feature space;
transforming the training feature vectors to reduce features in the training feature vectors, wherein transformed training feature vectors comprise a principal component axis;
inputting the training feature vectors with selected features to an initialized neural network;
adjusting, in backpropagation, the weighting factors of the machine learning model;
d) assessing….the feature tensor that is transformed from the sequencing dataset of 1,000 or more sequence reads to determine a consensus base call for the respective base position, wherein the consensus base call comprises a predicted nucleotide base; and
generating, based on the ML-based calling process, a sequence of the biological sample as an end result of the sequencing.
The identified claim limitations falls into one of the groups of abstract ideas of mathematical concepts, mental processes, and/or certain methods of organizing human activity for the following reasons. First, the machine-learning based calling process to determine consensus base calls recites a mental process and mathematical concept as applied below regarding each of the various steps in this process. Next, the step of transforming the sequencing dataset into a feature tensor representing a distribution of a plurality of features, with a plurality of feature dimension each corresponding to a feature in the recited list of features can be practically performed in the mind by organizing the feature data in the sequencing data set into a tensor data structure (e.g. a two-dimensional matrix) with rows and columns representing two features. With respect to the sub-steps of applying a discounted value, the step of assigning each base read a discounted value according to a quality score rank can be practically performed in the mind by ordering numerical quality scores in order of rank, and an associated a score corresponding to a rank to a base read, and then multiplying a given score by some value from a discount function.
Furthermore, the steps pertaining to training the neural network can be practically performed in the mind as follows. Initiating weighting factors of the model involves assigning a number to each input variable of the model. Generating training feature vectors from training data comprising base reads involves organizing information from the training data into vectors corresponding to features of a respective given read of the training data. Selecting features from the vectors based on a misclassification analysis encompasses analyzing classification results of the neural network and selecting a combination of features that yielded the lowest misclassifications. Performing a principal component analysis on a distribution of values to transform a feature space to a reduced feature space can be practically performed in the mind aided by pen and paper by carrying out linear algebra techniques, including calculating eigenvalues and eigenvectors. Therefore, the above steps of training the neural network can be practically performed in the mind aided by pen and paper. Assessing the feature tensor to determine the consensus base call for the respective base position amounts to a mere analysis of data, which can be practically performed in the mind. Last, generating a sequence of the biological sample involves analyzing the base calls of the 1,000 base positions an arranging them into a sequence, which is a mental process. That is, other than reciting the steps are carried out by a processor, nothing in the claims precludes the steps from being practically performed in the mind. As such, these limitations recite a mental process. See MPEP 2106.04(a)(2).
Furthermore the limitations of weighting additional base reads for a given score according to a discount function, c) training a neural network by initiating weighting factors, selecting features using a misclassification analysis, inputting training feature vectors to an initialized neural network, analyzing a distribution of the training feature vectors in a principal component analysis, transforming the training feature vectors to reduce features, adjusting the weighting factors of the model in backpropagation further recite a mathematical concept. A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. See MPEP 2106.04(a)(2) C. In the instant case, weighting a score according to a discount function amounts to a textual equivalent of multiplying the score by a weight corresponding toa discount function. The limitations of c) training the neural network to determine consensus base calls amounts to a textual equivalent to performing mathematical calculations. That is, under the broadest reasonable interpretation of the claims in light of Applicant’s specification, the limitation of training the model encompass training the neural network using gradient descent or backward propagation, which involves inputting vectors containing numerical values the model, performing addition and multiplication to determine an output, adjusting weighting factors and performing backpropagation to minimize a loss function during the training (which involves iteratively repeating the application of the model using adjusted weighting factors), which amounts to a textual equivalent to performing mathematical calculations. Furthermore, performing a principal component analysis to on features to transform features into a transformed feature space is a mathematical dimensionality reduction technique that amounts to a textual equivalent of performing mathematical calculations. Therefore, these limitations recite a mathematical concept. See MPEP 2106.04(a)(2) I.
Dependent claims 2, 6, 9, 11-13, 15, 28, 30, and 33 further recite an abstract idea. Dependent claim 2 further recites the mental process of repeating the above identified abstract idea of claim 2. Dependent claim 6 further recites the mental process and mathematical concept of training and using the neural network to determine a model quality score for the consensus base call. Dependent claim 9 further recites the mental process of identifying a quality bin obtained during training of the neural network, and the mental process and mathematical concept of interpolating the model quality score based on an actual error rate to generate a recalibrated model quality score. Dependent claim 11 further recites the mental process of generating a feature tensor representing a quantified distribution of the plurality of features for each respective base read. Dependent claim 12, further recites the mental process of generating a feature tensor representing a quantified distribution of the plurality of features for each base read, the respective plurality of features comprising a respective nucleotide base and a respective read quality score, and the quantified distribution comprising a discounted distribution. Dependent claim 13 further recites the mental process and mathematical concept of computing a plurality of class probabilities based on the feature tensor, wherein the plurality of class probabilities correspond to a set of at least four classes, and determining the consensus base call based on a highest class probability. Dependent claim 15 further recites the mental process and mathematical concept of using a penalized logistic regression model to derive a plurality of model coefficients for training. Dependent claim 28 further recites the mental process of removing each base position that fails to satisfy a selection criterion of a measure of central tendency. Dependent claim 30 further recites the mental process removing, from the plurality of sequence reads, each base position in the plurality of base positions that satisfy a filtering criterion, wherein the filtering criterion is satisfied when the plurality of base reads comprises a threshold number of alternative nucleotides. Dependent claim 33 further recites the mental process of determining the sequence reads provide a coverage of 20x. Therefore, claims 1-3, 6, 9, 11-13, 15, 18, 22-23, 25, 28, 30, 33, 35, and 37 a 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.
Dependent claims 2, 6, 9, 11-13, 15, 28, 30, and 33 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, 35, and 37 include:
a computer system comprising at least one processor and a memory (claim 37);
a non-transitory computer readable storage medium (claim 35);
causing sequencing of a biological sample of sample of a subject, the sequencing being the parallel sequencing that results in varying sequencing depths at various base positions (i.e. transmitting data, as discussed above in claim interpretation);
a) obtaining a respective sequencing dataset of 1,000 or more sequence reads in a biological sample of a subject, the respective sequencing dataset corresponding to a plurality of base reads for a respective base position, in electronic form, wherein each base read of the plurality of base reads is captured in a respective sequence read among a plurality of sequence reads, wherein the respective sequencing dataset includes a respective plurality of features for each respective base read of the plurality of base reads for the respective base position, and wherein the respective plurality of features for each respective base read comprises at least two features selected from a feature group, the feature group comprising: a nucleotide base of the respective base read, a read quality score associated with the respective base read, a strand identifier indicating whether the respective sequence read of the respective base read corresponds to a forward or a reverse strand, a trinucleotide context based on three consecutive nucleotide bases including the respective base read, a confidence score associated with the trinucleotide context, a quality score for each baes read, a bag depth count denoting a total number of base reads at the respective base position, and a low bag depth flag indicating when the total number of base reads at a respective base position is below a minimum threshold (i.e. receiving data) (claims 1, 35, and 37);
assessing, with a neural network (claims 1, 35, and 37).
The additional element of claim 3 includes:
wherein the plurality of base positions is at least 10 base positions…(i.e. receiving data) (claim 3)
The additional element of claim 18 includes:
wherein the sequencing dataset further comprises one or more additional features selected from the group consisting of…(including the list of features) (i.e. receiving data).
The additional element of claim 22 includes:
…wherein the biological sample includes a cell-free nucleic acid molecule (i.e. receiving data).
The additional element of claim 23 includes:
…wherein the sequencing dataset is obtained from a methylation sequencing or a targeted sequencing (i.e. receiving data).
The additional element of claim 25 includes:
prior to obtaining (a) the sequencing dataset, obtaining a mapping string for each sequence read in a plurality of sequence reads, thereby obtaining a plurality of mapping strings, wherein each mapping string (i) is determined…and (ii) comprises….(including the full limitations) (i.e. receiving data).
The additional elements of at least one processor, memory, a non-transitory computer readable medium, receiving data, transmitting data, and using a neural network are generic computer components and/or merely recite the abstract idea as being performed on a computer. Claims 3, 18, and 22-23 only serve to further limit the data being received by claim 1 and thus are part of the step of receiving data. 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, reciting that the assessing is performed “with a neural network” provide nothing more than mere instructions to implement an abstract idea (i.e. assessing) on a generic computer (see MPEP 2106.05(f)), and furthermore, merely indicates a field of use or technological environment (i.e. neural networks) in which the judicial exception is performed
Last, the step of transmitting data to cause sequencing and obtaining/receiving the sequencing dataset only serve to collect information for use by the abstract idea, which amounts to insignificant extra-solution activity that does not integrate the recited judicial exception into a practical application. See MPEP 2106.05(g).
Therefore, the additionally recited elements merely invoke computers as a tool, link the judicial exception to a particular technological environment, and/or amount to insignificant extra-solution activity and, as such, the claims as a whole do no integrate the abstract idea into practical application. Thus, claims 1-3, 6, 9, 11-13, 15, 18, 22-23, 25, 28, 30, 33, 35, and 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 for the following reasons. Dependent claims 2, 6, 9, 11-15, 28, 30, and 33 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, 3, 18, 22-23, 25, 35, and 37 are outlined above.
The additional elements of at least one processor, memory, a non-transitory computer readable medium, receiving data, transmitting data, and using a neural network are conventional computer components and/or merely recite the abstract idea as being performed on a computer. Claims 3, 18, and 22-23 only serve to further limit the data being received by claim 1 and thus are part of the step of receiving data. 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 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 amount to significantly more than the above-identified judicial exception(s). 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 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 25 Feb. 2026 regarding 35 U.S.C. 101 have been fully considered but they are not persuasive.
Applicant remarks the claims involve a type of improvement that involves a specific way to train models and includes unique adjustments to parameters of machine learning models associated with sequencing, and that the claim does not use a generic neural network but instead defines (i) a structured feature set uniquely tailored to base calling in sequencing environments, (ii) a transformation of 1,000 or more sequence reads into a multi-dimensional feature tensor representing a distribution of these features, and (iii) a discounted weighting mechanism that assigns each base read a value based on its quality score (Applicant’s remarks at pg. 14, para. 7 to pg. 15, para. 2). Applicant further remarks the claims recite a specific training workflow including misclassification based feature selection, PCA in a feature space, dimensionality reduction, and backpropagation-based adjustment of weighting factors, which recite concrete adjustments to machine learning parameters and model inputs that are engineered to address uneven read depth and base position variability in sequencing, and under the Desjardins guidance, such recitations are directed to improving computer functionality (Applicant’s remarks at pg. 15, para. 2).
This argument is not persuasive. First, MPEP 2106.05(a) explains an indication that the claimed invention provides an improvement can include a discussion in the specification that identifies a technical problem and explains the details of an unconventional technical solution expressed in the claim, or identifies technical improvements realized by the claim over the prior art. It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements, alone or in combination with the judicial exception.
Furthermore, in Ex Parte Desjardins, the claims recited a specific way of training a machine learning model to learn knew tasks while protecting knowledge about previous tasks to overcome the problem of “catastrophic forgetting”. “Such improvements were tantamount to how the machine learning model itself would function in operation and therefore not subsumed in the identified mathematical calculation (see pg. 2, second bullet point of the Advance notice of change to the MPEP in light of Ex Parte Desjardins published 05 Dec. 2025).
In the instant case, the only details with how the neural network is actually trained is after the training features are input into the neural network, reciting “adjusting, in backpropagation, the weighting factors of the neural network”. Example 47 of the 2024 AI-related SME examples discusses that backpropagation is a mathematical calculation conventionally used for supervised learning of neural networks (see pg. 3, para. 2 and pg. 6, para. 2 of the AI SME examples). Unlike the claims in Ex Parte Desjardins, the instant claims do not describe a specific way of training a machine learning model to overcome a particular technological problem, such as catastrophic forgetting. Instead, the instant claims merely use a conventional “backpropagation” algorithm for training the neural network, which does not reflect any improved way of training a neural network that solves a technical problem. Furthermore, applying a standard principal component analysis to perform feature selection prior to inputting features into a neural network is part of the abstract idea, discussed above, and furthermore, does not specify details on how the neural network itself actually operates.
Regarding (i)-(iii) discussed by Applicant above, the creation of a structured feature set containing the desired features and discounted values are part of the abstract idea, and furthermore, only prepare input data for the neural network.
Overall, the claims do not recite additional elements that integrate the recited judicial exception into a practical application, including an improvement to technology.
Applicant remarks the ordered combination of claim limitations produce concrete, technical benefits in the operation of a sequencing analysis pipeline, beginning with parallel sequencing which results in varying sequencing depths at different base positions, which is a technical limitation, and the claimed transformation of raw sequencing reads into a feature tensor followed by neural-network assessment enables more accurate consensus base calls across high and low depth positions, which is not an abstract idea divorced from technology, but a technical enhancement to computational processing of sequencing data that improves performance and reliability of a sequencing system (Applicant’s remarks at pg. 15, para. 3 to pg. 16, para. 1). Applicant further remarks the claims are directed to a specific practical improvement in nucleic acid sequence analysis, particularly for cell-free DNA, and states weighted aggregation cannot be replicated by mental processes or conventional counting, at relates to a tangible, technical data processing that resolves a challenge in the art of molecular diagnostics and sequencing (Applicant’s remarks at pg. 16, para. 2).
This argument is not persuasive. MPEP 2106.05(a) states It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements, or by the additional elements in combination with the 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.
Here, Applicant’s improvement in consensus base calls, or the field of nucleic acid sequencing analysis, is an improved analysis in already generated sequence reads, as discussed above, which is an abstract idea and not a technology. In addition, as discussed above in claim interpretation, the claims do not even require any physical steps of sequencing a sample, let alone a cell-free DNA sample. Instead, the claims encompass a generic computer (as recited in claims 35 and 37) performing the steps of transmitting data (e.g. causing sequencing), receiving sequencing reads, and then analyzing the reads. Generating a consensus base call is not a “tangible” output, and is simply data. Furthermore, the clams do not affect how a sequencing system actually sequences a sample, such that performing and reliability of “a sequencing system” is improved, as merely alleged by Applicant.
In addition, performing “weighted aggregation” recites both mathematical concept and mental process. Regarding the mental process, it is not persuasive that the human mind is not practically capable of multiplying one number (i.e. a given score) by a weighting factor (e.g. 3*0.5 = 1.5), for each of different base positions or reads.
Last, reciting that the assessing is performed “with a neural network” provide nothing more than mere instructions to implement an abstract idea (i.e. assessing) on a generic computer (see MPEP 2106.05(f)), and furthermore, merely indicates a field of use or technological environment (i.e. neural networks) in which the judicial exception is performed, as discussed in the above rejection.
Applicant remarks the inventive method encodes both bag depth and low-depth flags as features used for neural network classification, and by feeding these features into the feature tensors, this powers the underlying neural network not only to assess bases but to distinguish and adapt to contexts with severely limited evidenced, and further remarks the training pipeline employs PCA and/or sparce PCA to select features (Applicant’s remarks at pg. 16, para. 3 to pg. 17, para. 1). Applicant further remarks the technical solution is rooted in the integration of data transformations, feature constructions and machine learning and thus is a significant improvement of sequencing analysis itself (Applicant’s remarks at pg. 17, para. 1).
This argument is not persuasive for the reasons already discussed above. The generating of the feature tensor and performing PCA for feature selection are part of the abstract idea, and not additional elements. The claims recite the mental process of assessing the feature tensor to determine a consensus base call, and simply links the abstract idea to the technological environment of neural networks, as discussed above. The claims do not recite any details regarding how the neural network is trained and/or operates that would clearly reflect an improvement to computer technology. Last, the alleged improvement in “sequencing analysis itself” is an improvement in the abstract idea itself, which is not an improvement to technology.
Claim Rejections - 35 USC § 103
The rejection of claims 1-3, 6, 11-13, 18, 25, 28, 30, 33, 35, and 37 under 35 U.S.C. 103 as being unpatentable over Luo (2019) in view of Zhang (2019) in the Office action mailed 27 Aug. 2025 has been withdrawn in view of claim amendments received 25 Feb. 2026.
The rejection of claim 15 under 35 U.S.C. 103 as being unpatentable over Luo in view of Zhang as applied to claim 1 above, and further in view of Kingma (2015) in the Office action mailed 27 Aug. 2025 has been withdrawn in view of claim amendments received 25 Feb. 2026.
The rejection of claims 22-23 are rejected under 35 U.S.C. 103 as being unpatentable over Luo in view of Zhang as applied to claim 1 above, and further in view of Namsaraev (2018) in the Office action mailed 27 Aug. 2025 has been withdrawn in view of claim amendments received 25 Feb. 2026.
Conclusion
No claims are allowed.
Claims 1-3, 6, 9, 11-13, 15, 18, 22-23, 25, 28, 30, 33, 35, and 37 are free of the prior art. Independent claims 1, 35, and 37 recite “the respective plurality of features comprise: a trinucleotide context…; a confidence score…, a quality score…, a bag depth count denoting a total number of base reads at the respective base position, and a low bag depth flag indicating when the total number of base reads at a respective base position is below a minimum threshold”. Claims 1, 35, and 37 further recite “transforming…into the feature tensor comprises: applying a discounted value by assigning each base read a discounted value based on its quality score rank…”.
The closest prior art of record, Luo, discloses using features including base mapping quality (i.e. a mapping quality of the base read) (pg. 7, col. 2, para. 2 to pg. 8, col. 1, para. 1), the 16 flanking base pairs on both sides of a candidate (i.e. a trinucleotide context based on three consecutive nucleotides including the respective base read), and the strength of a variant signal at each position of the 16 flanking base pairs including the candidate (i.e. a confidence score associated with the trinucleotide context) (pg. 8, col. 2, para. 5; Figure 4, e.g. see varying strengths for background noise and true variants). However, Luo does not disclose using the bag depth count and low bag depth flag indicating when depth is low as claimed, as additional features in the neural network. Furthermore, while Luo does disclose a process of applying discounted values to features to weight the features (pg. 7, col. 2, para. 2 to pg. 8, col. 1, para. 1, e.g. base quality imposed as a weight on the depth/read counts; pg. 8, col. 2, para. 6), Luo does not disclose applying discounted values based on its quality score rank as claimed.
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