DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Status
Claims 1-20 are currently pending and under exam herein.
Claims 1-20 are rejected.
Claims 2-13 are objected to.
Priority
Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy of the foreign priority application was received on December 22, 2022. Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55. At this point in the examination, the effective filing date of claims 1-20 is August 20, 2021.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on August 19, 2022 and January 20, 2026 are in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statements are being considered by the examiner.
Nucleotide and/or Amino Acid Sequence Disclosures
REQUIREMENTS FOR PATENT APPLICATIONS CONTAINING NUCLEOTIDE AND/OR AMINO ACID SEQUENCE DISCLOSURES
Items 1) and 2) provide general guidance related to requirements for sequence disclosures.
37 CFR 1.821(c) requires that patent applications which contain disclosures of nucleotide and/or amino acid sequences that fall within the definitions of 37 CFR 1.821(a) must contain a "Sequence Listing," as a separate part of the disclosure, which presents the nucleotide and/or amino acid sequences and associated information using the symbols and format in accordance with the requirements of 37 CFR 1.821 - 1.825. This "Sequence Listing" part of the disclosure may be submitted:
In accordance with 37 CFR 1.821(c)(1) via the USPTO patent electronic filing system (see Section I.1 of the Legal Framework for Patent Electronic System (https://www.uspto.gov/PatentLegalFramework), hereinafter "Legal Framework") as an ASCII text file, together with an incorporation-by-reference of the material in the ASCII text file in a separate paragraph of the specification as required by 37 CFR 1.823(b)(1) identifying:
the name of the ASCII text file;
ii) the date of creation; and
iii) the size of the ASCII text file in bytes;
In accordance with 37 CFR 1.821(c)(1) on read-only optical disc(s) as permitted by 37 CFR 1.52(e)(1)(ii), labeled according to 37 CFR 1.52(e)(5), with an incorporation-by-reference of the material in the ASCII text file according to 37 CFR 1.52(e)(8) and 37 CFR 1.823(b)(1) in a separate paragraph of the specification identifying:
the name of the ASCII text file;
the date of creation; and
the size of the ASCII text file in bytes;
In accordance with 37 CFR 1.821(c)(2) via the USPTO patent electronic filing system as a PDF file (not recommended); or
In accordance with 37 CFR 1.821(c)(3) on physical sheets of paper (not recommended).
When a “Sequence Listing” has been submitted as a PDF file as in 1(c) above (37 CFR 1.821(c)(2)) or on physical sheets of paper as in 1(d) above (37 CFR 1.821(c)(3)), 37 CFR 1.821(e)(1) requires a computer readable form (CRF) of the “Sequence Listing” in accordance with the requirements of 37 CFR 1.824.
If the "Sequence Listing" required by 37 CFR 1.821(c) is filed via the USPTO patent electronic filing system as a PDF, then 37 CFR 1.821(e)(1)(ii) or 1.821(e)(2)(ii) requires submission of a statement that the "Sequence Listing" content of the PDF copy and the CRF copy (the ASCII text file copy) are identical.
If the "Sequence Listing" required by 37 CFR 1.821(c) is filed on paper or read-only optical disc, then 37 CFR 1.821(e)(1)(ii) or 1.821(e)(2)(ii) requires submission of a statement that the "Sequence Listing" content of the paper or read-only optical disc copy and the CRF are identical.
Specific deficiencies and the required response to this Office Action are as follows:
Specific deficiency - This application fails to comply with the requirements of 37 CFR 1.821 - 1.825 because it does not contain a "Sequence Listing" as a separate part of the disclosure or a CRF of the “Sequence Listing.”. A sequence listing is required because sequences are present in Figures 4 and 8-9.
Required response - Applicant must provide:
A "Sequence Listing" part of the disclosure; together with
An amendment specifically directing its entry into the application in accordance with 37 CFR 1.825(a)(2);
A statement that the "Sequence Listing" includes no new matter as required by 37 CFR 1.821(a)(4); and
A statement that indicates support for the amendment in the application, as filed, as required by 37 CFR 1.825(a)(3).
If the "Sequence Listing" part of the disclosure is submitted according to item 1) a) or b) above, Applicant must also provide:
A substitute specification in compliance with 37 CFR 1.52, 1.121(b)(3) and 1.125 inserting the required incorporation-by-reference paragraph, consisting of:
A copy of the previously-submitted specification, with deletions shown with strikethrough or brackets and insertions shown with underlining (marked-up version);
A copy of the amended specification without markings (clean version); and
A statement that the substitute specification contains no new matter.
If the "Sequence Listing" part of the disclosure is submitted according to item 1) c) or d) above, applicant must also provide:
A CRF in accordance with 37 CFR 1.821(e)(1) or 1.821(e)(2) as required by 1.825(a)(5); and
A statement according to item 2) a) or b) above.
Specific deficiency – Nucleotide and/or amino acid sequences appearing in the drawings are not identified by sequence identifiers in accordance with 37 CFR 1.821(d). Sequence identifiers for nucleotide and/or amino acid sequences must appear either in the drawings or in the Brief Description of the Drawings. Sequences are present in figures 4 and 8-9.
Required response – Applicant must provide:
Replacement and annotated drawings in accordance with 37 CFR 1.121(d) inserting the required sequence identifiers;
AND/OR
A substitute specification in compliance with 37 CFR 1.52, 1.121(b)(3) and 1.125 inserting the required sequence identifiers into the Brief Description of the Drawings, consisting of:
A copy of the previously-submitted specification, with deletions shown with strikethrough or brackets and insertions shown with underlining (marked-up version);
A copy of the amended specification without markings (clean version); and
A statement that the substitute specification contains no new matter.
Drawings
The drawings are objected to because they include nucleotide and/or amino acid sequences with sequence identifiers (see above).
The drawings are objected to as failing to comply with 37 CFR 1.84(p)(4) because reference character “83” has been used to designate both a first convolutional layer and a reduction layer in Figure 8. Corrected drawing sheets in compliance with 37 CFR 1.121(d) 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.
Claim Objections
Claims 2-13 are objected to because of the following informalities:
Regarding claims 2-9, in the preamble of the claims the phrase “The computer-implemented method for machine-learning method of claim…” is grammatically incorrect because it has a redundant noun issue. It is recommended to amend the preambles in a similar way as in claims 15-17 (e.g., The method of claim…, The computer-implemented method of claim…, The computer-implemented method for variant calling of claim…, etc.).
Regarding claim 10, the phrase “…the neural network being machine-learnt according to machine-learning a neural network…” in lines 29-30 is grammatically incorrect because it has a noun-as-a-verb error (machine-learning is a noun, but the claim is using it as a verb) and a mismatched voice and tense (being machine-learnt sounds unnatural as a passive progressive verb). The phrase is suggested to be amended to something similar to “… the neural network being trained to perform variant calling with respect to a reference genome…”
Regarding claims 11-13, in the preamble of the claims the phrase “The computer-implemented method for variant-calling method of claim…” is grammatically incorrect because it has a redundant noun issue. It is recommended to amend the preambles in a similar way as in claims 15-17 (e.g., The method of claim…, The computer-implemented method of claim…, The computer-implemented method for variant calling of claim…, etc.).
Appropriate correction is required.
Claim Warning
Applicant is advised that should claim 14 be found allowable, claim 18 will be objected to under 37 CFR 1.75 as being a substantial duplicate thereof. When two claims in an application are duplicates or else are so close in content that they both cover the same thing, despite a slight difference in wording, it is proper after allowing one claim to object to the other as being a substantial duplicate of the allowed claim. See MPEP § 608.01(m).
Claim Interpretation
In accordance with MPEP § 2111, pending claims must be given their broadest reasonable interpretation consistent with the specification as it would be interpreted by one of ordinary skill in the art. As such, the following interpretations of the claim language are noted:
Claims 1, 10, and 14 recite a symmetric function. The special definition in paragraph 0061, the explanation in paragraph 0114, and Figure 8 of the published specification will be used to define the claim limitations. Essentially, the neural network operates in such a way that it does not matter what order the data pieces in the sets are organized in when being input into the neural network and the output will always be the same if the input data pieces are the same even if they are ordered differently.
The preambles of claims 1-9 recite a computer-implemented method for machine-learning a neural network for variant calling seems to direct the claims toward a method of training a neural network variant caller, but the bodies of the claims do not recite steps to train such a neural network. Rather, the bodies of the claims recite steps to perform variant calling, describe the input data, and the architecture of the neural network and how it processes data. Therefore, the preambles are interpreted to recite an intended use that is not a limitation of the claims and the claims will be examined with the limitations present in the bodies of the claims.
Claim Rejections - 35 USC § 112
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.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 3 and 16 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding claims 3 and 16, a broad range or limitation together with a narrow range or limitation that falls within the broad range or limitation (in the same claim) may be considered indefinite if the resulting claim does not clearly set forth the metes and bounds of the patent protection desired. See MPEP § 2173.05(c). In the present instance, the claims recite the broad recitation “wherein the base descriptors include one or more descriptors representing insertion size and/or deletion size…”, and the claims also recites “…including an insertion size descriptor and a deletion size descriptor” which is the narrower statement of the range/limitation. The claims are considered indefinite because there is a question or doubt as to whether the feature introduced by such narrower language is (a) merely exemplary of the remainder of the claim, and therefore not required, or (b) a required feature of the claims.
Therefore, the claims are indefinite and are rejected under 35 U.S.C. 112(b) or pre-AIA 35 U.S.C. 112, second paragraph.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception without significantly more.
Step 1:
The first part of the eligibility analysis evaluates whether a claim falls withing any statutory
category (MPEP 2106.03). Claims 1-13 and 18-19 recite a series of steps to take on a computer to use a neural network variant caller to determine the presence and types of variants present in sequencing data. The claims are directed to a method and fall within one of the statutory categories of invention. Claims 14-17 and 20 recite a computer system which implements the method to use a neural network variant caller. The claims are directed to a computer system, which is a machine, and falls within one of the statutory categories of invention. Therefore, the claims belong to statutory categories of invention (Step 1: YES).
Step 2A, prong 1:
In accordance with MPEP § 2106, claims found to recite statutory subject matter (Step 1: YES)
are then analyzed to determine if the claims recite any concepts that equate to an abstract idea, law of
nature, or natural phenomenon (Step 2A, prong 1). In the instant application, the claims recite the
following limitations that equate to an abstract idea:
Claims 1 and 18 recite: … including, for each set of data pieces, a respective function configured to take as input and process the set of data pieces, the respective function being symmetric.
Claim 2 recites: … and each respective function comprises one or more convolutional layers for processing each respective sequence of base descriptors, each convolutional layer applying one or more one-dimensional convolutional filters.
Claim 4 recites: wherein each respective function includes a reduction layer for collapsing order- independently the output of the one or more convolutional layers into a set of features.
Claim 6 recites: wherein the reduction layer includes one or more order-independent operators, including a mean and/or a standard deviation.
Claim 7 recites: … includes one or more fully connected layers configured to process output of each reduction layer and perform classification.
Claim 9 recites: … including: a first function for the first set of reads and a second function for the second sets of reads, and a layer for aggregating the output of the first and second functions.
Claims 10 and 19 recite: … including, for each set of data pieces, a respective function configured to take as input and process the set of data pieces, the respective function being symmetric, application of the neural network comprising applying, for each set of data pieces, the respective symmetric function of the neural network configured to take as input and process the set of data pieces.
Claim 11 recites: … determining a set of regions of interest in the reference genome by comparing the one or more sets of reads with the reference genome; and for each given region of interest: performing a haplotype reconstruction based on the obtained one or more sets of reads of the given region to determine two or more haplotypes, re-aligning the one or more sets of reads of the given region based on the two or more haplotypes, deducing potential variants of the given region based on the re-aligned one or more sets of reads and the two or more haplotypes, performing a coarse grain filtering to detect candidate variants from the potential variants, each detected candidate variant corresponding to a respective genomic position, and for each detected candidate variant, determining one or more sets of data pieces each specifying a respective read aligned relative to the genomic position corresponding to the detected candidate variant…
Claim 12 recites: inferring a set of potential haplotypes by enumerating a predetermined number of longest paths in a directed acyclic graph; and/or selecting a subset of haplotypes from the set of potential haplotypes, the subset of haplotypes being potential haplotypes of the set having the highest number of supporting reads, the subset of haplotypes corresponding to the two or more haplotypes that the haplotype reconstruction determines.
Claim 13 recites: … the deducing of potential variants of the given region including, for germline variants, evaluating a probability that a variant is more probable than the reference genome…; … the deducing of potential variants of the given region considering, for somatic variants: a presence of a germline variant and/or the presence of a somatic variant, and a somatic variant frequency.
Claim 14 recites: … for each set of data pieces, a respective function configured to take as input and process the set of data pieces, the respective function being symmetric.
Claim 15 recites: … and each respective function includes one or more convolutional layers for processing each respective sequence of base descriptors, each convolutional layer applying one or more one-dimensional convolutional filters.
Claim 17 recites: wherein each respective function includes a reduction layer for collapsing order-independently the output of the one or more convolutional layers into a set of features.
The limitations of a respective symmetric function, the functions comprising convolutional layers applying one-dimensional filters, the functions including a reduction layer to order-independently collapse output, the reduction layer including order-independent operators including mean or standard deviation, fully connected layers for processing output and performing classification, a first and second function, an aggregating layer, and applying the symmetric function to a set of data pieces recited in claims 1-2, 4, 6-7, 9-10, 14-15, and 17 are mathematical formulas or equations and calculations used to process data and arrive at an output. Additionally, the claim 7 limitation of a fully connected layer configured to perform classification and the claim 9 limitation of a layer for aggregating output encompass mental processes of observing data and evaluating its’ class and combining data into one set, respectively.
The limitations of determining regions of interest by comparing reads with a reference genome, performing haplotype reconstruction, re-aligning reads to the haplotypes, deducing potential variants, and determining sets of data pieces aligned to a genomic position of a candidate variant recited in claims 11 are recitations of mental processes of observing and choosing desired points for further analysis, observing where individual reads overlap and combining them to form haplotypes, evaluating where reads are most similar to the haplotype sequences, observing deviations from the alignment to identify potential variants, and selecting which reads are aligned to a candidate variant position. Also, the limitations of performing haplotype reconstruction, re-aligning reads to the haplotypes, and performing coarse grain filtering encompasses performing mathematical calculations.
The limitations of inferring potential haplotypes by enumerating longest paths in a directed acyclic graph, deducing potential germline variants by evaluating probability, and deducing potential somatic variants by considering presence of germline or somatic variant and somatic variant frequency recited in claims 12 and 13 are mathematical relationships and calculations. The limitations of selecting a subset of haplotypes recited in claim 12 encompasses the mental process of observing data and deciding which haplotypes are desired.
Therefore, these limitations fall under the “Mathematical concepts” and “Mental processes” groupings of abstract ideas (Step 2A, prong 1: YES).
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
(Step 2A, prong 2). The claims recite the following additional elements:
Claim 1 recites: A computer-implemented method for machine-learning a neural network…; … the neural network taking as input one or more sets of data pieces each specifying a respective read aligned relative to a genomic position of the reference genome; and the neural network outputting information with respect to presence of a variant at the genomic position.
Claim 2 recites: wherein each data piece includes a respective sequence of base descriptors of the respective read…
Claim 3 recites: wherein the base descriptors include one or more descriptors representing insertion size and/or deletion size, including an insertion size descriptor and a deletion size descriptor.
Claim 5 recites: wherein the reduction layer further takes as input read descriptors for each respective read, and/or wherein the read descriptors include a haplotype support descriptor.
Claim 8 recites: wherein the neural network further takes as input pile descriptors, and/or wherein the pile descriptors include a descriptor representing depth and/or a descriptor representing a Bayesian variant estimation.
Claim 9 recites: wherein the one or more sets of reads include a first set of reads for germline variants and a second set of reads for somatic variants…
Claim 10 recites: A computer-implemented method…; obtaining, as input, one or more sets of data pieces each specifying a respective read aligned relative to a genomic position of the reference genome; and applying, to the input, a neural network, to output information with respect to presence of a variant at the genomic position, the neural network being machine-learnt according to machine-learning a neural network for variant calling with respect to a reference genome, the neural network being configured to take as input one or more sets of data pieces each specifying a respective read aligned relative to a genomic position of the reference genome, and to output information with respect to presence of a variant at the genomic position…
Claim 11 recites: obtaining one or more sets of reads aligned relative to the reference genome; and … wherein the obtaining of the one or more sets of data pieces and the applying of the neural network are performed for each detected candidate variant with the determined one or more sets of data pieces.
Claim 13 recites: wherein the one or more sets of reads includes a first set of reads for germline variants…; and … wherein the one or more sets of reads comprises a second set of reads for somatic variants…
Claim 14 recites: A device comprising: a non-transitory computer-readable data storage medium having recorded thereon a program comprising instructions for performing machine-learning or for using a neural network for variant calling with respect to a reference genome, that when executed by a processor causes the processor to be configured to via the neural network: take as input one or more sets of data pieces each specifying a respective read aligned relative to a genomic position of the reference genome, and output information with respect to presence of a variant at the genomic position…
Claim 15 recites: wherein each data piece includes a respective sequence of base descriptors of the respective read…
Claim 16 recites: wherein the base descriptors include one or more descriptors representing insertion size and/or deletion size, including an insertion size descriptor and a deletion size descriptor.
Claim 18 recites: A non-transitory computer readable medium having stored thereon a program that when executed by a computer causes the computer to implement the method for machine-learning the neural network for variant calling with respect to the reference genome…
Claim 19 recites: A non-transitory computer readable medium having stored thereon a program that when executed by a computer causes the computer to implement the method for variant calling with respect to the reference genome…
Claim 20 recites: The device of claim 14, further comprising the processor.
The additional elements of performing the method of variant calling using a neural network recited in claims 1 and 10 are recited at a high level of generality and invokes computers to perform the variant calling. These additional elements amount to mere instructions to apply the judicial exception in a generic computer environment (MPEP 2106.05f). In the same manner, the additional elements of non-transitory computer-readable data storage mediums and a processor recited in claims 14 and 18-20 are just limitations used to apply the judicial exception in a generic computer. The additional elements of inputting data into the neural network, the neural network outputting information about potential variants, the reduction layer taking input read descriptors, the neural network taking pile descriptors as input, and obtaining datasets to input into the neural network recited in claims 1, 5, 8, 10-11, and 14 are nominal or tangential additions to the invention that don’t impose significant limits and are necessary data gathering and outputting required for the uses of the recited judicial exception. As such, those additional elements are insignificant extra-solution activity (MPEP 2106.05g). The additional elements describing the specifics of the input datasets recited in claims 2-3, 9, 13, and 15-16 are only further limiting the additional elements of the data gathering steps and are extra-solution activity.
Therefore, the judicial exception is not integrated into a practical application because the claims do not recite an additional element that reflects an improvement to technology or applies/uses the recited judicial exception in some other meaningful way and the claims are directed to the judicial exception (Step 2A, prong 2: NO).
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 recite additional elements that equate to mere instructions to apply the recited
judicial exception in a generic computing environment. Claims that amount to nothing more than
instructions to apply the judicial exception using a generic computer do not render an abstract idea
eligible. Alice Corp., 576 U.S. at 223, 110 USPQ2d at 1983. See also 573 U.S. at 224, 110 USPQ2d at 1984. Since the data input and output steps are recited at a high level of generality, the claims also recite computer functions that the courts have ruled to be well-understood, routine, and conventional such as: Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network) and Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93.
Additionally, paragraph 0004 of the published specification states that it is general practice to call variant with sequencing data by comparing to a standard genome. Paragraph 0063 of the published specification states that it is “known per se” in the field of machine-learning that data input into a neural network is processed by applying operations to the data. Thus, demonstrating the conventionality of variant calling and inputting data into neural networks to train them and obtain output.
As such, the combination of additional elements recited in the claims is well-understood,
routine and conventional. The additional elements do not comprise an inventive concept when
considered individually or as an ordered combination that transform the claimed judicial exception into
a patent-eligible application of the judicial exception. Therefore, the claims do not amount to
significantly more than the judicial exception itself (Step 2B: NO) and claims 1-20 are not patent eligible.
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-4, 10, and 14-20 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Yakovenko et al. (arXiv:2003.07220, arXiv). The italicized text corresponds to the instant claim limitations.
Regarding claim 1, 10, 14, 18-19, and 20, Yakovenko et al. teach a neural network variant caller that processes individual reads independently and then average pooling across all reads in a pileup (p. 3-4, 3.1 Differences with DeepVariant, paragraphs 1-2; Figure 2). Yakovenko et al. teach reads are aligned to a reference genome, candidate variants are generated, and then variant probabilities are scored based on a read pileup around the variant in question (p. 2, Calling variants; Figure 1). These teachings disclose the claim 1 limitations of A computer-implemented method for machine-learning a neural network for variant calling with respect to a reference genome, comprising: the neural network taking as input one or more sets of data pieces each specifying a respective read aligned relative to a genomic position of the reference genome; and the neural network outputting information with respect to presence of a variant at the genomic position… Since the neural network in Yakovenko et al. processes reads individually and then pools them by averaging, the claim 1 limitation of …the neural network including, for each set of data pieces, a respective function configured to take as input and process the set of data pieces, the respective function being symmetric reads on the neural network taught. This is because the order of the reads input into Yakovenko et al.’s neural network does not matter to the final output since information between reads is not shared until the data are pooled by averaging (p. 4, 3.1 Differences with DeepVariant, paragraph 2). This feature of the taught neural network satisfies the definition of a symmetric function, as explained in the claim interpretation section, in paragraph 0061 of the published specification and the example of a symmetric function 90 in paragraph 0114 of the published specification referring to Figure 8. The limitations of claim 10 also read on the teachings of Yakovenko et al. in the same way they were applied to claim 1 and additional the teaching of obtaining datasets to train their model (p. 4, 4.2 Training with additional data, paragraph 1-3) discloses the claim 10 limitation of obtaining, as input, one or more sets of data pieces each specifying a respective read aligned relative to a genomic position of the reference genome.
Additionally, the limitations of claim 18 of a non-transitory computer readable medium having stored thereon a program that when executed by a computer causes the computer to implement the method for machine-learning the neural network for variant calling with respect to the reference genome according to claim 1 and the claim 19 limitation of a non-transitory computer readable medium having stored thereon a program that when executed by a computer causes the computer to implement the method for variant calling with respect to the reference genome according to claim 10 read on Yakovenko et al. since a non-transitory computer readable medium is inherently required to train and use a neural network model using the PyTorch framework and a graphics processing unit (GPU) (p. 7, 5.3 Training, paragraph 4). The limitations of claim 14 and 20 also read on the cited teachings of Yakovenko et al. since their neural network was trained and used on a computer which would have a non-transitory computer readable data storage medium and a processor in order to use the PyTorch software and GPU.
Regarding claims 2 and 15, Yakovenko et al. teach input data into the neural network consists of base quality scores, strand direction, and masks representing the reference and variant proposal (p. 5, Encoding individual reads; Table 2) which discloses the claims 2 and 15 limitation of wherein each data piece includes a respective sequence of base descriptors of the respective read… Yakovenko et al. also teach individual reads are transformed through convolutional layers that have one-dimensional filters (p. 5, One-dimensional convolutional layers; Figure 2) which discloses the claims 2 and 15 limitation of each respective function comprises one or more convolutional layers for processing each respective sequence of base descriptors, each convolutional layer applying one or more one-dimensional convolutional filters.
The claims 3 and 16 limitation of wherein the base descriptors include one or more descriptors representing insertion size and/or deletion size, including an insertion size descriptor and a deletion size descriptor also read on the previous teaching as shown in Table 2 of an example read encoding a potential insertion mutation. The encoding in table 2 shows a proposed insertion mutation A -> ATT and has a row for the variant length mask which describes the size of the insertion mutation. Table 1 (p. 3) demonstrates several insertion mutations of varying length correctly identified by the neural network. Yakovenko et al. teach insertions and deletions are collectively referred to as indels (p. 2, Human genome, paragraphs 1-2) and their model predicted both (p. 7, Table 4) which shows that the encoded reads for deletion mutations would inherently have deletion descriptors.
Regarding claims 4 and 17, Yakovenko et al. teach combining single reads by performing mean and max pooling operations which ignores read order in the pileup (p. 5, Final layer pooling) which discloses the claims 4 and 17 limitation of wherein each respective function includes a reduction layer for collapsing order-independently the output of the one or more convolutional layers into a set of features.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 5-9 are rejected under 35 U.S.C. 103 as being unpatentable over Yakovenko et al. (arXiv:2003.07220, arXiv) as applied to claims 1-4 above, and further in view of Lam et al. (US20210257050A1; IDS document 8/19/2022). The italicized text corresponds to the instant claim limitations.
The limitations of claims 1-4 have been taught by Yakovenko et al.
Regarding claim 6, the limitation of claim 6 of wherein the reduction layer includes one or more order-independent operators, including a mean… reads on the teaching by Yakovenko et al. of the final pooling layer performing mean pooling operations (p. 5, Final layer pooling; Figure 2).
Regarding claim 7, Yakovenko et al. teach a fully connected network made several layers that takes the output of the pooling layer and out puts a predicted variant call (p. 5, Fully connected network for variant candidate classification) which discloses the limitation of claim 7 of wherein the neural network includes one or more fully connected layers configured to process output of each reduction layer and perform classification.
The claim 8 limitations of wherein the neural network further takes as input pile descriptors, and/or wherein the pile descriptors include a descriptor representing depth… read on the teachings of Yakovenko et al. that the input to their neural network consist of a pileup of aligned reads and a variant candidate proposal, including a mask representing the variant proposal (p. 5, Encoding individual reads; Figure 1; Table 2) and that possible variants are generated with a high-recall and low precision that the neural network then scores variant probabilities based on a reads pileup around the variant in question (p. 2, Calling variants; p. 6, Candidate generation); Additionally, Table 1 shows a measurement of depth is present for correctly called variants. These teachings demonstrate the input data has metadata (i.e., pile descriptors) describing the pileup of reads used to score proposed variants.
Yakovenko et al. is silent on the claim 5 limitation of wherein the reduction layer further takes as input read descriptors for each respective read… and the claim 9 limitation of wherein the one or more sets of reads include a first set of reads for germline variants and a second set of reads for somatic variants, the neural network including: a first function for the first set of reads and a second function for the second sets of reads, and a layer for aggregating the output of the first and second functions. However, these limitations were known in the art at the effective filing date of the invention, as taught by Lam et al.
Regarding claim 5, Lam et al. teach a convolutional neural network for variant calling that may have pooling layers and identity shortcut layers ‘515’ that connect inputs of some layers to outputs of other layers (paragraph 0156; Figure 5). Figure 5 shows that the shortcut identify layers can be connected to the output of a convolutional layer which then feeds into a pooling layer and Figure 4 shows one embodiment of the various input into the neural network detailing information about the reads, such as alignment feature matrices which detail quality of sequencing reads and their alignment (paragraphs 0161-0164). These shortcut layers allow data to bypass layers of a neural network just like the highway layers in Yakovenko et al. and the teaching shows they can be connected to the input of a pooling layer, thus disclosing the limitation of claim 5 of wherein the reduction layer further takes as input read descriptors for each respective read…
Regarding claim 9, Lam et al. teach in some embodiments performing somatic variant calling, their neural network takes as input a plurality of tumor sequence reads (e.g., obtained from cancerous tissue of a patient) and a plurality of normal sequence reads (e.g., obtained from normal tissue of the same patient) (paragraph 0167; Figure 6; Figure 8) which discloses the claim 9 limitation of wherein the one or more sets of reads include a first set of reads for germline variants and a second set of reads for somatic variants… Lam et al. teach that their neural network can have the same or substantially the same architecture to predict germline and somatic variants, but the architecture can be optimized/modified for either type of variant calling (paragraph 0178) thus teaching the claim 9 limitation of … the neural network including: a first function for the first set of reads and a second function for the second sets of reads… Lastly, the claim 9 limitation of … a layer for aggregating the output of the first and second functions reads on the previously mentioned pooling layers (paragraph 0156) that may be between convolutional layers and right before fully connected layers as in Figure 9.
The prior art contained a convolutional neural network with a symmetric function that could take in sequence reads in any order to perform variant calling as taught by Yakovenko et al. That neural network had highway layers that bypassed convolutional layers and concatenated read information, but instead of passing the concatenated reads into a pooling layer, the highway layer passed the concatenated read information directly into the fully connected network (p. 5, Highway layers). This differs from the claimed invention in claims 5-8, however this different architecture was known in the art as taught by Lam et al. which showed identity shortcut layers that bypassed convolutional layers and fed into pooling layers (paragraph 0156; Figure 5). One of ordinary skill in the art could have substituted the architecture of the highway layers in Yakovenko et al. with the architecture of the identify shortcut layers in Lam et al. to have the pooling layer take the concatenated read information as input because the modification only changes where data is output and the purpose of bypassing convolutional layers is still achieved. The results would be predictable in that the final pooling layer would output both the concatenated read information and pooled reads to the fully connected network. The invention of claims 5-8 is therefore prima facie obvious.
The prior art contained a base neural network with a symmetric function as disclosed in Yakovenko et al. that performed only germline variant calling, whereas the claimed invention in claim 9 is an improvement over the prior art in that the invention is configured to perform somatic variant calling. However, the prior art taught a different neural network in Lam et al. that performed somatic variant calling using a technique applicable to the base neural network in Yakovenko et al. One of ordinary skill in the art would recognize that applying the technique of somatic variant calling to the base neural network would predictably result in an improved system because the prior art already showed other convolutional neural network performing both types of variant calling. The results would yield a convolutional neural network with a symmetric function that could perform germline and somatic variant calling. The invention of claim 9 is therefore prima facie obvious.
Claims 11-13 are rejected under 35 U.S.C. 103 as being unpatentable over Yakovenko et al. (arXiv:2003.07220, arXiv) as applied to claim 10 above, and further in view of Poplin et al. (Nature Biotechnology, vol. 36, no. 10, pp.983-87; IDS document 8/19/2022) and Garrison et al. (arXiv:1207.3907). The italicized text corresponds to the instant claim limitations.
The limitations of claim 10 have been taught by Yakovenko et al.
Yakovenko et al. is silent on the limitations of claims 11-13. However, these limitations were known in the art at the effective filing date of the invention as taught by Poplin et al. and Garrison et al.
Regarding claim 11, Poplin et al. teach, for haplotype-aware read realignment, selecting locations across the genome for reassembly by looking for evidence of possible variation, using overlapping reads to reconstruct haplotypes, selecting the top two haplotypes based read evidence, and performing a read realignment (p. 6, Haplotype-aware realignment of reads) which discloses the limitations …prior to the obtaining of the one or more sets of data pieces and the applying of the neural network: obtaining one or more sets of reads aligned relative to the reference genome; determining a set of regions of interest in the reference genome by comparing the one or more sets of reads with the reference genome; and for each given region of interest: performing a haplotype reconstruction based on the obtained one or more sets of reads of the given region to determine two or more haplotypes, re-aligning the one or more sets of reads of the given region based on the two or more haplotypes. Furthermore, Poplin et al. teach finding candidate variants to be evaluated with their deep learning model by looking at the CIGAR string of reads overlapping a site in the genome and determining the allele aligned to the site for each read (p. 6, Finding candidate variants, paragraph 1) which discloses the limitation …deducing potential variants of the given region based on the re-aligned one or more sets of reads and the two or more haplotypes… Then, the number of occurrences of each distinct allele across all reads is counted and if a candidate passes the calling thresholds at a site in the genome, a VCF-like record is emitted (p. 6, Finding candidate variants, paragraph 2) which teaches the limitation …performing a coarse grain filtering to detect candidate variants from the potential variants, each detected candidate variant corresponding to a respective genomic position… This is done for each site in the genome (p. 6, Finding candidate variants, paragraph 1) and as such disclose the limitation for each detected candidate variant, determining one or more sets of data pieces each specifying a respective read aligned relative to the genomic position corresponding to the detected candidate variant, wherein the obtaining of the one or more sets of data pieces and the applying of the neural network are performed for each detected candidate variant with the determined one or more sets of data pieces.
Regarding claim 12, Poplin et al. teach that haplotypes are made with a local De-Bruijn-graph-based read assembly procedure where graphs are constructed using multiple fixed k-mer sizes, candidate haplotypes are generated by traversing the assembly graphs, and the top haplotype candidates are selected that best explain the read evidence (p. 6, Haplotype-aware realignment of reads) which discloses the limitations of wherein the performing of the haplotype reconstruction further comprises: inferring a set of potential haplotypes by enumerating a predetermined number of longest paths in a directed acyclic graph; and/or selecting a subset of haplotypes from the set of potential haplotypes, the subset of haplotypes being potential haplotypes of the set having the highest number of supporting reads, the subset of haplotypes corresponding to the two or more haplotypes that the haplotype reconstruction determines.
Regarding claim 13, Garrison et al. teach calculating probabilities to identify potential variants represented in a set of reads for a particular locus (p. 3-4, 2.3 Estimating the probability of a sample genotype given sequencing observations, P(Ri|Gi)) which discloses the limitation of wherein the one or more sets of reads includes a first set of reads for germline variants, the deducing of potential variants of the given region including, for germline variants, evaluating a probability that a variant is more probable than the reference genome…
The prior art taught a convolutional neural network variant caller with a symmetric function as taught by Yakovenko et al. and the prior art taught a convolutional neural network variant caller that performed haplotype reassembly as taught by Poplin et al. which is a comparable neural network variant caller. Garrison et al. teach that haplotype-based variant detection methods offer several benefits over methods which operate on a single position at a time (p. 1, Motivation, paragraph 2). Yakovenko et al. were aware of the variant caller (i.e., DeepVariant) of Poplin et al. and taught that it demonstrated deep neural networks can be competitive with traditional variant calling methods (p. 1, Introduction, paragraph 2) and compared their neural network variant caller with DeepVariant (p. 3-4, Related Work; p. 7, Table 4). One of ordinary skill in the art could have applied the known haplotype reconstruction and realignment technique to the neural network of Yakovenko et al. because it is a known bioinformatics technique to improve variant calling from sequencing data which would yield the predictable result of an improved variant caller. The invention of claims 11-12 is therefore prima facie obvious.
Furthermore, Yakovenko et al. teach that more sophisticated candidate generation procedures would improve the accuracy of their neural network (p. 6, footnote). One of ordinary skill in the art would be motivated to combine the Bayesian approach to estimate probabilities of potential variants taught by Garrison et al. with the combined teachings of Yakovenko et al. and Poplin et al. in order to create a more accurate variant caller. There would be a reasonable expectation of success because Garrison et al. teach implementing their framework in a variant caller named FreeBayes (p.2, Motivation, paragraph 4). The invention of claim 13 is therefore prima facie obvious.
Conclusion
No claims are allowed.
E-mail Communications Authorization
Per updated USPTO Internet usage policies, applicant and/or applicant’s representative is encouraged to authorize the USPTO examiner to discuss any subject matter concerning the above application via Internet e-mail communications. See MPEP 502.03. To approve such communications, applicant must provide written authorization for e-mail communication by submitting the following statement via EFS-Web (using PTO/SB/439) or Central Fax (570-273-8300):
“Recognizing that Internet communications are not secure, I hereby authorize the USPTO to communicate with the undersigned and practitioners in accordance with 37 CFR 1.33 and 37 CFR 1.34 concerning any subject matter of this application by video conferencing, instant messaging, or electronic mail. I understand that a copy of these communications will be made of record in the application file.”
Written authorizations submitted to the examiner via e-mail are NOT proper. Written authorizations must be submitted via EFS-Web (using PTO/SB/439) or Central Fax (570-273-8300). A paper copy of e-mail correspondence will be placed in the patent application when appropriate. E-mails from the USPTO are for the sole use of the intended recipient, and may contain information subject to the confidentiality requirement set forth in 35 USC § 122. See also MPEP 502.03.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to TIMUR Y OLJUSKIN whose telephone number is (571)272-4006. The examiner can normally be reached Mon - Fri; 0800-1630 EST.
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.
/T.Y.O./Examiner, Art Unit 1685
/OLIVIA M. WISE/Supervisory Patent Examiner, Art Unit 1685
Prosecution Insights