DETAILED ACTION
Applicant’s response, filed 13 November 2025, 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 .
Claim Status
Claims 1-20 are pending and examined herein.
Claims 1-20 are rejected.
Priority
Claims 1-20 do not claim the benefit of priority to an earlier application. Thus, the effective filling date of claims 1-20 is 28 December 2021.
Information Disclosure Statement
The information disclosure statement (IDS) was 15 August 2025. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement has been considered by the examiner.
Specification
The objection to the disclosure for containing an embedded hyperlink in paragraph [0003] in Office action mailed 14 August 2025 is withdrawn in view of the amendment to the disclosure received 13 November 2025.
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.
The rejection below has been modified necessitated by amendment.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
(Step 1)
Claims 1-7 and 15-20 fall under the statutory category of a machine and claims 8-14 fall under the statutory category of a process.
(Step 2A Prong 1)
Under the BRI, the instant claims recite judicial exceptions that are an abstract idea of the type that is in the grouping of a “mental process”, such as procedures for evaluating, analyzing or organizing information, and forming judgement or an opinion. The instant claims further recite judicial exceptions that are an abstract idea of the type that is in the grouping of a “mathematical concept”, such as mathematical relationships and mathematical equations.
Independent claim 1 recites mental processes of “determining sequencing metrics for an initial genotype call from nucleotide reads…” and “determine a final genotype call for the multiallelic genomic coordinate…”.
Independent claim 1 recites a mathematical concept of “generate, utilizing a call recalibration machine learning model to process the sequencing metrics, a set of variant call classifications comprising a reference probability…”.
Independent claims 8 recites mental processes of “determining sequencing metrics for an initial genotype call…” and “determining a final genotype call indicating a haploid genotype…”.
Independent claim 8 recites a mathematical concept of “generate, utilizing a call recalibration machine learning model and based on the sequencing metrics, a first genotype probability of a haploid reference genotype, with respect to a reference genome, at the genomic coordinate, and a second genotype probability of a haploid alternate genotype…”.
Independent claim 15 recites mental processes of “determine, for one or more nucleotide reads, an initial genotype call indicating a homozygous reference genotype…”, “determine sequencing metrics for the initial genotype call…”, “determine a final genotype call comprising a variant call for the genomic coordinate…”.
Independent claim 15 recites a mathematical concept of “generate, utilizing a call recalibration machine learning model to process the sequencing metrics, for the initial genotype call a set of variant call…”.
Dependent claim 2 recites a mental process of “determine the final genotype call for the multiallelic genomic coordinate by modifying, based on the set of variant call classifications, an initial genotype call generated…”. Dependent claim 3 recites a mathematical concept of “generate updated genotype likelihoods…”. Dependent claim 5 recites a mathematical concept of “generate the reference probability…”. Dependent claim 6 recites a mathematical concept of “generate the differing genotype probability…”. Dependent claim 7 recites a mathematical concept of “generate the variant-call probability…”. Dependent claim 9 recites mathematical concepts of “generating the first genotype probability…” and “generating the second genotype probability…”. Dependent claim 10 recites mathematical concepts of “generating, for the genomic coordinate utilizing one or more layers of the call recalibration machine learning model, a first confidence score…” and “normalizing the first confidence score and the third confidence score…”. Dependent claim 10 recites “excluding the second confidence score…”. Dependent claim 12 recites mental processes of “converting a haploid reference genotype call generated by a call generation model to a diploid homozygous reference genotype call…” and “converting a haploid alternate genotype call generated by the call generation model to a diploid homozygous alternate genotype call…”. Dependent claim 12 recites a mathematical concept “generating, utilizing the call recalibration machine learning model, the first genotype probability and the second genotype probability…”. Dependent claim 13 recites mental processes “down sampling diploid sequencing metrics to simulate haploid sequencing…”, “selecting a subset of diploid nucleotide reads…”, and “selecting, based on nucleotide base calls of the subset of diploid nucleotide reads, a subset of genomic coordinates…”. Dependent claim 14 recites a mental process of “selecting, to simulate the haploid sequencing metrics corresponding to the haploid nucleotide sequence…”. Dependent claim 16 recites a mental process of “determine the variant call for the genomic coordinate by modifying…”. Dependent claim 17 recites a mental process of “determine the sequencing metrics by…”. Dependent claim 18 recites mental processes of “identify a previous homozygous reference genotype call…”, “identify a ground truth base call…”, and “modify the call recalibration machine learning model based on a comparison of the variant call…”. Dependent claim 19 recites a mental process and mathematical concept of “generate the set of variant call classifications, including a homozygous reference classification indicating a probability… a homozygous alternate classification indicting a probability of a homozygous alternate genotype… a heterozygous genotype classification indicating a probability of a heterozygous genotype…”. Dependent claim 20 recites a mental process of “update one or more of a call quality field…”.
The claims recite analyzing/evaluating data, organizing data, and making observations/ judgments of determining sequencing metrics, determine final nucleotide base calls based on classifications, modifying quality metrics, updating genotype likelihoods, determining the indication of homozygosity from nucleotide bases, excluding confidence scores, determining haploid genotypes, converting haploid genotype call to a diploid genotype call, selecting subset of reads from diploid data, determining variant call by modification, determining genotypes for genomic coordinates, and updating information. The human mind is capable of analyzing/ evaluating data, organizing data, and making observations/ judgments about data. Further the claims recite mathematical concepts of generating probabilities using a machine learning model (which encompasses models such as support vector machines, a linear regression or a logistic regression see instant disclosure [0064]) these models take in numerical input them perform mathematical calculations and produce numerical output. Thus, the steps of determining probabilities are mathematical calculations. Dependent claim 4 further limit the mental process/mathematical concept recited in the independent claim but do not change their nature as a mental process/mathematical concept.
(Step 2A Prong 2)
Claims found to recite a judicial exception under Step 2A, Prong 1 are then further analyzed to determine if the claims as a whole integrate the recited judicial exception into a practical application or not (Step 2A, Prong 2). Integration into a practical application is evaluated by identifying whether there are any additional elements recited in the claim and evaluating those additional elements to determine whether they integrate the exception into a practical application.
The additional elements in claims 1 and 8 of using a generic computer to perform judicial exceptions and the additional element in claim 15 of a non-transitory computer readable medium hold instructions when executed causes a generic computer to perform judicial exceptions does not integrate the judicial exceptions into a practical application because these additional elements interact with the judicial exceptions in a manner where the generic computer environment is a tool to perform the judicial exceptions. Therefore, these additional elements amount to applying the judicial exceptions to a generic computer environment without an improvement to computer functionality.
The additional element in claim 3 of generate a variant call file that includes the updated genotype likelihoods does not integrate the judicial exceptions into a practical application because these additional elements interact with the judicial exceptions in a manner of outputting the results of the judicial exceptions. Therefore, these additional elements amount to adding insignificant extra solution activity of outputting data. The data being input into the variant call file does not change the additional element of generating a variant call file.
The additional element in claim 16 of receive, from a call generation model, an indication of the homozygous reference genotype at the genomic coordinate does not integrate the judicial exceptions into a practical application because this additional element interacts with the judicial exceptions in a manner of providing data to the judicial exceptions. Therefore, this additional element amount to adding insignificant extra solution activity of data gathering data. The data being gathered does not change the step of receiving data.
Thus, the additional elements do not integrate the judicial exceptions into a practical application and claims 1-20 are directed to the abstract idea.
(Step 2B)
Claims found to be directed to a judicial exception are then further evaluated to determine if the claims recite an inventive concept that provides significantly more than the judicial exception itself (Step 2B). The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because:
The additional element in claims 1 and 8 of using a generic computer to perform judicial exceptions and the additional element in claim 15 of a non-transitory computer readable medium hold instructions when executed causes a generic computer to perform judicial exceptions are conventional see MPEP 2106.05(b) and MPEP 2106.05(d)(II).
The additional element in claim 3 of generate a variant call file that includes the updated genotype likelihoods is conventional as shown by DePristo et al. (US 10,354,747 B1; cited in IDS 17 August 2023; previously cited) in col. 16 generating a VCF file and in Ramachandran et al. (BMC Bioinformatics 22, 404 (2021); previously cited) on page 27 generating a VCF file.
The additional element in claim 16 of receive, from a call generation model, an indication of the homozygous reference genotype at the genomic coordinate is conventional see MPEP 2106.05(b) and MPEP 2106.05(d)(II). The data being gathered does not change the step of receiving data.
Thus, the additional elements are not sufficient to amount to significantly more than the judicial exception because they are conventional.
Response to Arguments
Applicant's arguments filed 13 November 2025 have been fully considered but they are not persuasive.
Applicant argues claim 1 improves the computing speed and accuracy of genotype recalibration for genotype calls at multiallelic genomic coordinates by repurposing sequencing metrics into specialized variant-call classifications and uses them to improve speed and accuracy of genotype recalibration for multiallelic sites (Reply p. 15-16). Applicant further argues the claim integrates any recite mathematical operations into a practical application that improves the functioning of a computer-implemented variant calling at multi-allelic genomic coordinates (Reply p. 16).
This argument has been fully considered but found to be not persuasive. The MPEP states at 2106.05(a) “It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements… In addition, the improvement can be provided by the additional element(s) in combination with the recited judicial exception”. The determination of an improvement to technology has two steps, the identification of additional elements (which define the technology) and the evaluation of the additional elements to determine if the improvement is provided by or realized in the additional elements either by the additional elements themselves or the additional element in combination with the judicial exception (i.e. the interaction between the judicial exceptions and the additional elements). In the instant case, claim 1 recites the additional element of a generic computer (i.e., a system comprising at least one processor and non-transitory computer readable medium) to perform judicial exceptions. The additional element of the computer only interacts with the judicial exceptions in a manner of being utilized as a tool to perform abstract ideas. It is noted that variant calling at multiallelic genomic coordinates falls under the abstract idea. Although the abstract idea of variant calling at multiallelic genomic coordinates may be more computationally efficient to perform and more accurate, the speed and accuracy comes solely from the abstract idea itself and not from the computer functioning in a different/improved manner. Thus, the judicial exceptions in claim 1 are not integrated into a practical application because the additional element of the generic computer does not provide the improvement and the combination of the generic computer and abstract ideas do not provide an improvement which is realized in the functionality of the computer itself.
Applicant argues that claim 8 improves the computing speed and accuracy of Genotype Recalibration for Genotype calls at a Haploid Genomic Coordinate by being directed to a targeted improvement in haploid variant-calling workflows, not to generic probability calculations (Reply p. 16-17). Applicant argues that claim 8 applies a particular machine-learning model to real-world sequencing reads to generate and use specialized variant call classifications in a manner that increases the speed and accuracy of haploid genotype determination and uses AI to improve the detection of physical/biological conditions through enhanced model performance (Reply p. 17).
This argument has been fully considered but found to be not persuasive. The MPEP states at 2106.05(a) “It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements… In addition, the improvement can be provided by the additional element(s) in combination with the recited judicial exception”. The determination of an improvement to technology has two steps, the identification of additional elements (which define the technology) and the evaluation of the additional elements to determine if the improvement is provided by or realized in the additional elements either by the additional elements themselves or the additional element in combination with the judicial exception (i.e. the interaction between the judicial exceptions and the additional elements). In the instant case, claim 8 recites the additional element of using a generic computer to perform judicial exceptions. The additional element of the computer only interacts with the judicial exceptions in a manner of being utilized as a tool to perform abstract ideas. It is noted that machine learning model itself falls under the abstract idea because it encompasses mathematical calculations (the machine learning model itself encompasses a linear regression or a logistic regression see instant disclosure [0064]) this machine learning model takes in numerical input (sequencing metric data) to perform a mathematical calculation a linear regression using a linear regression equation (or logistic regression using a logistic regression equation) and to produce numerical output (a probability). Although the abstract idea of genotype recalibration of sequencing data may be more computationally efficient to perform and more accurate, the speed and accuracy comes solely from the abstract idea itself and not from the computer functioning in a different/improved manner. Further, due to the machine learning model encompassing a linear regression model (i.e., a linear regression equation) or a logistic regression model (i.e., a logistical regression equation), the use of these models for improving detection encompasses improving abstract mathematical calculations to improve sequence data analysis. Thus, the judicial exceptions in claim 8 are not integrated into a practical application because the additional element of the generic computer does not provide the improvement and the combination of the generic computer and abstract ideas do not provide an improvement which is realized in the functionality of the computer itself.
Applicant argues claim 15 improves the computing speed and throughput of genotype recalibration for initial genotype calls indicating homozygous reference at genomic coordinates (Reply p. 17). Applicant argues that the steps of the claim makes up a computer-implemented recalibration pipeline that corrects false homozygous-reference calls by using learned variant call classifications faster than existing sequencing systems (Reply p. 17). Applicant argues the steps cannot be reasonably performing as a mental process given the volume of sequencing data and the recited machine-learning model and is not directed merely to mathematical concepts because it represents a technical solution to the technical problem of misclassified homozygous reference coordinates in high-throughput sequencing systems (Reply p. 18).
This argument has been fully considered but found to be not persuasive. The MPEP states at 2106.05(a) “It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements… In addition, the improvement can be provided by the additional element(s) in combination with the recited judicial exception”. The determination of an improvement to technology has two steps, the identification of additional elements (which define the technology) and the evaluation of the additional elements to determine if the improvement is provided by or realized in the additional elements either by the additional elements themselves or the additional element in combination with the judicial exception (i.e. the interaction between the judicial exceptions and the additional elements).
In the instant case, claim 15 recites the additional element of a non-transitory computer readable medium hold instructions when executed causes a generic computer to perform judicial exceptions. This additional element of interacts with the judicial exceptions in a manner of being utilized as a tool to perform abstract ideas. It is noted that the steps of the pipeline fall under the abstract idea. It is further noted that under the BRI of the claim the sequencing data encompasses one nucleotide read (even if the data volume is large the human mind is still capable of making determinations utilizing sequencing data) and the machine-learning model encompasses performing a calculation using a linear regression equation (or logistic regression equation) (which are mathematical calculations). Therefore, the solution to the problem of misclassified homozygous reference coordinates in high-throughput systems is an abstract process of analyzing sequencing data. Although the abstract idea of genotype recalibration for initial genotype calls indicating homozygous reference at genomic coordinates may be more computationally efficient to perform and more accurate, the speed and accuracy comes solely from the abstract idea itself and not from the computer functioning in a different/improved manner. Thus, the judicial exceptions in claim 15 are not integrated into a practical application because the additional element does not provide the improvement and the combination of the additional element and abstract ideas do not provide an improvement which is realized in the functionality of the computer itself.
Claim Rejections - 35 USC § 103
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.
The rejection below has been modified necessitated by amendment.
Claims 1-7 are rejected under 35 U.S.C. 103 as being unpatentable over DePristo et al. (US 10,354,747 B1; previously cited) in view of Lam et al. (US 20210257050 A1; previously cited).
Independent claim 1 determine sequencing metrics for nucleotide base calls of nucleotide reads corresponding to a multiallelic genomic coordinate of a sample nucleotide sequence
DePristo et al. shows determining sequence metrics such as information about sample read data, identified alleles, and variant types for nucleotide base calls for a region (DePristo et al. col. 13-14). DePristo et al. shows that two or more candidate alleles may be identified for a particular nucleotide position in the genome and implicitly shows that the genomic coordinate can be a multiallelic site containing three different alleles (DePristo et al. col. 14).
generate, utilizing a call recalibration machine learning model and based on the sequencing metrics, a set of variant call classifications
DePristo et al. shows utilizing a neural network for determining if a candidate variant is a true mutation based on the sequencing metrics (DePristo et al. col. 14-15).
DePristo et al. does not show that variant call classification comprising a reference probability of a homozygous reference genotype, with respect to a reference genome, at the multiallelic genomic coordinate, a differing genotype probability of a genotype error at the multiallelic genomic coordinate, and a variant-call probability of a variant call genotype, with respect to the reference genome, being correct at the multiallelic genomic coordinate or determining a final genotype call for the multiallelic genomic coordinate based on these variant call classifications
Like DePristo et al., Lam et al. shows utilizing a neural network model that intakes sequence metrics and candidate variants to output probabilities of predictions. Lam et al. shows utilizing a neural network model that intakes candidate variants and outputs probability values being NONE (no variant, i.e., a false-positive call), SNP/SNV (single nucleotide polymorphism/variant), INS (insertion variant), and DEL (deletion variant) (Lam et al. [0157]). Lam et al. further shows that the output of the neural network can also output a predicted genotype association with the variant such as probabilities of a homozygous reference (a non-variant), a heterozygous variant (where only one of maternal or paternal copies has a variant, a homozygous variant (where both copies have the same variant, or other (where each copy has a different variant) and that these outputs can be omitted and/or additional outputs can be added (Lam et al. [0160]). Therefore, this shows a machine learning model that outputs a reference probability of a homozygous reference (the homozygous reference probability), a differing genotype probability of a genotype error (the probability of a false-positive call), and a variant-call probability of a variant call genotype, with respect to the reference genome, being correct (the probability that the candidate variant which was initially identified is a heterozygous variant (where only one of maternal or paternal copies has a variant, a homozygous variant (where both copies have the same variant, or other (where each copy has a different variant)).
and determine a final genotype call for the multiallelic genomic coordinate based on the set of variant call classifications.
Lam et al. shows based on these probability values it can be determined which type of variant is the most likely, as well as the confidence level of the particular prediction (Lam et al. [0157] and [0160]).
Dependent claim 2 is directed to determine the final genotype call for the multiallelic genomic coordinate by modifying, based on the set of variant call classifications, an initial genotype call generated by a call generation model for a variant call file.
DePristo et al. shows the variant data set (which contains quality values and genotype likelihoods) may be used to iteratively realign sequence reads to clean up ambiguous regions to improve alignment confidence or the accuracy of allele determination (DePristo et al. col. 14-16). DePristo et al. shows the variant data set is a proto data object and that the proto data moves through the pipeline, information within the proto data may be removed, added, or modified (DePristo et al. col. 14). DePristo et al. shows the realignment pipeline generates a VCF to encode variant calling information into a data file (DePristo et al. col. 16).
Dependent claim 4 is directed to wherein the sequencing metrics for the initial genotype call include one or more of: read-sequencing metrics comprising sequencing metrics derived from nucleotide reads of the sample nucleotide sequence, externally sources metrics comprising sequencing metrics stored in one or more databases external to the system or call model generated sequencing metrics comprising sequencing metrics generated by a call generation model.
Lam et al. shows an output from the network may be a heterozygous variant (which is a mixture of bases at a variant location) (Lam et al. [0160]). Lam et al. based on the probabilities from the network it can be determined which type of variant is most likely (Lam et al. [0157]. It would have been obvious to one of ordinary skill in that art that in the case of heterozygosity two nucleotide bases will be predicted.
Dependent claim 5 is directed to generate the reference probability by determining a probability that a genotype at the multiallelic genomic coordinate is a homozygous genotype with respect to a reference genome.
Lam et al. further shows that the output of the neural network can also output a predicted genotype association with the variant such as probabilities of a homozygous reference (a non-variant) which is in respect to a reference (Lam et al. [0160]).
Dependent claim 6 is directed to determining a probability that a predicted variant call genotype initially determined with respect to the reference genome by a call generation model for the multiallelic genomic coordinate is an incorrect genotype or an incorrect allele.
Lam et al. shows utilizing a neural network model that intakes candidate variants and outputs a probability value being NONE (no variant, i.e., a false-positive call) (Lam et al. [0157]).
Dependent claim 7 is directed to generate the variant-call probability by determining a probability that a predicted variant-call genotype initially determined with respect to the reference genome by a call generation model for the multiallelic genomic coordinate is a correct genotype.
Lam et al. shows utilizing a neural network model that intakes candidate variants with sequencing metrics with respect to a reference genome and outputs probability values being SNP/SNV (single nucleotide polymorphism/variant), INS (insertion variant), and DEL (deletion variant) (Lam et al. [0155] and [0157]) (the probability that the candidate variant which was initially identified is a SNP/SNV, insertion, or deletion). Lam et al. further shows that the output of the neural network can also output a predicted genotype association with the variant such as a heterozygous variant (where only one of maternal or paternal copies has a variant, a homozygous variant (where both copies have the same variant, or other (where each copy has a different variant) and that these outputs can be omitted and/or additional outputs can be added (Lam et al. [0160]).
An invention would have been obvious to one or ordinary skill in the art if some motivation in the prior art would have led that person to modify reference teachings to arrive at the claimed invention. It would have been obvious to one of ordinary skill in the art before the effective filling date to have modified the neural network for performing variant classifications of DePristo et al. with the neural network which produces several predictions such as reference probability of a homozygous reference (the homozygous reference probability), a differing genotype probability of a genotype error (the probability of a false-positive call), and a correct variant probability (the probability that the candidate variant which was initially identified is a SNP/SNV, insertion, deletion, a heterozygous variant (where only one of maternal or paternal copies has a variant, a homozygous variant (where both copies have the same variant, or other (where each copy has a different variant)) because this allows for a neural network which can produce several probability outputs capturing different aspects from candidate variants and thus provides a comprehensive analysis on the candidate variant to provide a final variant genotype call (Lam et al. [0157]-[0160]). One would have a reasonable expectation of success because DePristo et al. shows formatting data compatible for being analyzed by a neural network for variant calling while Lam et al. provides a neural network model in which several predictions can be obtained for variant calling.
The rejection below has been modified necessitated by amendment.
Claims 8-14 are rejected under 35 U.S.C. 103 as being unpatentable over Ramachandran et al. (BMC Bioinformatics 22, 404 (2021); previously cited) in view of Heller et al. (Bioinformatics, Volume 36, Issue 22-23, December 2020, Pages 5519–5521; previously cited).
Claim 8 is directed to determining sequencing metrics for an initial genotype call from of nucleotide reads corresponding to a genomic coordinate of a haploid nucleotide sequence from a sample;
Ramachandran et al. shows determining sequence metrics such as the read base, the reference base, the base quality, the mapping quality of the read, the strand identity of the read, and a position flag of a genomic coordinate which is encoded into a vector to be processed by a deep learning model (Ramachandran et al. page 24-25 and figure 10).
generating, utilizing a call recalibration machine learning model to process the sequencing metrics, a set of variant call classifications comprising a first genotype probability of a haploid reference genotype, with respect to a reference genome, at the genomic and a second genotype probability of a haploid alternate genotype, with respect to the reference genome, at the genetic coordinate,
Ramachandran et al. shows a convolutional neural network which intakes these sequencing metrics to produce probabilities that the alleles are true given the evidence for one allele to the allelic evidence of all other alleles and shows haplotag values which is utilized in the prediction and shows utilizing haplotype pairs when making predictions (Ramachandran et al. pages 24-27 sections “DNN architecture” and “Single-platform architecture”, page 28 “determining hotspots”, and page 29 “Preparing training data”). Ramachandran et al. shows that the probabilities are based on the alleles that pass an initial screening and may be only two genotype probabilities (Ramachandran et al. page 28 “determining hotspots”).
And determining a final genotype call indicating a haploid genotype for the genomic coordinate based on the first genotype probability and the second genotype probability coordinate and a second genotype probability of a second genotype at the genomic coordinate
Ramachandran et al. shows determining the final base call based on the generated probabilities of true alleles (Ramachandran et al. page 27 with in the section of “Single-platform architecture”).
Ramachandran et al. does not explicitly show a genomic coordinate of a haploid nucleotide sequence from a sample
Like Ramachandran et al., Heller et al. shows using sequencing reads to predict variants and genotypes. Heller et al. shows a pipeline to call structural variant candidates using haploid assembly alignments and diploid assembly alignments which are paired haplotypes based on similarity (Heller et al. Supplementary material Fig. S2).
Claim 9 is directed to generating the first genotype probability comprises utilizing a layer of the call recalibration machine learning model to modify a homozygous reference probability of a homozygous reference genotype at the genomic coordinate to generate a haploid reference probability of a reference genotype at the genomic coordinate and generating the second genotype probability comprises utilizing the layer of the call recalibration machine learning model to modify a homozygous alternate probability of a homozygous alternate genotype at the genomic coordinate to generate a haploid alternate probability of an alternate genotype at the genomic coordinate.
Ramachandran et al. in view of Heller et al. shows a convolutional neural network which intakes sequencing metrics from aligned reads to assemblies to produce probabilities that the alleles are true given the evidence for one allele to the allelic evidence of all other alleles and shows haplotag values which is utilized in the prediction and shows utilizing haplotype pairs when making predictions (Ramachandran et al. pages 24-27 sections “DNN architecture” and “Single-platform architecture”, page 28 “determining hotspots”, and page 29 “Preparing training data”). It would have been obvious that these probabilities are modified probabilities that are based on haploids produced by the haploid assembly. Further, since Ramachandran et al. utilizes reference bases and read base information the probabilities may include a reference probability and a alternate allele probability when determining the true allele ate a genomic coordinate.
Claim 10 is directed to wherein generating the first genotype probability and the second genotype probability comprises: generating, for the genomic coordinate utilizing one or more layers of the call recalibration machine learning model, a first confidence score corresponding to a first genotype, a second confidence score corresponding to a second genotype, and a third confidence score corresponding to a third genotype;
Ramachandran et al. shows generating probabilities which are the confidence in a particular allele identified in the sequencing data if an allele passes an initial threshold those reads will be processed for prediction and shows haplotag values which is utilized in the prediction and shows utilizing haplotype pairs when making predictions (Ramachandran et al. pages 24-27 sections “DNN architecture” and “Single-platform architecture”, page 28 “determining hotspots”, and page 29 “Preparing training data”).
excluding the second confidence score corresponding to the second genotype; and normalizing the first confidence score and the third confidence score utilizing a SoftMax model to generate the first genotype probability and the second genotype probability.
Ramachandran et al. shows passing these probabilities (or confidence values) through a set of equations that returns a value of 1 for true alleles and 0 for false alleles based on the probabilities (Ramachandran et al. page 27 in “Single-platform architecture” section). This shows excluding some genotypes while normalizing the true values as 1 based on the probabilities.
Claim 11 is directed to determining the final genotype call indicating the haploid genotype for the genomic coordinate comprises determining one of: the haploid alternate genotype for the genomic coordinate, a modified base call quality metric, a modified genotype metric, and a modified genotype quality metric based on determining that the second genotype probability exceeds the first genotype probability; or the haploid reference genotype for the genomic coordinate, a modified base call quality metric, and a modified genotype quality metric based on determining that the first genotype probability exceeds the second genotype probability.
Ramachandran et al. in view Heller et al. of shows a convolutional neural network which intakes these sequencing metrics derived from haplotype realignment and base quality score recalibration to produce probabilities that the alleles are true given the evidence for one allele to the allelic evidence of all other alleles and shows haplotag values which is utilized in the prediction and shows utilizing haplotype pairs when making predictions (Ramachandran et al. pages 24-27 sections “DNN architecture” and “Single-platform architecture”, page 28 “determining hotspots”, and page 29 “Preparing training data”). Ramachandran et al. shows the output may identify which allele is true based on a comparison of one allele to other alleles from assemblies these assemblies carry base call quality metrics which can be modified using GATK indel realigner (Ramachandran et al. page 21 “Use of GATK indel realigner”), modified genotype type metric (label of true), and the modified genotype quality (probability of genotype output from the model).
Claim 12 is directed to converting a haploid reference genotype call generated by a call generation model to a diploid homozygous reference genotype call as an input for the call recalibration machine learning model; or converting a haploid alternate genotype call generated by the call generation model to a diploid homozygous alternate genotype call as an input for the call recalibration machine learning model; and generating, utilizing the call recalibration machine learning model, the first genotype probability and the second genotype probability based further on the diploid homozygous reference genotype call or the diploid homozygous alternate genotype call.
Ramachandran et al. in view of Heller et al. shows converting a haploid alternate genotype call to a diploid homozygous alternative genotype call to predict candidate variant candidates (Heller et al. Supplementary materials Fig. S2) which can be used as inputs to the neural network model to generate probabilities of candidate alleles. Ramachandran et al. in view of Heller et al. shows haplotag values which is utilized in the prediction and shows utilizing haplotype pairs when making predictions (Ramachandran et al. page 24-27 sections “DNN architecture”, “Single-platform architecture” and “Preparing training data”).
Claim 13 is directed to down sampling diploid sequencing metrics to simulate haploid sequencing metrics corresponding to the haploid nucleotide sequence by: selecting a subset of diploid nucleotide reads from the sample to simulate haploid nucleotide reads; and selecting, based on nucleotide base calls of the subset of diploid nucleotide reads, a subset of genomic coordinates exhibiting homozygous reference genotypes or homozygous alternate genotypes as indicated by a call generation model or as indicated by a ground-truth base-call dataset.
Heller et al. shows that the diploids are paired signatures containing similar structural variation and contain genotypes as initially called by a model (Heller et al. Supplementary materials Fig. S2 and page 5520 left col.). It would have been obvious to one of ordinary skill in the art that these diploid pairs may be genotyped and then split into single assemblies for further processing.
Claim 14 is directed to selecting, to simulate the haploid sequencing metrics corresponding to the haploid nucleotide sequence, the diploid sequencing metrics corresponding to the subset of genomic coordinates exhibiting homozygous reference genotypes or homozygous alternate genotypes.
Ramachandran et al. in view of Heller et al. shows a convolutional neural network which intakes sequencing metrics from aligned reads to an assembly to produce probabilities that the alleles are true given the evidence for one allele to the allelic evidence of all other alleles and shows haplotag values which is utilized in the prediction and shows utilizing haplotype pairs when making predictions (Ramachandran et al. pages 24-27 sections “DNN architecture” and “Single-platform architecture”, page 28 “determining hotspots”, and page 29 “Preparing training data”). It would have been obvious to one of ordinary skill that the probability being output by the neural network model may be a haploid reference genotype and a haploid alternate genotype when two probabilities are used based aligned reads to an assembly.
An invention would have been obvious to one or ordinary skill in the art if some motivation in the prior art would have led that person to modify reference teachings to arrive at the claimed invention. It would have been obvious to one of ordinary skill in the art before the effective filling date to have modified the sequence reads used to predict true alleles which utilizes specific processing steps depending on ploidy states of Ramachandran et al. with the assembly pipeline which initially calls structural variants in sequence read data of Heller et al. because this would allow for a method that can predict if a structural variant allele is true that has been initially been identified by a model (Heller et al. Supplementary material Fig. S2). One would have a reasonable expectation of success because Heller et al. produces candidate structural variants from read alignments while Ramachandran et al. uses sequencing metrics to from read alignments to produce a prediction if an initially called allele is true.
The rejection below has been modified necessitated by amendment.
Claims 15-20 are rejected under 35 U.S.C. 103 as being unpatentable over Ramachandran et al. (BMC Bioinformatics 22, 404 (2021); previously cited) in view of DePristo et al. US 10,354,747 B1; previously cited).
Claim 15 is directed to determine, for one or more nucleotide reads, an initial genotype call indicating a homozygous reference genotype at a genomic coordinate of a sample nucleotide sequence;
Ramachandran et al. shows determining hotspots if any non-reference candidate allele has support from a sufficient fraction of reads aligning to the site, the site is determined to be a hotpot and is analyzed using the neural network for determination of true alleles (Ramachandran et al. page 28 “determining hotspots” section). Ramachandran et al. shows this pipeline is software implemented using processors (Ramachandran et al. page 4 “tool versions…” and page 21 paragraph 1).
determine sequencing metrics for the initial genotype call corresponding to the genomic coordinate;
Ramachandran et al. shows determining sequence metrics such as the read base, the reference base, the base quality, the mapping quality of the read, the strand identity of the read, and a position flag of a genomic coordinate which is encoded into a vector to be processed by a deep learning model (Ramachandran et al. page 24-25 and figure 10).
generate, utilizing a call recalibration machine learning model to process the sequencing metrics for the initial genotype call a set of variant call classifications indicating an accuracy of identifying a variant at the genomic coordinate;
Ramachandran et al. shows a convolutional neural network which intakes these sequencing metrics to produce probabilities that the alleles are true given the evidence for one allele to the allelic evidence of all other alleles (Ramachandran et al. page 24-27 sections “DNN architecture” and “Single-platform architecture”).
and determine a final genotype call comprising a variant call for the genomic coordinate based on the set of variant call classifications.
Ramachandran et al. shows determining a variant call for the genomic coordinate based on the classifications by determining which candidate allele is true by the convolutional neural network (Ramachandran et al. page 26 last full paragraph through page 27 third full paragraph).
Ramachandran et al. does not explicitly show determining an initial genotype call indicating a homozygous reference genotype at a genomic coordinate of a sample nucleotide sequence
Like Ramachandran et al., DePristo et al. shows the use of deep learning to analyze sequencing data in the context of determining variants. DePristo et al. shows a neural network model that uses read pileup data of overlapping sequences to determine the likelihood of the allele being homozygous reference, the likelihood of the allele being heterozygous variant, and the likelihood of the allele being homozygous reference (DePristo et al. col. 21-22).
Claim 16 is directed to receive, from a call generation model, an indication of the homozygous reference genotype at the genomic coordinate; and determine the variant call for the genomic coordinate by modifying the homozygous reference genotype to a different genotype based on the one or more variant call classifications.
Ramachandran et al. shows a convolutional neural network which intakes these sequencing metrics to produce probabilities that the alleles are true given the evidence for one allele to the allelic evidence of all other alleles (Ramachandran et al. page 24-27 sections “DNN architecture” and “Single-platform architecture”)
Claim 17 is directed to determine the sequencing metrics by determining one or more of read-based sequencing metrics, externally sourced sequencing metrics, or call model generated sequencing metrics for the genomic coordinate indicated as having a homozygous reference genotype with respect to a reference genome.
Ramachandran et al. shows determining sequence metrics such as the read base, the reference base, the base quality, the mapping quality of the read, the strand identity of the read, and a position flag of a genomic coordinate which is encoded into a vector to be processed by a deep learning model (Ramachandran et al. page 24-25 and figure 10).
Claim 18 is directed to identify a previous homozygous reference genotype call with respect to a reference genome from a call generation model for the sample nucleotide sequence at the genomic coordinate; identify a ground truth base call for the sample nucleotide sequence at the genomic coordinate; and modify the call recalibration machine learning model based on a comparison of the variant call for the genomic coordinate and the ground truth base call for the genomic coordinate.
Ramachandran et al. shows performing candidate allele scanning reads at a hotspot and prepares data then compares this data to ground truth data for labeling these labels will give candidate alleles labels of true or false based on ground truth data (Ramachandran et al. page 29 “preparing training data”). Ramachandran et al. shows a training data preparation process and the output of the process is labeled data that trains (or modifies) the convolutional neural network for identifying true alleles (Ramachandran et al. page 29 “preparing training data”).
Claim 19 is directed to generate the set of variant call classifications, including a homozygous reference classification indicating a probability of a homozygous reference genotype, with respect to a reference genome, at a genomic coordinate, a homozygous alternate classification indicating a probability of a homozygous alternate genotype, with respect to the reference genome, at the genomic coordinate, and a heterozygous genotype, with respect to the reference genome, at the genomic coordinate.
DePristo et al. shows a neural network model that uses read pileup data of overlapping sequences to determine the likelihood of the allele being homozygous reference, the likelihood of the allele being heterozygous variant, and the likelihood of the allele being homozygous variant (DePristo et al. col. 21-22). DePristo et al. shows using these likelihoods to classify the coordinate homozygous reference, heterozygous variant, or homozygous variant (DePristo et al. col. 23).
Claim 20 is directed to update one or more of a call quality field, a genotype field, or a genotype quality field corresponding to a variant call file based on the set variant call classifications.
Ramachandran et al. shows a convolutional neural network which intakes these sequencing metrics to produce probabilities that the alleles are true given the evidence for one allele to the allelic evidence of all other alleles and further shows once the an allele is determined the pipeline convert is to a variant record (Ramachandran et al. page 24-27 sections “DNN architecture” and “Single-platform architecture”).
An invention would have been obvious to one or ordinary skill in the art if some motivation in the prior art would have led that person to modify reference teachings to arrive at the claimed invention. It would have been obvious to one of ordinary skill in the art before the effective filling date to have modified that hotspot determination algorithm of Ramachandran et al. with the neural network to predict likelihoods and determinations of homozygous reference, heterozygous variant, or homozygous variant for initial calling of DePristo et al. because this would give a method that can discover false positives for homozygous reference calls due to the neural network of Ramachandran et al. utilizing allele based features that rely on base and mapping quality to discover if the initial call true (Ramachandran et al. page 24 section “DNN architecture”). Since the model of Ramachandran et al. utilizes allele-based features such as base and mapping quality initial calls of homozygous reference may be determined as false positives due to the allele-based features supporting a different allele at that particular position. One would have a reasonable expectation of because DePristo et al. shows a model to call zygosity from aligned reads while Ramachandran et al. utilizes features from identified genomic coordinates to determine which allele is true or false.
Response to Arguments
Applicant's arguments filed 13 November 2025 have been fully considered but they are not persuasive.
Arguments pertaining to Depristo et al. in view of Lam et al. for claims 1-7:
Applicant argues that Lam’s probabilities do not correspond to the claimed “variant call classifications” for the recited “multiallelic coordinate” where “three or more alleles” are possible (Reply p. 19-21).
This argument has been fully considered but found to be not persuasive. Lam et al. further shows that the output of the neural network can also output a predicted genotype association with the variant such as probabilities of a homozygous reference (a non-variant), a heterozygous variant (where only one of maternal or paternal copies has a variant), a homozygous variant (where both copies have the same variant, or other (where each copy has a different variant) (Lam et al. [0160]). Lam et al. shows an output probability of “other” which includes situations which occur at multiallelic coordinates where each copy has a different variant. The variant call probabilities of Lam show producing probabilities for more than three alleles that are possible at a genomic coordinate. Thus, Lam shows predicting these variant call classifications for the recited multiallelic coordinate.
Applicant argues that neither Lam nor Depristo describe or suggest the specific two-stage recalibration framework in which sequencing metrics of an “initial genotype call” are processed by a machine learning model to produce a “final genotype call” (Reply p. 21).
This argument has been fully considered but found to be not persuasive. DePristo et al. shows determining sequence metrics such as information about sample read data, identified alleles, and variant types for nucleotide base calls for a region (DePristo et al. col. 13-14) which shows that sequencing metrics are determined for an initial call (i.e., identified alleles) and the neural network of Depristo et al. in view of Lam et al. determines a final genotype call utilizing these sequencing metrics.
Applicant argues that a person of ordinary skill in the art would not have had a reasonable expectation of success in combining (a) DePristo's DeepVariant model, which performs variant calling by processing an image-based pileup representation of aligned sequencing reads, with (b) Lam's convolutional neural network, which operates on a non-image-based numerical feature matrix derived from augmented sequence alignments. The two systems employ fundamentally different data representations, preprocessing pipelines, and neural network architectures. Nothing in either reference suggests that their approaches could be merged, nor that doing so would yield predictable results (Reply p. 22).
This argument has been fully considered but found to be not persuasive. It is noted that both the neural networks in Depristo et al. and Lam et al. are convolutional neural networks and both features used as input into the model represent sequence alignments. Further, Convolutional Neural Networks process numerical data in the form of tensors (which include matrices) and images are treated as tensors when used as inputs into these models. Whether the data used for inputs to these models where images or non-image matrices before both are recognized by the convolutional neural networks as a tensor holding numerical data. Thus, one of ordinary skill in the art would recognize both models operate by processing numerical tensors (due to both models being convolutional neural networks) and would recognize images and non-image matrices are both able to be represented as a tensor (i.e., matrix holding values for the CNN’s to process).
Applicant argues that Lam expressed criticism of the model described in Depristo et al. would have discouraged a skilled artisan from attempting a combination (Reply p. 22).
This argument has been full consider but found to be not persuasive. It is noted that the rationale relied upon was the modification of the model taught in Depristo et al. with the architecture and elements of the model of Lam et al. (not a combination of these two models). Thus, the criticism provided by Lam et al. along with the description of a model capable of processing data to produce multiple genotypes calls on candidate variant sites could reasonably encourage a skilled artisan to modify the model of Depristo et al. in the manner provided by Lam et al. (i.e., the elements and architecture of the model).
Arguments pertaining to Ramachandran et al. in view of Heller et al. for claims 8-14:
Applicant argues neither Heller nor Ramachandran teach or suggest distinctive metrics catered to “a final genotype call indicating a haploid genotype for the genomic coordinate” (Reply p. 24). Applicant argues that none of Ramachandran’s output probabilities indicate a haploid reference genotype with respect to a reference genome at genomic coordinate and a second probabilities of a haploid alternate genotype with respect to a reference genome (Reply p. 24-25). Applicant argues neither Ramachandran nor Heller describe or suggest the two-stage recalibration framework recited in claim 8, in which sequencing metrics of an initial genotype call are processed to produce a final genotype call (Reply p. 25-26).
This argument has been fully considered but found to be not persuasive. As described above, Ramachandran et al. in view of Heller et al. shows the use of a haplotag value in the predictions of determining if a candidate variant is a true allele at a hotspot region (Ramachandran et al. pages 24-27 sections “Single-platform architecture”) while the pipeline in Heller et al. provides a pipeline for determining these hotspot regions (which is interpreted as an sequencing metrics for an initial genotype call) to be used for determining if the candidate allele is true allele (which is interpreted as the final genotype call based on the metrics from the hotspot region including the haplotag value).
Arguments pertaining to Ramachandran et al. in view of Depristo et al. for claims 15-20:
Applicant argues by contrast, independent claim 15 as amended requires determining "an initial genotype call indicating a homozygous reference genotype at a genomic coordinate" as an input for a subsequent machine-learning-processing step, not generating a probability of that genotype as the DePristo model's output in the first instance. The Office Action's reasoning thus reverses the order of operations described in the claim and fails to identify any disclosure in DePristo, Ramachandran, or the other cited references that teaches or suggests the claimed operations (Reply p. 26-27).
This argument has been fully considered but found to be not persuasive. Depristo et al. describes a neural network model which infers genotypes directly from read pileup images and states “in addition to the pipeline variant caller… the present application also describes using neural network modeling variant genotype… using images constructed from data representing read pileup image” (Depristo et al. col. 22). Depristo et al. shows performing genotype labeling utilizing predictions from the neural network. Depristo et al. is relied upon for showing an initial genotype call which includes initially calling a homozygous reference genotype and generating a quality score for this genotype at a location (Depristo et al. col. 22-23). Further, the rational as described above provides modifying the identification hotspot regions Ramachandran’s with the initial genotype calling (which includes a situation where the initial call is a homozygous reference call) and produced quality scores of Depristo et al. where the identified regions with the initial genotype call are then processed by Ramachandran to determine if the initial call is true utilizing quality values encoded into the feature vector (Ramachandran et al. pages 24-27). Thus, Ramachandran et al in view of Depristo et al. show an initial call (encompassing calling a homozygous reference) from read alignment to a reference genome to produce sequencing metrics such as quality which are utilized by Ramchandran et al. to produce the probability the initial call is true utilizing quality along with additional sequencing metrics.
Double Patenting
The provisional rejection of claims 1-7 on the ground of nonstatutory double patenting as being unpatentable over claims 1-5 of copending Application No. 17/384,423 in view of DePristo et al. (US 10,354,747 B1; previously cited) in view of Lam et al. (US 20210257050 A1; previously cited) in Office action mailed 14 August 2025 is withdrawn in view of the amendments of “generate, utilizing a call recalibration machine learning model to process the sequencing metrics… a reference probability of a homozygous reference genotype, with respect to a reference genome… a variant call probability of a variant call genotype with respect to the reference genome, being correct at the multiallelic genomic coordinate” received 13 November 2025.
The provisional rejection of claims 8-14 on the ground of nonstatutory double patenting as being unpatentable over claim 16 of copending Application No. 17/384,423 in view of DePristo et al. (US 10,354,747 B1; previously cited) in view of Lam et al. (US 20210257050 A1; previously cited) in Office action mailed 14 August 2025 is withdrawn in view of the amendments of “generate, utilizing a call recalibration machine learning model to process the sequencing metrics… a first genotype probability of a haploid reference genotype, with respect to a reference genome… a second genotype probability of a haploid alternate genotype, with respect to the reference genome, at the genomic coordinate” received 13 November 2025.
The provisional rejection of claims 15-20 on the ground of nonstatutory double patenting as being unpatentable over claims 11, 13, and 15 of copending Application No. 17/384,423 in view of DePristo et al. (US 10,354,747 B1; previously cited) in view of Lam et al. (US 20210257050 A1; previously cited) in Office action mailed 14 August 2025 is withdrawn in view of the amendments of “determine, for one or more nucleotide reads, an initial genotype call indicating a homozygous reference genotype” received 13 November 2025.
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
The provisional rejection below is newly recited necessitated by amendment.
Claims 1-7 are provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claims 10 and 12 of copending Application No. 18/654914
Regarding instant claim 1, the copending application 18/654914 shows determining sequence metrics for an initial genotype call as extract, from the one or more sequencing data files, sequencing metrics for the genotype call (copending application claim 10),
the copending application 18/654914 shows generate utilizing a call recalibration machine learning model a reference probability, a differing probability of a genotype error, and a variant-call probability as identifying the genomic coordinate as a multiallelic genomic coordinate; generating, utilizing the call-recalibration-machine-learning model, the one or more variant-call classifications comprising one or more of a reference probability that the genotype call comprises a homozygous reference genotype at the multiallelic genomic coordinate, a zygosity-error probability that the genotype call comprises a genotype-zygosity error at the multiallelic genomic coordinate, or a true-positive variant probability that the genotype call constitutes a true positive variant at the multiallelic genomic coordinate (copending application claim 12).
The copending application 18/654914 shows determine a final genotype call for the multiallelic genomic coordinate based on the set of variant call classifications as determining the updated genotype call at the multiallelic genomic coordinate based on one or more of the reference probability, the zygosity-error probability, or the true-positive variant probability (copending application claim 12).
Regarding instant claims 2 and 3, copending application 18/654914 shows generating a recalibrated sequencing data file comprising an updated genotype call at the genomic coordinate and shows that the updated genotype call is based on probabilities of genotypes at a multiallelic coordinate (copending application claims 10 and 12).
Regarding instant claim 4, copending application 18/654914 shows accessing for a sample nucleotide sequence, one or more sequencing data files comprising a genotype call at a genomic coordinate which are read-sequencing metrics derived from nucleotide reads of a sample (copending application 18/654914 claim 10).
Regarding instant claims 5-7, copending application 18/654914 shows generate utilizing a call recalibration machine learning model a reference probability, a differing probability of a genotype error, and a variant-call probability as identifying the genomic coordinate as a multiallelic genomic coordinate; generating, utilizing the call-recalibration-machine-learning model, the one or more variant-call classifications comprising one or more of a reference probability that the genotype call comprises a homozygous reference genotype at the multiallelic genomic coordinate, a zygosity-error probability that the genotype call comprises a genotype-zygosity error at the multiallelic genomic coordinate, or a true-positive variant probability that the genotype call constitutes a true positive variant at the multiallelic genomic coordinate (copending application claim 12).
This is a provisional nonstatutory double patenting rejection.
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.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JONATHAN EDWARD HAYES whose telephone number is (571)272-6165. The examiner can normally be reached M-F 9am-5pm.
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 at 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.
/J.E.H./Examiner, Art Unit 1685
/KAITLYN L MINCHELLA/Primary Examiner, Art Unit 1685