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 3, 9-11, 14-15, 18, 23, 29 and 35-59 are cancelled by Applicant.
Claims 1-2, 4-8, 12-13, 16-17, 19-22, 24-28 and 30-34 are currently pending and are herein under examination.
Claims 1-2, 4-8, 12-13, 16-17, 19-22, 24-28 and 30-34 are rejected.
Claims 1-2, 4-9, 12-13, 16-17, 19-22, 24-28, and 30-34 are objected.
Priority
The instant application claims domestic benefit as a 371 filing of PCT/CN2020/132331 filed 27 Nov 2020. However, the PCT application has been submitted in Chinese and has not been translated to English. MPEP 1893.01(d) requires the PCT application to be translated into English if filed in another language. Thus, the claim to domestic benefit is not acknowledged. The effective filing date for claims 1-2, 4-8, 12-13, 16-17, 19-22, 24-28 and 30-34 is 26 May 2023, which is the date the 371 was filed.
Information Disclosure Statement
The IDSs filed 05/26/2023, 05/31/2024, 11/22/2024 and 02/20/2025 follow the provisions of 37 CFR 1.97 and have been considered in full. A signed copy of the list of references cited from these IDSs is included with this Office Action.
Drawings
The drawings filed 05/26/2023 are accepted.
Abstract
The abstract of the disclosure is objected to because it recites 181 words instead of 50 to 150 words and because it recites the following implied phrase “The present invention relates to”. A corrected abstract of the disclosure is required and must be presented on a separate sheet, apart from any other text. See MPEP § 608.01(b).
Applicant is reminded of the proper language and format for an abstract of the disclosure.
The abstract should be in narrative form and generally limited to a single paragraph on a separate sheet within the range of 50 to 150 words in length. The abstract should describe the disclosure sufficiently to assist readers in deciding whether there is a need for consulting the full patent text for details.
The language should be clear and concise and should not repeat information given in the title. It should avoid using phrases which can be implied, such as, “The disclosure concerns,” “The disclosure defined by this invention,” “The disclosure describes,” etc. In addition, the form and legal phraseology often used in patent claims, such as “means” and “said,” should be avoided.
Claim Objections
Claims 1-2, 4-9, 12-13, 16-17, 19-22, 24-28, and 30-34 are objected to because of the following informalities:
Claim 1, step 3, should recite “sequence; and”.
Claims 2, 4-8, 12-13, 16-17, 19-22, 24-28, 30 and 32-34 are missing the word “wherein” after “The method of claim X”. For example, claim 2, line 1, should recite “The method of claim 1, wherein in (1), the”. The same should be done to the other claims.
Claim 2, line 2, should recite “woman, and”.
Claim 6 recites “a.”, “b.”, “c.”, “d.” and “e.”. The periods should be replaced by parentheses. See MPEP 608.01(m).
Claim 6, step e, recites “correcte” which should be “corrected”.
Claim 7, line 2, recites “one or combination of more of” which should be either “one or more of” or “one or a combination of one or more of”.
Claim 8, line 4, recites “the phenotypic data of the pregnant women” which should recite “the phenotypic feature data of the pregnant women” to maintain claim terminology consistency.
Claim 8 recites “a.”, “b.”, “c.” and “d.” The periods should be replaced by parentheses. See MPEP 608.01(m).
Claim 12, line 4, should recite “; and (2.2)”.
Claim 21, line 4, should recite “(I); and”.
Claim 21, line 5, should recite “a ReLU activation layer”.
Claim 22, line 15, should recite “pre-module; and”.
Claim 24, line 13, should recite “kernels; and”.
Claim 25, line 4, should recite “function; and”.
Claim 25 recites “a.” and “b.” The periods should be replaced by parentheses. See MPEP 608.01(m).
Claim 31, step 3, should recite “sequence; and”.
Claim 34, line 5, should recite “1; and”.
Appropriate correction is required.
Claim Interpretation
MPEP 2113.I recites "[e]ven though product-by-process claims are limited by and defined by the process, determination of patentability is based on the product itself. The patentability of a product does not depend on its method of production. If the product in the product-by-process claim is the same as or obvious from a product of the prior art, the claim is unpatentable even though the prior product was made by a different process."
Claim 6 is being interpreted as a product by process limitation. Claim 6 further limits the read segments obtained in claim 1, step 1. The broadest reasonable interpretation of claim 1, step 1, included obtaining read segments that were previously generated. As such, claim 6 defines a GC content correction process (i.e. claim 6 steps a-e) that was previously performed to generate the product of the read segments. A prior art references that discloses read segments corrected for GC content, even if GC content corrected by a different method than claim 6, will read on the read segments.
35 USC 112(f)
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. The following claims recite the following means-plus-function limitations:
Claim 20 recites a pre-module for performing the first convolution and activation operation of the sequence feature matrix from the input to obtain a feature map.
Claim 20 recites a core module for further abstraction and feature extraction of the feature map from the pre-module and strengthening the expression ability of the neural network by effectively increasing the depth of the neural network model.
Claim 20 recites a post-module for feature abstraction and representation of the feature map from the core module.
Claim 21 recites structure for the pre-module.
Claims 22 and 24-25 recite structure for the core module.
Because these claim limitations are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, they are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof. These limitations are being interpreted as computer-implemented means-plus-function limitations which requires implementation by hardware or a combination of hardware and software, and equivalents thereof (MPEP 2181.II.B). Below is the corresponding structure of the pre-module, core module, and post-module in claim 20.
Pre-module: the structure recited in specification paras. [31] [117] and equivalents thereof will be interpreted as the structure.
Core module: the structure recited in specification paras. [33-36] [117] and equivalents thereof will be interpreted as the structure.
Post-module: the structure recited in specification para. [117] and equivalents thereof will be interpreted as the structure.
If applicant does not intend to have these limitations interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitations to avoid them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitations recites sufficient structure to perform the claimed function so as to avoid them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 112
35 USC 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
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 1-2, 4-8, 12-13, 16-17, 19-22, 24-28 and 30-34 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.
Claims dependent from a rejected claim are also rejected, unless otherwise noted.
Claim 1, step 4, recites “the fetal chromosomal abnormality state” which lacks antecedent. The preamble of claim 1 recites “fetal chromosomal abnormality” but does not recite “state”. Provide antecedent basis for the phrase.
Claim 5, line 2, recites “the unique mapping reads” which lacks antecedent basis. Provide antecedent basis.
Claim 6, step a, recites “the human reference genome” which lacks antecedent basis. Claim 1, step 2, recites “a reference genome” but does not specify it as human. Provide antecedent basis.
Claim 6, step c, recites “the fragment”. It is unclear which fragment is being referenced because claim 1, step a, recites “nucleic acid fragments”, claim 6, step a, recites “m fragments”, claim 6, step b, recites “the number of fragments” and “fragment k”, and claim 6, step c, recites “fragment k”. Clarify which fragment is being referenced.
Claim 6, step d, recites “the global scaling factor” which lacks antecedent basis. To overcome this rejection, amend to “a global scaling factor”.
Claim 6, step e, recites “the expected number of sequencing reads segments with a correcte GC content of i” which lacks antecedent basis. To overcome this rejection, amend to “an expected number of sequencing reads segments with a corrected GC content of i”.
Claim 8, line 3, recites “the missing values and null values” which lacks antecedent basis. Claim 8, lines 1-2, recites “missing value” and “null value” but does not recite a plurality of missing or null values. Provide antecedent basis for the recitation.
Claim 8, lines 6-9, recites xage, xheight, xweight, and xGW which render the claim indefinite. It is unclear how these variables are defined. For example, what does GW means? Is age, height, and weight measured in years, inches, and pounds? Clarify the definition of these variables within the claim.
Claim 12, line 2, recites “the window with length b” which lacks antecedent basis. Claim 1, step 2, recites sliding windows but does not recite a window with length b. Provide antecedent basis.
Claim 12, lines 2-3, recites “the chromosome sequence with length L of the reference genome” which lacks antecedent basis. Although claim 1, step 2, recites “at least a part of a chromosome sequence of a reference genome”, it does not recite that it has a length L. Provide antecedent basis.
Claim 12, line 5, recites “the sliding windows.” It is unclear which sliding windows are being referenced because both claim 1, step 2, and claim 12, line 3, recite “sliding windows”. Clarify which sliding windows are being referenced.
Claim 12, line 6, recites “the chromosome sequence”. It is unclear if this recitation refers to claim 1, step 2, of “a chromosome sequence” or claim 12, lines 2-3, “the chromosome sequence with length L”. Clarify which chromosome sequence is being referenced.
Claim 13 recites “the base quality and the mapping quality of read segments” which lacks antecedent basis. To overcome this rejection, amend the recitation to “a base quality and a mapping quality of read segments”.
Claim 17, lines 4-5, recites “the jth sequence eigenvalue in the ith sliding window of sample k” which lacks antecedent basis. Provide antecedent basis for the phrase.
Claim 17, lines 5-6, recites “the mean and standard deviation of the jth sequence eigenvalue in the ith sliding window of all samples” which lacks antecedent basis. Provide antecedent basis for the phrase.
Claim 20, lines 1-2, recites “the neural network model” which lacks antecedent basis. For examination, claim 20 depends on claim 19 which is where “a neural network” first appears. Provide antecedent basis or change the dependency of the claim.
Claim 20, line 5, recites “the first convolution” which lacks antecedent basis. To overcome this rejection, amend to “a first convolution”.
Claim 20, lines 8-9, recites “the expression ability” which lacks antecedent basis. To overcome this rejection, amend to “an expression ability”.
Claim 20, lines 10-11, recites “the feature abstraction representation” which lacks antecedent basis. Claim 20, lines 7-8, recites further abstraction and feature extraction of the feature map, but there is no mention of a feature abstraction representation. Provide antecedent basis or clarify what the phrase refers to.
Claim 22, lines 2, recites “the same structure” which lacks antecedent basis. To overcome this rejection, amend to “a same structure”.
Claim 22, lines 2-3, recites “the output of each residual module is the input of the next residual model” which lacks antecedent basis. Claim 22, lines 1-2, recites one or more residual “submodules” but there is no mention of residual “modules”. Provide antecedent basis or clarify what this recitation refers to.
Claim 22, line 3, recites “the residual submodule includes”. It is unclear which residual submodule is being referenced because claim 22, lines 1-2, recites “one or more residual submodules”. For examination, this recitation is being interpreted as “wherein each of the one or more residual submodules includes”. Clarify which residual submodule is being referenced.
Claim 22, step a, recites “the residual submodule includes: (A) a pre-submodule of the core module, each of which includes …” It is unclear if each residual submodule includes more than one pre-submodule because of the phrase “each of which includes”, or if the limitation means to read as “(A) a pre-submodule of the core module, wherein the pre-submodule includes …”. For examination, step A is being interpreted to mean that each residual submodule includes a single pre-submodule. Clarify the meaning “each of which includes”.
Claim 24, steps a and e, recites “the residual submodule”. It is unclear which residue module is being referenced because claim 22, lines 1-2, recites “one or more residual submodules”. Clarify which residual submodule is being referenced.
Claim 24, step b, recites “the size of the output feature map … the number of 1D convolutional kernels” which lacks antecedent basis. Provide antecedent basis.
Claim 24, step c, recites “the decline speed” which lacks antecedent basis. Provide antecedent basis.
Claim 25, line 4, recites “the activation function” which lacks antecedent basis. To overcome this rejection, amend to “an activation function”.
Claim 27, line 3, recites “the ith sequence eigenvalue” which lacks antecedent basis. To overcome this rejection, amend to “an ith sequence eigenvalue”.
Claim 27, line 4, recites “the mean of the ith sequence eigenvalues … the standard deviation of the ith sequence eigenvalues” which lacks antecedent basis. To overcome this rejection, amend to “a mean of the ith sequence eigenvalues … a standard deviation of the ith sequence eigenvalues”.
Claim 31, step 3, recites “the training data set” and “the machine learning model” which lack antecedent basis. To overcome this rejection, amend to “a training data set” and “a machine learning model”.
Claim 34, lines 1-2, recites “the training data set” which renders the claim indefinite. It is unclear which training data set is being referenced because claim 31, step 3, recites that a training data set is constructed for each pregnant woman. Clarify which training data set is being referenced.
Claim 34, lines 3-4, recites “the normalized sequence feature matrix of training sample k” which lacks antecedent basis. To overcome this rejection, amend to “a normalized sequence feature matrix of training sample k”.
35 USC 112(d)
The following is a quotation of 35 U.S.C. 112(d):
(d) REFERENCE IN DEPENDENT FORMS.—Subject to subsection (e), a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
The following is a quotation of pre-AIA 35 U.S.C. 112, fourth paragraph:
Subject to the following paragraph [i.e., the fifth paragraph of pre-AIA 35 U.S.C. 112], a claim in dependent form shall contain a reference to a claim previously set forth and then specify a further limitation of the subject matter claimed. A claim in dependent form shall be construed to incorporate by reference all the limitations of the claim to which it refers.
Claim 26 is rejected under 35 U.S.C. 112(d) or pre-AIA 35 U.S.C. 112, 4th paragraph, as being of improper dependent form for failing to further limit the subject matter of the claim upon which it depends, or for failing to include all the limitations of the claim upon which it depends.
Claim 26 fails to further limit claim 1 because claim 1, step 4, already recites “combining the sequence feature vector and the phenotypic feature vector of the pregnant woman to form a combined feature vector”.
Applicant may cancel the claim(s), amend the claim(s) to place the claim(s) in proper dependent form, rewrite the claim(s) in independent form, or present a sufficient showing that the dependent claim(s) complies with the statutory requirements.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-2, 4-8, 12-13, 16-17, 19-22, 24-28 and 30-34 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1:
Step 1 asks whether the claims recite statutory subject matter. In the instant application, claims 1-2, 4-8, 12-13, 16-17, 19-22, 24-28 and 30 recite a method, and claims 31-34 recite a method. As such, these claims recite statutory subject matter (Step 1: YES).
Step 2A, Prong 1:
Claims that recite statutory subject matter are analyzed under Step 2A, Prong 1 to determine if they recite any concepts that equate to an abstract idea, law of nature or natural phenomena. The instant claims recite the following limitations that equate to one or more categories of judicial exception:
Claim 1 recites “(1) obtaining sequencing data of cell-free nucleic acid fragments and clinical phenotypic feature data from a pregnant woman to be detected, wherein the sequencing data comprise a plurality of read segments, and the clinical phenotypic feature data of the pregnant woman to be detected form a phenotypic feature vector of the pregnant woman; (2) performing window division on at least part of a chromosome sequence of a reference genome to obtain sliding windows, counting the read segments falling within the sliding windows, and generating a sequence feature matrix of the chromosome sequence; (3) inputting the sequence feature matrix into a trained machine learning model to extract a sequence feature vector of the chromosome sequence; (4) combining the sequence feature vector and the phenotypic feature vector of the pregnant woman to form a combined feature vector, and inputting the combined feature vector into a classification detection model to obtain the fetal chromosomal abnormality state of the pregnant woman to be detected.”
Claim 2 recites “in (1), the cell-free nucleic acid fragments are derived from the peripheral blood, liver, and/or placenta of the pregnant woman, the cell-free nucleic acid fragments are cell-free DNA.”
Claim 4 recites “in (1), the sequencing data are derived from ultra-low depth sequencing”.
Claim 5 recites “in (1), the read segments are aligned to the reference genome to obtain the unique mapping reads, the subsequent steps are carried out with the unique mapping reads”
The entirety of claim 6.
Claim 7 recites “in (1), the phenotypic feature data of the pregnant woman are selected from one or combination of more of: age, gestational week, height, weight, BMI, biochemical test results of prenatal examination, ultrasonic diagnosis results, and cell-free fetal DNA concentration in plasma.”
Claim 8 recites “in (1), the phenotypic feature data of the pregnant woman are subjected to outlier processing, missing value processing and/or null value processing, and the missing values and null values are padded by missForest algorithm, the phenotypic data of the pregnant woman sample will be judged as outliers if the following records appear:”
Claim 12 recites “(2.2) counting the read segments falling within each of the sliding windows, and generating a sequence feature matrix of the chromosome sequence.”
Claim 13 recites “in (2), the sequence feature matrix includes the number, the base quality and the mapping quality of read segments within the sliding windows.”
Claim 16 recites “in (2), the sequence feature matrix is:
PNG
media_image1.png
46
134
media_image1.png
Greyscale
wherein h represents the number of sliding windows, w represents the number of sequence features within a single sliding window, and Xij represents the t 1 sequence eigenvalue in the ith sliding window.”
Claim 17 recites “in (3), the sequence feature matrix is normalized, the sequence feature matrix is normalized by using formula (I):
PNG
media_image2.png
116
173
media_image2.png
Greyscale
wherein Zz5) is the normalized sequence feature matrix of sample k:Xz5) represents the jth sequence eigenvalue in the ith sliding window of sample k~i,j and CJi,j represent the mean and standard deviation of the jth sequence eigenvalue in the ith sliding window of all samples, respectively.”
Claim 26 recites “in (4), the combined feature vector is obtained by combining the sequence feature vector and the phenotypic feature vector of the pregnant woman.”
Claim 27 recites “in (4), the combined feature vector xis normalized by:
PNG
media_image3.png
57
128
media_image3.png
Greyscale
wherein xf is the ith sequence eigenvalue of the normalized combined feature vector x, xi is the ith sequence eigenvalue of the combined feature vector, μi is the mean of the ith sequence eigenvalues of the combined feature vector x, and (Ji is the standard deviation of the ith sequence eigenvalues of the combined feature vector.”
Claim 28 recites “in (4), the classification detection model is an ensemble learning model, the ensemble learning model is an ensemble learning model based on Stacking or Majority Voting.”
Claim 30 recites “the chromosomal abnormality includes at least one or more of: trisomy 21 syndrome, trisomy 18 syndrome, trisomy 13 syndrome, 5p-syndrome, chromosomal microdeletion and chromosomal microduplication.”
Claim 31 recites “(1) obtaining sequencing data of cell-free nucleic acid fragments and clinical phenotypic feature data from a plurality of pregnant women, wherein the sequencing data comprise a plurality of read segments, and the fetal chromosomal state of each of the pregnant women is known, and the clinical phenotypic feature data of each of the pregnant women form a phenotypic feature vector of the pregnant woman; (2) for each of the pregnant women, performing window division on at least part of a chromosome sequence of a reference genome to obtain sliding windows, counting the read segments falling within the sliding windows, and generating a sequence feature matrix of the chromosome sequence; (3) for each of the pregnant women, constructing the training data set using the sequence feature matrix and the fetal chromosomal state, and training the machine learning model to extract a sequence feature vector of the chromosome sequence; (4) combining the sequence feature vector and the phenotypic feature vector of each of the pregnant women to form a combined feature vector, and training the classification model with the combined feature vectors and the fetal chromosomal states of the pregnant women to obtain a trained classification detection model.”
Claim 32 recites “the fetal chromosomal state of each of the pregnant women is one or more of: normal diploid, chromosomal aneuploid, partial monosomy syndrome, chromosomal microdeletion and chromosomal microduplication.”
Claim 33 recites “the number of the pregnant women is greater than 10, and the ratio of the number of fetuses with normal diploid to that of fetuses with chromosomal aneuploidy is to 2.”
Claim 34 recites “in (3), the training data set is represented as:
PNG
media_image4.png
113
522
media_image4.png
Greyscale
wherein N represents the number of training samples, and N is an integer~ 1; Zz5) is the normalized sequence feature matrix of training sample k, and kE [l,N], wherein i is an integer~ 1 and j is an integer>1; the chromosomal abnormality includes at least one or more of: trisomy 21 syndrome, trisomy 18 syndrome, trisomy 13 syndrome, Sp-syndrome, chromosomal microdeletion and chromosomal microduplication.”
Limitations reciting a mental process.
Claims 1-2, 4-8, 13, 16-17, 26-27 and 30-34 contain limitations recited at such a high level of generality that they equate to a mental process because they are similar to the concepts of collecting information, analyzing it, and displaying certain results of the collection and analysis in Electric Power Group, LLC, v. Alstom (830 F.3d 1350, 119 USPQ2d 1739 (Fed. Cir. 2016)), which the courts have identified as concepts that can be practically performed in the human mind. The paragraphs below discuss the broadest reasonable interpretation (BRI) of the limitations in these claims that recite a mental process.
Claim 1, step 1, and claim 31, step 1, include collecting previously generated information. Claim 1, step 2, and claim 31, step 2, include writing on pen and paper a portion of a chromosome sequence, selecting locations to divide the sequence into windows, aligning sequence reads to the sequence, counting a number of sequence reads in each window, then generating a vector containing read counts for each window. Claim 1, step 3, and claims 31, step 3, includes inputting a matrix into a trained linear regression to generate a vector output, which can be done on pen and paper. Claim 1, step 4, and claim 31, step 4, include creating a vector by combining two separate vectors on pen and paper, then inputting the combined vector into a trained logistic regression which outputs probabilities of a particular chromosomal abnormality.
Claim 5 includes aligning sequences to a genome and determining if they are unique which can be done on paper. The claim does not require any software implementation. Claim 6, step a, includes selecting data. Steps b-e include performing a calculation on pen and paper. Claim 8 recites outlier processing by judging if phenotypic data falls outside an interval which requires mental evaluation. Claim 26 includes combing two vectors, which can be done on pen and paper. Claim 16 includes organizing data into a matrix. Claims 17 and 27 include performing the calculations of the functions on pen and paper. Claim 34 includes organizing data.
Claims 2, 4, 7 are included in the mental process of claim 1, step 1, of collecting information because they limit the source of the previously generated information. Claim 13 is included in the judicial exception in claim 1, step 2, because it further limits the sequence feature matrix. Claims 30 and 32 limit the output of the classification model in claim 1, step 4, and claim 30, step 4, which themselves recite a judicial exception. Claim 33 is included in the judicial exception in claim 30, step 1, of collecting information.
Limitations reciting a mathematical concept.
Claims 1, 6, 8, 16-17, 26-28, 31 and 34 recite limitations that equate to a mathematical concept because they are similar to the concepts of organizing and manipulating information through mathematical correlations in Digitech Image Techs., LLC v Electronics for Imaging, Inc. (758 F.3d 1344, 111 U.S.P.Q.2d 1717 (Fed. Cir. 2014)), which the courts have identified as mathematical concepts. The paragraphs below discuss the broadest reasonable interpretation (BRI) of the limitations in these claims that recite a mathematical concept.
Claim 1, step 2, and claim 31, step 2, include math because generating a sequence feature matrix is a mathematical correlation of numbers, and specification para. [27] recites that the sequence feature matrix is a mathematical formula. Claim 1, step 3, includes a mathematical calculation and function because a numerical matrix is inputted to a machine learning model to calculate a vector of numbers, such as a linear regression that uses matrix-vector multiplication to output a 1D vector.
Claim 1, step 2, and claim 31, step 2, recite math by counting reads and generating a matrix which is a mathematical correlation of data.
Claim 31, step 3, of training a machine learning model includes using a gradient descent in a linear regression to output a vector of numbers. Claim 31, step 4, of training a classification model includes training a logistic regression using gradient descent. Claim 6, steps b-e, and claims 17 and 27 recite mathematical equations that require calculations. Claim 8 recites missing or null value processing by padding using a missForest algorithm which requires an equation and calculation to determine values. See specification para. [128] for the mathematical algorithm. Claim 16 recites an equation and requires mathematical correlations. Claim 26 includes organizing and manipulating numbers into a vector. Claim 28 includes an ensemble of logistic regression models, as recited in specification para. [40]. Claim 34 organizes and manipulates the training data using mathematical correlations.
As such, claims 1-2, 4-8, 12-13, 16-17, 19-22, 24-28 and 30-34 recite an abstract idea (Step 2A, Prong 1: YES).
Additional Elements:
Once limitations have been identified that recite a judicial exception, the claims are evaluated for additional elements. The additional elements are then analyzed under Step 2A, Prong 2 then Step 2B. The instant claims recite the following additional elements:
Claim 12 recites “(2.1) using the window with length b to overlap and slide the chromosome sequence with length L of the reference genome at step size of t to obtain sliding windows, wherein b is a positive integer and b=[10000,10000000], t is any positive integer, L is a positive integer and L>b;”
Claim 19 recites “in (3), the trained machine learning model is a neural network model or an AutoEncoder model;”
Claim 20 recites “the neural network model is a deep neural network model, and the structure of the deep neural network model includes: an input layer, for receiving the sequence feature matrix; a pre-module, being connected with the input layer, for performing the first convolution and activation operation of the sequence feature matrix from the input layer, to obtain a feature map; a core module, being connected with the pre-module, for further abstraction and feature extraction of the feature map from the pre-module, and strengthening the expression ability of the neural network by effectively increasing the depth of the neural network model; a post-module, being connected with the core module, for the feature abstraction representation of the feature map from the core module; a first global average pooling layer, being connected with the post-module, for vectorizing the feature map of the feature abstraction representation and outputting the sequence feature vector of the chromosome sequence.”
Claim 21 recites “the pre-module includes: (I) a 1D convolution layer; (II) a batch normalization layer, being connected with the 1D convolution layer described in (I); (III) ReLU activation layer, being connected with the batch normalization layer described in (II).”
Claim 22 recites “the core module consists of one or more residual submodules with the same structure, wherein the output of each residual module is the input of the next residual module, the residual submodule includes: (A) a pre-submodule of the core module, each of which includes a 1D convolution layer, a Dropout layer connected with the 1D convolution layer, a batch normalization layer connected with the Dropout layer, and a ReLU activation layer connected with the batch normalization layer;(B) a first 1D average pooling layer, being connected with the pre-submodule of the core module described in (A): (C) a Squeeze-Excite module, and/or a Spatial Squeeze-Excite module, being connected with the first 1D average pooling layer described in (B): (D) a first Addition layer, being connected with the Squeeze-Excite module and/or Spatial Squeeze-Excite module described in (C): (E) a second 1D average pooling layer, being connected with the ReLU activation layer in the pre-module: (F) a second Addition layer, being connected with the first Addition layer described in (D) and the second 1D average pooling layer described in (E).”
Claim 24 recites “the Squeeze-Excite module includes: (a) a second global average pooling layer, being connected with the first lD average pooling layer in (B) of the residual submodule; (b) a Reshape layer, being connected with the second global average pooling layer described in (a), and the size of the output feature map of the Reshape layer is 1 X f, wherein f is the number of lD convolution kernels; (c) a first fully connected layer, being connected with the Reshape layer described in (b), and the number of output neurons of the first fully connected layer is L, wherein f is the rsE number of lD convolution kernels, and rsE is the decline speed of the Squeez-Excite module; (d) a second fully connected layer, being connected with the first fully connected layer described in (c), and the number of output neurons of the second fully connected layer is f, wherein f is the number of lD convolution kernels; (e) a Multiply layer, being connected with the second fully connected layer described in (d) and the first lD average pooling layer in (B) of the residual submodule.”
Claim 25 recites “the Spatial Squeeze-Excite module includes: a. a 1 x 1 1D convolution layer, being connected with the first 1D average pooling layer in (B), which uses sigmoid function as the activation function; b. a Multiply layer, being connected with the first 1D average pooling layer in (B) and the 1 x 1 1D convolution layer in a.”
These above recited additional elements are analyzed below under both Step 2A, Prong 2 and Step 2B:
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). The judicial exception is not integrated into a practical application because the claims do not recite additional elements that reflect an improvement to a computer, technology, or technical field (MPEP § 2106.04(d)(1) and 2106.5(a)), require a particular treatment or prophylaxis for a disease or medical condition (MPEP § 2106.04(d)(2)), implement the recited judicial exception with a particular machine that is integral to the claim (MPEP § 2106.05(b)), effect a transformation or reduction of a particular article to a different state or thing (MPEP § 2106.05(c)), nor provide some other meaningful limitation (MPEP § 2106.05(e)). Rather, the claims include limitations that equate to an equivalent of the words “apply it” and/or to instructions to implement an abstract idea on a computer (MPEP § 2106.05(f)), insignificant extra-solution activity (MPEP § 2106.05(g)), and field of use limitations (MPEP § 2106.05(h)). The paragraphs below discuss the additional elements recited above in the instant claims.
Claim 12 equates to insignificant extra-solution activity of necessary data gathering because it gathers data necessary to perform the judicial exception in claim 1, step 2, of counting reads.
Claim 19 recites “the trained machine learning model is … an AutoEncoder model”. The BRI of this limitation includes mere instructions to implement an abstract idea on a generic computer. MPEP 2106.05(f) provides the following considerations for determining whether a claim simply recites a judicial exception with the words “apply it” (or an equivalent), such as mere instructions to implement an abstract idea on a computer: (1) whether the claim recites only the idea of a solution or outcome i.e., the claim fails to recite details of how a solution to a problem is accomplished; (2) whether the claim invokes computers or other machinery merely as a tool to perform an existing process; and (3) the particularity or generality of the application of the judicial exception. The two following paragraphs provide analysis under these considerations.
The AutoEncoder performs the abstract idea of step 3 in claim 1. The AutoEncoder is used to generally apply the abstract idea without placing any limits on how the AutoEncoder functions. Rather, this limitation only recites the outcome of the abstract idea of step 3 in claim 1 without including any details about how the abstract idea is accomplished. See MPEP 2106.05(f).
This limitation also merely indicates a field of use or technological environment in which the abstract idea is performed. Although limiting the abstract idea, this limitation merely confines the use of the abstract idea to a particular technological environment (AutoEncoder models) and thus fails to add an inventive concept to the claims. See MPEP 2106.05(h).
Claims 20-22 and 24-25 are optional limitations. MPEP 2143.03 recites “[l]anguage that suggests or makes a feature or step optional but does not require that feature or step does not limit the scope of a claim under the broadest reasonable claim interpretation.” As discussed in section 35 USC 112(b), claim 20 depends on claim 19 where the trained model is either a neural network (NN) or an AutoEncoder. Claim 20 and its dependents further limit the NN but do not require that the trained model be the NN. Subsequently, claims 20-22 and 24-25 are not required when claim 19 requires the trained model be the AutoEncoder. As such, claims 20-22 and 24-25 are not required and are not being evaluated on their merits under Step 2A, Prong 2.
As such, claims 1-2, 4-8, 12-13, 16-17, 19-22, 24-28 and 30-34 are directed to an abstract idea (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). These claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because these claims recite additional elements that equate to instructions to apply the recited exception in a generic way and/or in a generic computing environment (MPEP § 2106.05(f)) and to well-understood, routine and conventional (WURC) limitations (MPEP § 2106.05(d)). The paragraphs below discuss the additional elements recited above in the instant claims.
Claims 20-22 and 24-25 are not required, and are thus not being examined on their merits under Step 2B for the same reasons discussed above in section Step 2A, Prong 2.
Claim 12 recites well-understood, routine, and conventional (WURC) limitations of sliding a window over a sequence of a reference genome at a step size to obtain sliding windows as taught by Chen et al. (“Chen”; Prenatal diagnosis 33, no. 6 (2013): 584-590) and Becker et al. (“Becker”; Molecular cell 68, no. 6 (2017): 1023-1037).
Chen noninvasively detects fetal deletions and duplications using sequencing from maternal plasma DNA (abstract) (Figure 1) (Table 1). A human reference genome was divided into 308,789 sliding windows where adjacent windows shared 99% overlap (pg. 585, col. 2, last para.) (pg. 587, col. 2, para. 2). An average window length was 0.94 ± 0.68 Mb (587, col. 1, para. 2).
Becker isolates and maps heterochromatin (summary). Enriched genomic domains were called using a 10-kb sliding window algorithm with a sliding step of 500bp (pg. e11, para. 5). For every 10-kb window in the reference genome, a number of reads falling into a window were counted (pg. e11, para. 5). The reference genome was divided into 10-kb sliding windows (pg. e12, para. 3).
Claim 19 equates to instructions to “apply” the abstract idea, which cannot provide an inventive concept. See MPEP 2106.05(f). See above in section Step 2A, Prong 2 for further discussion.
When these additional elements are considered individually and in combination, they do not provide an inventive concept because they equate to mere instructions to apply the judicial exception and to WURC limitation as taught by Chen and Becker. Therefore, these additional elements do not transform the claimed judicial exception into a patent-eligible application of the judicial exception and do not amount to significantly more than the judicial exception itself (Step 2B: No).
As such, claims 1-2, 4-8, 12-13, 16-17, 19-22, 24-28 and 30-34 are not patent eligible.
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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
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 1-2, 4-7, 12, 19-22, 24-26, 28 and 30 are rejected under 35 USC 103 for being unpatentable over Jensen et al. (“Jensen”; US 2017/0342477 A1) in view of Chen et al. (“Chen”; Prenatal diagnosis 33, no. 6 (2013): 584-590), Notley et al. (“Notley”; arXiv preprint arXiv:1805.02294 (2018)), and Kim et al. (“Kim”; Obstetrics & gynecology science 56, no. 3 (2013): 160).
The bold and italicized text below are the limitations of the instant claims, and the italicized text serves to map the prior art onto the instant claims.
Claim 1:
A method of detecting a fetal chromosomal abnormality, comprising:
Jensen discloses non-invasive fetal aneuploidy detection methods [6].
(1) obtaining sequencing data of cell-free nucleic acid fragments and clinical phenotypic feature data from a pregnant woman to be detected, wherein the sequencing data comprise a plurality of read segments, and the clinical phenotypic feature data of the pregnant woman to be detected form a phenotypic feature vector of the pregnant woman;
Jensen discloses non-invasive prenatal testing based on sequencing cell-free DNA from a pregnant woman (sequencing data of cell-free nucleic acid fragments from a pregnant women) [25] [54] [76-77]. Sequencing produces read segments [34] [55]. Fetal aneuploidy detection is based on sequencing data and additional information such as maternal age (clinical phenotypic feature data from a pregnant woman) [73].
However, Jensen does not teach that maternal age forms a feature vector.
It would have been prima facie obvious over the teachings of Jensen to have transformed maternal age into a feature vector. This is because maternal age is used in combination with read counts [54] for fetal aneuploidy detection [191], and because classification models used for the detection include neural networks and SVMs [68], which take vectors as input. There would have been a reasonable expectation of success to convert maternal age into a vector because machine learning models use vectors as input.
(2) performing window division on at least part of a chromosome sequence of a reference genome to obtain sliding windows, counting the read segments falling within the sliding windows, and generating a sequence feature matrix of the chromosome sequence;
Jensen discloses read counts derived from a target chromosome are organized into matrices (a sequence feature matrix of the chromosome sequence) [56] [60] [285]. Reads are aligned to a reference genome to perform counting (counting the read segments) [54-55]. A specific chromosome such as 18 and 21 can be targeted (at least a part of a chromosome) [25].
However, Jensen does not count reads falling within sliding windows generated from a chromosome of a reference genome.
Chen noninvasively detects fetal deletions and duplications using sequencing from maternal plasma DNA (abstract) (Figure 1) (Table 1). A human reference genome was divided into sliding windows (pg. 585, col. 2, last para.), and the number of unique reads within each window were counted (pg. 586, col. 1, last para.) (pg. 587, col. 1, last para. – col. 2, para. 2).
It would have been prima facie obvious to modify Jensen’s method of counting reads to detect fetal chromosomal aneuploidy by using the sliding window method for read counting of Chen. Motivation is taught by Chen who teaches that the sliding window strategy accurately detects copy-number variations with relativity low coverage whole genome sequencing, which enables detection of large fetal chromosomal deletion/duplication using less sequencing data of maternal plasma (pg. 585, col. 2, para. 1). There would have been a reasonable expectation of success to use the sliding window method that outputs read counts because Jensen uses read counts to detect fetal chromosomal aneuploidy [257], and because Chen discloses that the sliding window method detects fetal deletion/duplication (abstract).
(3) inputting the sequence feature matrix into a trained machine learning model to extract a sequence feature vector of the chromosome sequence;
Jensen teaches read counts organized into matrices are inputted into a trained classification model to determine fetal aneuploidy, which may be a neural network (inputting the sequence feature matrix into a trained machine learning model) [16] [56] [68] [197] [285] [391].
However, Jensen does not extract a sequence feature vector from the trained classification model.
Notley extracts features from numeric data using a neural network then inputs the extracted features into an SVM or KNN to enhance classification (abstract). The NN was trained then used for experimental results (pg. 3, col. 2, para. 2). The extracted features were taken from the fully connected layer of the NN which necessitates them being a feature vector.
It would have been prima facie obvious to have modified Jensen who inputs read counts into a classifier to detect fetal chromosomal abnormality by using a NN to extract a read count feature vector from a matrix of read counts as taught by Notley. Motivation is taught by Notley who recites that using a neural network to extract features for use in other classification models is a viable technique to produce strong learning models (6, col. 2, para. 2). This aligns with Jensen who teaches that read counts can be processed by multiple model types to detect abnormality [197]. There would have been a reasonable expectation of success because Notley teaches that the NN can extract features from numeric datasets for use in classification (abstract), wherein Jensen uses numeric data of read counts for classification [257] and uses neural networks to process data [197].
(4) combining the sequence feature vector and the phenotypic feature vector of the pregnant woman to form a combined feature vector, and inputting the combined feature vector into a classification detection model to obtain the fetal chromosomal abnormality state of the pregnant woman to be detected.
Jensen detects fetal chromosomal abnormalities from read counts [257] and states that maternal age is used as an additional variable for abnormality detection [73] [191]. However, Jensen does not generate a combined feature vector from a sequence feature vector and a phenotypic feature vector to then input the combined feature vector into a classification model.
As discussed above, Jensen in view of Notley teach the sequence feature vector as features extracted from a NN, and the teachings of Jensen render obvious using maternal age as a phenotypic feature vector. Notley teaches extracting features from a NN to then input the extracted features into a second classification model (abstract).
Kim associates maternal age-specific rates of fetal chromosomal abnormalities in pregnant women of advanced maternal age (title). Kim teaches that advanced maternal age is associated with fetal chromosomal abnormalities (pg. 160, col. 1, last para. – pg. 161, col. 1, para. 1).
It would have been prima facie obvious, after generating the sequence feature vector of Jensen and Notley, to have concatenated the sequence feature vector with the maternal age feature vector of Jensen, then input the concatenated feature vector into a second classification model distinct from a trained NN feature extractor as taught by Notley. Motivation for doing so is that Notley teaches features extracted from a NN are used to as input for a subsequent classification model such as a SVM or KNN, thereby enhancing classification abilities (abstract). This aligns with Jensen who teaches that classification can be performed with a NN, SVM, or KNN [68], and that maternal age is used as a variable to detect fetal chromosomal abnormalities [191]. Kim teaches motivation for adding maternal age into the input vector for classification by stating that there is a correlation between advance maternal age and fetal chromosomal abnormalities (pg. 160, col. 1, last para. – pg. 161, col. 1, para. 1).
There would have been a reasonable expectation of success because a classification model such as a KNN or SVM takes vector, wherein the input vector would contain read counts and maternal age, both of which can be used to detect fetal chromosomal abnormality as taught by Jensen [191] [257].
Claim 2:
Jensen teaches cfDNA is derived from placenta or peripheral blood from a pregnant woman [5] [76] [78] (claims 2-3)
Claim 4:
Jensen teaches sequencing depth may be 0.2 to 01 [138].
Claim 5:
Jensen teaches “a read may uniquely or non-uniquely map to portions in a reference genome. A read is considered as ‘uniquely mapped’ if it aligns with a single sequence in the reference genome. A read is considered as ‘non-uniquely mapped’ if it aligns with two or more sequences in the reference genome. In some embodiments, non-uniquely mapped reads are eliminated from further analysis (e.g. quantification)” [153].
Claim 6:
Jensen states that sequence reads are corrected for GC content [191] [193]. As discussed in Claim Interpretation, claim 6 recites a product by process where the product of read segments was previously GC content corrected by steps a-e. MPEP 2113.I recites "[e]ven though product-by-process claims are limited by and defined by the process, determination of patentability is based on the product itself. The patentability of a product does not depend on its method of production. If the product in the product-by-process claim is the same as or obvious from a product of the prior art, the claim is unpatentable even though the prior product was made by a different process." In the instant case, the GC corrected reads of Jensen read on the GC corrected reads of claim 6 even though produced by a different method because the product is the same.
Claim 7:
Jensen teaches maternal age and fetal fraction are used as variables to detect fetal chromosomal abnormalities [73] [191] [407].
Claim 12:
Jensen organizes data such as read counts derived from a target chromosome into matrices (generating a sequence feature matrix of the chromosome sequence) [56] [60] [285]. Reads are aligned to a reference genome and counted (counting the read segments) [54-55]. A specific chromosome such as 18 and 21 can be targeted (at least a part of a chromosome) [25].
However, Jensen does not count reads falling within sliding windows generated from a chromosome of a reference genome.
Chen noninvasively detects fetal deletions and duplications using sequencing from maternal plasma DNA (abstract) (Figure 1) (Table 1). A human reference genome was divided into 308,789 sliding windows where adjacent windows shared 99% overlap (pg. 585, col. 2, last para.) (pg. 587, col. 2, para. 2). An average window length was 0.94 ± 0.68 Mb (587, col. 1, para. 2). The number of unique reads within each window were counted (counting the read segments falling within each of the sliding windows) (pg. 586, col. 1, last para.) (pg. 587, col. 1, last para. – col. 2, para. 2).
It would have been prima facie obvious to modify Jensen’s method of counting reads to detect fetal chromosomal aneuploidy by using the sliding window method for read counting of Chen. Motivation is taught by Chen who teaches that the sliding window strategy accurately detects copy-number variations with relativity low coverage whole genome sequencing, which enables detection of large fetal chromosomal deletion/duplication using less sequencing data of maternal plasma (pg. 585, col. 2, para. 1). There would have been a reasonable expectation of success to use the sliding window method that outputs read counts because Jensen uses read counts to detect fetal chromosomal aneuploidy [257], and because Chen discloses that the sliding window method detects fetal deletion/duplication (abstract).
Claim 19:
Jensen discloses a trained neural network [391] [412].
Claims 20-22 and 24-25:
Claims 20-22 and 24-25 are optional limitations. MPEP 2143.03 recites “[l]anguage that suggests or makes a feature or step optional but does not require that feature or step does not limit the scope of a claim under the broadest reasonable claim interpretation.” As discussed in section 35 USC 112(b), claim 20 depends on claim 19 where the trained model is either a neural network (NN) or an AutoEncoder. Claim 20 and its dependents further limit the NN but do not require that the trained model be the NN. Subsequently, claims 20-22 and 24-25 are not required (i.e. optional) when claim 19 requires the trained model be the AutoEncoder. As such, claims 20-22 and 24-25 are not required and
are rejected for their dependency on rejected claim 19.
Claim 26:
Claim 26 is rejected for the same reasons applied above in claim 1, step 4.
Claim 28:
Jensen teaches multiple models of the same or different type can analyze the data sets to produce the classification [197] [241].
Claim 30:
Jensen teaches that the fetal chromosomal abnormality can any of the abnormalities listed in instant claim 30. [193] and [332].
Claim 8 is rejected under 35 USC 103 for being unpatentable over Jensen et al. (“Jensen”; US 2017/0342477 A1) in view of Chen et al. (“Chen”; Prenatal diagnosis 33, no. 6 (2013): 584-590), Notley et al. (“Notley”; arXiv preprint arXiv:1805.02294 (2018)), and Kim et al. (“Kim”; Obstetrics & gynecology science 56, no. 3 (2013): 160), as applied above to claim 1, and in further view of Kwak et al. (“Kwak”; Korean journal of anesthesiology 70, no. 4 (2017): 407) and Stekhoven et al. (“Stekhoven”; NPL ref 16 on IDS 05/26/2026; Bioinformatics 28, no. 1 (2012): 112-118).
The limitations of claim 1 have been taught in the rejection above by Jensen, Chen, Notley, and Kim.
The bold and italicized text below are the limitations of the instant claims, and the italicized text serves to map the prior art onto the instant claims.
Claim 8:
Jensen discloses maternal age as a phenotypic feature as discussed in claim 1, step 1. However, Jensen does not perform outlier processing, missing/null values processing, or padding missing/null values with missForest.
Kwak discloses methods for handling missing values which includes complete case analysis, available case analysis, and imputation analysis (missing value and/or null value processing) (pg. 409, col. 1). Kwak discloses methods for treating outliers which includes trimming, winsorization and robust estimation method (outlier processing) (pg. 410, col. 1).
It would have been prima facie obvious to perform outlier processing and missing/null value processing on the phenotypic data of Jensen using the methods described by Kwak. Motivation is taught by Kwak who teaches that missing values and outliers are frequently encountered in data collection (pg. 407, col. 1, para. 1) and should be adjusted for using different methods (pg. 407, col. 1, para. 2 – col. 2, para. 2). There would have been a reasonable expectation of success because Kwak teaches that handling missing values and outliers during data collection for natural science experiments is common (pg. 407, col. 1, para. 1) and because Kwak discloses specific methods to perform the data management techniques.
However, Jensen and Kwak do not pad missing/null values with missForest.
Stekhoven discloses missForest for non-parametric missing value imputation for mixed-typed data (the missing values and null values are padded by missForest algorithm) (title) (algorithm 1 on pg. 113). It would have been prima facie obvious to impute missing values in the phenotypic data in Jensen using Stekhoven’s missForest. Motivation for doing so is that imputation of missing values is a crucial step in data analysis, particularly in high-throughput data as taught by Stekhoven (abstract) (pg. 112, sec. 1, para. 1). There would have been a reasonable expectation of success to impute missing variables in Jensen using missForest because missForest works on both continuous and categorical variables of any type of data (abstract) (pg. 117, sec. 5, para. 1).
The following limitations in claim 8 recites a contingent limitation: “the phenotypic data of the pregnant women sample will be judged as outliers if the following records appear: … and these outliers are set as null values.” MPEP 2114.04.II recites “[t]he broadest reasonable interpretation of a method (or process) claim having contingent limitations requires only those steps that must be performed and does not include steps that are not required to be performed because the condition(s) precedent are not met.” In the instant case, the phenotypic data is not judged as outliers when the records do not appear. As such, these limitations are not required.
Claim 13 is rejected under 35 USC 103 for being unpatentable over Jensen et al. (“Jensen”; US 2017/0342477 A1) in view of Chen et al. (“Chen”; Prenatal diagnosis 33, no. 6 (2013): 584-590), Notley et al. (“Notley”; arXiv preprint arXiv:1805.02294 (2018)), and Kim et al. (“Kim”; Obstetrics & gynecology science 56, no. 3 (2013): 160), as applied above to claim 1, and in further view of Szatkiewicz et al. (“Szatkiewicz”; Nucleic acids research 41, no. 3 (2013): 1519-1532).
The limitations of claim 1 have been taught in the rejection above by Jensen, Chen, Notley, and Kim.
The bold and italicized text below are the limitations of the instant claims, and the italicized text serves to map the prior art onto the instant claims.
Claim 13:
Jensen discloses data such as read counts derived from a target chromosome are organized into matrices (a sequence feature matrix) [56] [60] [285], and that that maternal age is used as an additional variable for abnormality detection [73] [191]. Chen discloses a counting a number of read segments falling within each sliding window (pg. 585, col. 2, last para.) (pg. 586, col. 1, last para.) (pg. 587, col. 1, last para. – col. 2, para. 2). However, Jensen and Chen do not include base quality or mapping quality of read segments within sliding windows as part of the matrix.
Szatkiewicz improves detection of copy number variation (CNV) by simultaneous bias correction and read-depth correction (abstract). Szatkiewicz recites that input data for CNV detection “are a triplet of [read depth] signal, GC content and mappability score computed in sliding windows tiled along the genome” (pg. 1521, col. 1, para. 3), and recites “[u]sing reads passing the QC filter, the [read depth] signal is calculated as the number of sequenced DNA fragments in sliding windows, ensuring each fragment is counted only once” (pg. 1521, col. 1, para. 4).
It would have been prima facie obvious to modify the input matrix of Jensen and Chen comprising read counts to also include read-depth filtered by base QC and mappability score of reads in each sliding window as taught by Szatkiewicz for detecting CNV. Motivation is that Szatkiewicz recites their method simultaneously corrects for multiple sources of bias and infers CNV from read depth (pg. 1529, col. 2, last para.). There would have been a reasonable expectation of success to modify the input matrix of Jense and Chen comprising read counts to contain other variables for bias correction because Jensen states that data processing and normalization steps include bias relationships [191] [203].
Claims 31-34 are rejected under 35 USC 103 for being unpatentable over Enrich et al. (“Enrich”; WO 2019/191319 A1; FOR ref 7 on IDS filed 05/06/2023) in view of Chen et al. (“Chen”; Prenatal diagnosis 33, no. 6 (2013): 584-590).
The bold and italicized text below are the limitations of the instant claims, and the italicized text serves to map the prior art onto the instant claims.
Claim 31:
(1) obtaining sequencing data of cell-free nucleic acid fragments and clinical phenotypic feature data from a plurality of pregnant women, wherein the sequencing data comprise a plurality of read segments, and the fetal chromosomal state of each of the pregnant women is known, and the clinical phenotypic feature data of each of the pregnant women form a phenotypic feature vector of the pregnant woman;
Enrich discloses deep-learning based methods for prenatal testing (title). Enrich teaches “a) obtaining a biological sample from a subject, wherein the biological sample comprises nucleic acid molecules; b) sequencing at least a portion of the nucleic acid molecules to produce a set of sequencing reads” [7]. Fetal cell-free nucleic acids are collected from a pregnant subject [23] [399]. A training data set contains training pregnant subjects’ age, weight, and gestational age (clinical phenotypic feature data) [8-7] [209]. Training subjects have a known fetal aneuploidy or euploid state (fetal chromosomal state of the pregnant women is known) [8] [292] [309]. Input data for a machine learning algorithm is a vector [90].
(2) for each of the pregnant women, performing window division on at least part of a chromosome sequence of a reference genome to obtain sliding windows, counting the read segments falling within the sliding windows, and generating a sequence feature matrix of the chromosome sequence;
Enrich recites “step (c) includes alignment of the set of sequencing reads relative to a reference sequence, and counting the number of sequencing reads that are aligned with each of a series of pre-define subsections of the reference sequence, thereby generating a set of numeric values that form all or part of the input data set” [8]. Output of a model is a matrix [91] [368] (Fig. 16 and 31).
However, Enrich does not count reads within sliding windows generated from window division of a reference chromosome.
Chen noninvasively detects fetal deletions and duplications using sequencing from maternal plasma DNA (abstract) (Figure 1) (Table 1). A human reference genome was divided into sliding windows (pg. 585, col. 2, last para.), and the number of unique reads within each window were counted (pg. 586, col. 1, last para.) (pg. 587, col. 1, last para. – col. 2, para. 2).
It would have been prima facie obvious to modify Enrich’s method of counting reads to detect fetal chromosomal aneuploidy by using the sliding window method for read counting of Chen. Motivation is taught by Chen who teaches that the sliding window strategy accurately detects copy-number variations with relativity low coverage whole genome sequencing, which enables detection of large fetal chromosomal deletion/duplication using less sequencing data of maternal plasma (pg. 585, col. 2, para. 1). There would have been a reasonable expectation of success to use the sliding window method that outputs read counts because Enrich uses read counts to detect fetal chromosomal aneuploidy [187] [291] (Figure 16), and because Chen discloses that the sliding window method detects fetal deletion/duplication (abstract).
(3) for each of the pregnant women, constructing the training data set using the sequence feature matrix and the fetal chromosomal state, and training the machine learning model to extract a sequence feature vector of the chromosome sequence;
Enrich shows in Figure 16 the first NN taking a sequence read matrix as input then outputting a class probability vector per sequencing read [50], which may be count normalized SoftMax probability sum vectors [66]. Raw sequencing data is used to train an ANN to produce an output of a set of normalized bin count data vectors or read probabilities vectors (a sequence feature vector) [179]. Input into ML can be organized as matrices [90]. The training data contains the count vectors along with a label of its class which included fetal chromosomal state [179] [381].
(4) combining the sequence feature vector and the phenotypic feature vector of each of the pregnant women to form a combined feature vector, and training the classification model with the combined feature vectors and the fetal chromosomal states of the pregnant women to obtain a trained classification detection model.
Enrich recites “[a]n input vector comprising the normalized bin count data is applied to the input layer of the artificial neural network, and mapped to an output value (e.g., a sample classification result) by the ANN after the latter has been trained using one or more training data sets that comprise the normalized bin count data for a plurality of known euploid and/or aneuploid samples … the training data sets may comprise additional input and/or output values” [185]. The additional input data includes maternal age, gestational age, and weight [8]. Input into a model is a vector [90]. Thus inputting both a normalized bind count data vector and additional input data vector into a model would result in a combined vector used for training. The output of the second model is the fetal aneuploidy classification [187] [291] (Figure 16).
Claim 32:
Enrich teaches the training subjects are pregnant women with known fetal state of chromosomal euploid, aneuploidy, microdeletion and microduplication [8] [23] [179] [205-206] [294].
Claim 33:
Enrich teaches the training dataset contains more than 10 pregnant samples wherein half of the dataset was trisomy positive and the other half was trisomy negative [395].
Claim 34:
Claim 34 is being interpreted as an aesthetic design choice. MPEP 2144.04.I recites “[t]he court found that matters relating to ornamentation only which have no mechanical function cannot be relied upon to patentably distinguish the claimed invention from the prior art.”
Applicant has not disclosed that the arrangement of training data provides an advantage, is used for a particular purpose, or solves a stated problem when compared to the training data of Enrich, which contains the same training data of claim 34 albeit in a different form (Figure 16) [50] [66] [90] [179] [381]. Therefore, the training data of Enrich would perform the same as the training data arrangement of claim 34, and such a design modification fails to patentably distinguish over Enrich.
Conclusion
No claims are allowed.
Claims 16-17 and 27 are free from the prior art because the prior art does not fairly teach or suggest the following limitations: in claims 16-17 the sequence feature matrix containing sequence eigenvalue values in a sliding window, which is used in claim 1 to detect fetal chromosomal abnormality; and in claim 27 the sequence eigenvalue of the normalized feature vector, which is used in claim 1 to detect fetal chromosomal abnormality.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Noah A. Auger whose telephone number is (703)756-4518. The examiner can normally be reached M-F 7:30-4:30 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, Karlheinz Skowronek can be reached at (571) 272-9047. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/N.A.A./Examiner, Art Unit 1687
/KAITLYN L MINCHELLA/Primary Examiner, Art Unit 1685
Prosecution Insights