DETAILED ACTION
It is noted the Examiner of record has changed herein. The following rejections and/or objections are either reiterated or newly applied. They constitute the complete set presently being applied to the instant application.
This Office action is a non-final Office action that contains a new grounds of rejection under 35 U.S.C. 112(b), U.S.C. 101, and U.S.C. 103 after the filing of a notice of appeal.
In view of the Appeal Brief filed on 24 June 2026, PROSECUTION IS HEREBY REOPENED. New grounds of rejection are set forth below.
To avoid abandonment of the application, appellant must exercise one of the following two options:
(1) file a reply under 37 CFR 1.111 (if this Office action is non-final) or a reply under 37 CFR 1.113 (if this Office action is final); or,
(2) initiate a new appeal by filing a notice of appeal under 37 CFR 41.31 followed by an appeal brief under 37 CFR 41.37. The previously paid notice of appeal fee and appeal brief fee can be applied to the new appeal. If, however, the appeal fees set forth in 37 CFR 41.20 have been increased since they were previously paid, then appellant must pay the difference between the increased fees and the amount previously paid.
A Supervisory Patent Examiner (SPE) has approved of reopening prosecution by signing below:
/OLIVIA M. WISE/
Supervisory Patent Examiner, Art Unit 1685
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
Claims 1-15 and 18-23 are pending.
Claims 1-15 and 18-23 are rejected.
Claims 2-7, 10, 12, 18-20, and 23 are objected to.
Priority
Applicant’s claim for the benefit of a prior-filed application, PCT/EP2020/070472 filed 20 July 2020 and U.S. Provisional App. No. 62/877,238 filed 22 July 2019 under 35 U.S.C. 119(e) or under 35 U.S.C. 120, 121, 365(c), or 386(c) is acknowledged.
Accordingly, the effective filing date of the claimed invention is 22 July 2019.
Information Disclosure Statement
The information disclosure statement(s) (IDS) submitted on 15 May 2026 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the list of cited references was considered in full by the examiner.
Drawings
The drawings were accepted in the Office action mailed 09 April 2026.
Claim Objections
Claims 2-7, 10, 12, 18-20, and 23 are objected to because of the following informalities. This objection is newly recited.
Claims 2 and 18 recite “wherein the type of cancer is…”, which should be amended to recite “wherein the type of lymphoma is…” to increase clarity and use consistent language with claims 1 and 12.
Claims 3-6 recite “the group of genes consisting of EZH2…CD79B, MYD88.”, which is a grammatical error and should include an “and” after the penultimate member of the list to recite “…CD79B, and MYD88.”.
Claim 7 recites “(3)…a second cell of original class…”, which is a typographical error and should recite “a second cell of origin class”.
Claim 7 recites the abbreviation “GCB” before providing the full form of the term. Claim 7 should be amended to recite “germinal center B cell-like (GCB)”.
Claims 10 and 23 recite “(4) functional region of a protein…”, which is missing an article and should recite “(4) a functional region of a protein…”.
Claim 12 recites “A computer-implemented method for classifying a cell of origin for a type of lymhoma…”, which is a typographical error and should recite “a type of lymphoma”.
Claims 19-20 recite “MYD88L273P_SNV” which should be amended to recite “MYD88_L273P” to increase clarity that MYD88 is the gene and L273P is the variant.
Appropriate correction is required.
Claim Interpretation
Claims 1 and 11 recite “wherein the features comprise genes that were identified from cells of a plurality of blood or plasma samples from [human] subjects having a type of lymphoma”. The limitation regarding how the genes were previously identified from cells is interpreted as a product by process limitation defining the process in which the genes were previously identified. However, the claim does not require a step of identifying the genes from cells as claimed. "[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. See MPEP 2113.
Claim 7 recites “wherein the identifying a gene comprises…”, which further limits the process in which a gene was previously identified in claim 1 and thus is part of the product by process limitation.
Claim Rejections - 35 USC § 112(a)
The rejection of claims 1-15 and 18-23 under 35 U.S.C. 112(a) in the Office action mailed 09 April 2026 has been withdrawn after further consideration of the claims.
Claim Rejections - 35 USC § 112(b)
The previous rejection of claims 1-15 and 18-23 under 35 U.S.C. 112(b) in the Office action mailed 09 April 2026 has been withdrawn after further consideration of the claims. However, after further consideration of the claims, a new grounds of rejection is set forth below.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
Claims 1-15 and 18-23 are rejected under 35 U.S.C. 112(b) as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor regards as the invention. This rejection is newly recited.
Claim 1, and claims dependent therefrom, are indefinite for recitation of “wherein the features comprise genes that were identified…”. Claim 1 previously recites “a collection of features” and then “each decision tree comprising a plurality of nodes corresponding to a random subset of features from the collection of features”. As a result, it is not clear if “the features comprise genes” in the wherein clause is requiring one of the random subsets of features for a decision tree to comprise genes or the collection of features to comprise genes. If “the features” is intended to refer to one of the random subsets of features, it is further unclear which random subset of features is being referenced, given there is a random subset of features for each decision tree. For purpose of examination, the features is interpreted to refer to the collection of features, as suggested by dependent claim 3.
Claim 7 is indefinite for recitation of “about 2% of the subjects…” and “about 30% of all subjects…”. The term “about” is a relative term which renders the claim indefinite. The term “about” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. It is noted Applicant’s specification at para. [0083] states “the phrase “about” or “approximately” may be used when describing magnitude and/or position to indicate that the value and/or position described is within a reasonable expected range of values and/or positions. For example, a numeric value may have a value that is +/−0.1% of the stated value (or range of values), +/−1% of the stated value (or range of values), +/−2% of the stated value (or range of values), +/−5% of the stated value (or range of values), +/−10% of the stated value (or range of values), etc.”. However, the definition in Applicant’s specification does not serve to clarify the metes and bounds of the term “about”, and instead merely describes open-ended examples of what may be encompassed by the term “about” and does not clarify what would be considered “a reasonable expected range of values”. As a result, it is not clear if “about” strictly refers to +/−10% of the stated value, or if the term “about” may encompass +/- 12%, 15%, 20%, etc. Clarification is requested.
Claim 7 is indefinite for recitation of “(1)…the subjects from a first cell of origin class or a second cell of origin class…”. There is insufficient antecedent basis for this limitation in the claim because claim 1, from which claim 7 depends, recites “genes that were identified from…human subjects having a type of lymphoma”, but does not recite subjects from a first or second cell of origin class. For purpose of examination, the limitation is interpreted to mean “….2% of subjects from a first cell of origin class…”.
Claim 7 is indefinite for recitation of “(2)…30% of all subjects with variants in the gene are unclassified…”. It is unclear what set of subjects “all subjects” is referring to, and as a result one of ordinary skill in the art cannot ascertain when the condition of at most 30% of “all subjects” having a variant in the gene is met. For example, it is not clear if the claim (1) requires at most 30% of all subjects from the human subjects having the type of lymphoma have unclassified variants in the gene, (2) requires at most 30% of all subjects from subjects from a first cell of origin class or a second cell of origin class have variants in the gene, or (3) requires at most 30% of all subjects from some other set of subjects have variants in the gene (e.g. all subjects having a type of lymphoma). Clarification is requested via claim amendment.
Claim 7 is indefinite for recitation of “(3)…the second cell of origin class to the first cell of origin class…”. Claim 7 previously recites “(1)…subjects from a first cell of origin class or a second cell of origin[al] class GCB” and “(3) a ratio of a first cell of origin class to a second cell of origin class”, and thus recites two “first cell of origin” classes and two “second cell of origin” classes. As a result, it is not clear which first and second cell of origin class, “the second cell of origin class” and “the first cell of origin class” in step (3) is referring to, or if these are all intended to be the same first and second cell of origin classes. Furthermore, given (3) later refers to “the second cell of origin class” without mentioning “GCB”, it is not clear if “a second cell of origin class GCB” in (3) is requiring the second cell of origin class is a germinal center B cell-like class or if the second cell of origin class can be any cell origin class. Clarification is requested via claim amendment.
Claim 8 is indefinite for recitation of “…wherein the variant data was obtained from cells of a blood or plasma sample from the human subject”. Claim 8 previously recites “the human subjects”, “wherein the method further comprising receiving gene data comprising variant data from a human subject”, and claim 1, from which claim 8 depends, recites “cell gene data of a human subject”. It is not clear if “the human subject” from which the variant data was obtained in claim 8 is referring to one of the “human subjects”, the human subject recited in the training step of claim 1, or the human subject previously recited in the second limitation in claim 8. For purpose of examination, “the human subject” in claim 8 is interpreted to refer to the human subject recited in the second limitation of claim 8. If Applicant agrees with this interpretation, claim 8 can be amended to recite “comprising variant data from a given human subject, wherein the variant data was obtained…from the given human subject…based on the variant data from the given human subject”, or a similar amendment that distinguishes the human subject.
Claim 12, and claims dependent therefrom, are indefinite for recitation of “wherein the features comprise genes that were identified…”. Claim 12 previously recites “features of a collection of features” for each individual model and also “a subset of the collection of features” for each decision tree. As a result, it is not clear if “the features” is referring to the features of the collection of the features of an individual model, the collection of features, and/or the subset of the collection of features for a decision tree. Clarification is requested via claim amendment. For purpose of examination, the features is interpreted to refer to “the collection of features”.
Claim 15 is indefinite for recitation of “the selected number of iterations”. There is insufficient antecedent basis for this limitation in the claim because claims 12-13 do not previously recite a selected number of iterations. It is noted that claim 14 does recite “a selected number of iterations”. For purpose of examination, claim 15 is interpreted to mean that the iterative training of claim 12 is limited to 25 iterations.
Claims 19-20 are indefinite for recitation of “wherein the collection of features comprises at least…genes that are selected from the group of genes consisting of EZH2_SNV, GNA13_SNV, BCL2_Fusion…”. It is unclear if the collection features is only required to comprise genes (e.g. EZHG2, GNA13, etc. selected from the list of genes), as suggested by “the collection of features comprises…genes”, or if the collection of features is required to comprise particular variants in genes, as suggested by “EZH2_SNV”, etc. Clarification is requested via claim amendment. For purpose of examination, the features are interpreted to comprise the recited gene-variant pairs in light of Applicant’s specification at para. [0068].
Claim 21 is indefinite for recitation of “wherein the collection of features used to construct the plurality of decision trees…is increased by adding one or more additional features…based on determining an improvement in accuracy by a gain of at least 1 to the ensemble decision tree machine learning model”. Claim 21 previously recites “determining a prediction accuracy of the ensemble decision tree machine learning model”, such that only a single prediction accuracy is determined. As a result, it is unclear what the “improvement in accuracy by a gain of at least 1” is relative to, such that one of ordinary skill in the art could ascertain when accuracy is improved by a gain of at least 1. Furthermore, given the claims determine a single prediction accuracy, it is unclear if the wherein clause is merely expressing an intended result or use of the determined prediction accuracy, or if the claims intend to require determining multiple prediction accuracies and increasing the collection of features based on the accuracy improving. See MPEP 2111.04 I, stating “‘whereby clause in a method claim is not given weight when it simply expresses the intended result of a process step positively recited.’”. For purpose of examination, claim 21 is interpreted to require determining a prediction accuracy as claimed, and the wherein clause is interpreted as an intended use of the determined prediction accuracy.
Claim 22 is indefinite for recitation of “…wherein the variant data was obtained from cells of a blood or plasma sample from the human subject”. Claim 22 previously recites “human subjects”, “wherein the method further comprising receiving gene data comprising variant data from a human subject”, and claim 12, from which claim 22 depends, recites “cell gene data of a human subject”. It is not clear if “the human subject” from which the variant data was obtained in claim 22 is referring to one of the “human subjects”, the human subject recited in claim 12, or the human subject previously recited in the second limitation in claim 22. For purpose of examination, “the human subject” in claim 22 is interpreted to refer to the human subject recited in the second limitation of claim 22. If Applicant agrees with this interpretation, claim 22 can be amended to recite “comprising variant data from a given human subject, wherein the variant data was obtained…from the given human subject…based on the variant data from the given human subject”, or a similar amendment that distinguishes the human subject.
Response to Arguments
Applicant's arguments filed 24 June 2026 regarding 35 U.S.C. 112(b) have been fully considered but they are not persuasive because they do not pertain to the new grounds of rejection set forth above.
Claim Rejections - 35 USC § 101
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 1-15 and 18-23 are rejected under 35 U.S.C. 101 because the claimed invention is directed to one or more judicial exceptions without significantly more. This rejection is newly recited.
The Supreme Court has established a two-step framework for this analysis, wherein a claim does not satisfy § 101 if (1) it is “directed to” a patent-ineligible concept, i.e., a law of nature, natural phenomenon, or abstract idea, and (2), if so, the particular elements of the claim, considered “both individually and as an ordered combination,” do not add enough to “transform the nature of the claim into a patent-eligible application.” Elec. Power Grp., LLC v. Alstom S.A., 830 F.3d 1350, 1353 (Fed. Cir. 2016) (quoting Alice, 134 S. Ct. at 2355). Applicant is also directed to MPEP 2106.
Step 1: The instantly claimed invention (claims 1 and 12 being representative) is directed to a method of classifying a cell of origin of a type of lymphoma. Therefore, the instantly claimed invention falls into one of the four statutory categories. [Step 1: YES]
Step 2A: First it is determined in Prong One whether a claim recites a judicial exception, and if so, then it is determined in in Prong Two if the recited judicial exception is integrated into a practical application of that exception.
Step 2A, Prong 1: Under the MPEP § 2106.04, the Step 2A (Prong 1) analysis requires determining whether a claim recites an abstract idea, law of nature, or natural phenomenon.
Claim 1 recites the following steps which fall under the mathematical concepts and/or mental processes groupings of abstract ideas:
constructing a plurality of decision trees based on a collection of features, each decision tree comprising a plurality of nodes corresponding to a random subset of features from the collection of features, wherein the features comprise genes that were identified from cells of a plurality of blood or plasma samples from human subjects having a
training a machine learning model to classify the type of lymphoma from cell gene data of a human subject, the machine learning model trained by iterating through a set of training data using the plurality of decision trees based on the collection of features.
Claim 12 recites the following steps which fall under the mathematical concepts and/or mental processes groupings of abstract ideas:
constructing an ensemble decision tree machine learning model configured to classify a type of lymphoma from cell gene data of a human subject, the ensemble decision tree comprising
iteratively training each individual decision tree machine learning model through a set of training data to generate a plurality of successive decision trees that are added to each individual decision tree machine learning model respectively, wherein each successive decision tree is configured to correct for an error in a previous decision tree, wherein each decision tree comprises a subset of the collection of features, wherein the features comprise genes that were identified
The identified claim limitations falls into one of the groups of abstract ideas of mathematical concepts and/or mental processes for the following reasons. In this case, constructing a plurality of decision trees (i.e. an ensemble decision tree in claim 12) based on a collection of features with each decision tree comprising nodes corresponding to a subset of gene features can be practically performed in the mind by organizing gene information into genes associated with a type of lymphoma versus not associated with a type of lymphoma, and drawing via pen and paper nodes and edges in the form of a hierarchical tree corresponding to the genes associated or not associated with a type of lymphoma to reach leaf nodes corresponding to a classification. For example, Kingsford et al. (What are decision trees, 2008, Nat Biotechnol., 26(9), pg. 1-4; newly cited) describes the process in which decision trees are constructed by posing series of questions in the form of a hierarchy, including an example of a constructed tree (pg. 1, para. 3-5; Figure 1). That is, other than reciting the step is carried out by a computer, nothing in the claims preclude the step from being practically performed in the mind. See MPEP 2106.04(a)(2) III.
The above limitations of constructing decision trees and iteratively training each individual decision tree machine learning model recite a mathematical concept. First, the construction of decision trees using gene features for cell of origin classification of lymphoma recites a mathematical relationship between the presence or absence of particular features and the probability of a given classification, analogous to claims of organizing information and manipulating information through mathematical correlations in Digitech Image Techs., LLC v. Electronics for Imaging, Inc., 758 F.3d 1344, 1350, 111 USPQ2d 1717, 1721 (Fed. Cir. 2014). See MPEP 2106.04(a)(2) I. A. Iteratively training a machine learning model to classify the type of lymphoma by iterating through a set of training data using the plurality of decision trees (claim 1) and similarly training each individual decision tree model to generate successive decision trees that are added to each decision tree recites a mathematical calculation. MPEP 2106.04(a)(2) I. C states a mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation. The embodiment of training in claim 12 involving generating successive decision trees is merely using words to describe the mathematical concept of applying a gradient boosted decision tree algorithm for training using a gradient descent algorithm to minimize the error, as described in Applicant’s specification at para. [0057] and [0063]), and therefore recites a mathematical concept. The broader embodiment of training a machine learning model using the plurality of decision trees similarly encompasses the same mathematical calculation regarding applying a gradient descent algorithm to train the machine learning model in light of Applicant’s specification, and thus recites a mathematical concept.
Dependent claims 2-11, 13-15, and 18-22 further recite an abstract idea and/or are part of the abstract idea of claims 1 and 12 above. Dependent claims 2 and 18 further limit the abstract idea of classifying the type of lymphoma to classify a type of diffuse large B cell lymphoma, and thus are part of the abstract idea of claims 1 and 12. Dependent claims 2-6, 9-11, and 19-20 further limit the collection of features to comprise certain genes, variant location data, and variant allele fraction data, and thus are part of the abstract idea of constructing decision trees using the collection of features discussed above in claims 1 and 12. Dependent claim 7 defines the process in which the genes of the collection of features were previously identified, and thus is part of the abstract idea of constructing decision trees using the collection of features comprising genes discussed above in claim 1. Dependent claims 8 and 22 further limit the abstract idea of constructing decision trees to use gene variant data, and further recite the mental process and mathematical concept determining a type of lymphoma based on the variant data from the human subject using the machine learning model of the decision trees, given the human mind is equipped to analyze and follow a decision tree based on input data to arrive a leaf node/classification. Dependent claim 13 further limits the mental process of constructing the decision trees to have each initial tree comprising a random subset of features. Dependent claims 14-15 further limit the mathematical concept of training to be limited to a certain number of iterations. Dependent claim 21 further recites the mathematical concept of determining a prediction accuracy of the ensemble decision tree model, and constructing decision trees using a collection of features determined based on an improvement in accuracy.
Claims 1-15 and 18-23 further recite the law of nature of a natural correlation of the presence of particular gene features in blood or plasma samples of a subject and a cell of origin of a type of lymphoma in the subject, analogous to a correlation between the presence of myeloperoxidase in a bodily sample (such as blood or plasma) and cardiovascular disease risk in Cleveland Clinic Foundation v. True Health Diagnostics, LLC, 859 F.3d 1352, 1361, 123 USPQ2d 1081, 1087 (Fed. Cir. 2017). See MPEP 2106.04(b) I.
Therefore, claims 1-15 and 18-23 recite an abstract idea and law of nature. [Step 2A, Prong 1: YES]
Step 2A: Prong 2: Under the MPEP § 2106.04, the Step 2A, Prong 2 analysis requires identifying whether there are any additional elements recited in the claim beyond the judicial exception(s), and evaluating those additional elements to determine whether they integrate the exception into a practical application of the exception. This judicial exception is not integrated into a practical application for the following reasons.
Claims 2-7, 9-11, and 13-21 do not recite any elements in addition to the judicial exception and thus are part of the judicial exception.
The additional elements of claims 1 and 12 include:
a computer.
The additional elements of claims 8 and 22-23 include:
receiving gene data comprising variant data from a human subject, wherein the variant data was obtained from cells of a blood or plasma sample from the human subject (i.e. receiving data) (claims 8 and 22); and
wherein the variant data comprises at least one of: (1) a position of a variant in a gene; (2) information about a domain region of a protein which is modified in a variant; (3) information about a structural motif of a protein which is modified in a variant; or (4) functional region of a protein which is modified in a variant (i.e. receiving data) (claim 23).
The additional elements of claims 1, 8, 12, and 22-23 encompass a computer used to carry out the abstract idea and receiving particular variant data by the computer. The courts have found the use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application. See MPEP 2106.05(f).
Furthermore, the additional element of receiving the variant data in claims 22-23 only serves to collect information for use by the abstract idea of determining the type of lymphoma, which amounts to insignificant extra-solution activity that is not sufficient to integrate the recited judicial exception into a practical application. See MPEP 2106.05(g).
Therefore, the additionally recited elements merely invoke computers as a tool and/or amount to insignificant extra-solution activity and, as such, the claims as a whole do no integrate the abstract idea into practical application. Thus, claims 1-15 and 18-23 are directed to an abstract idea and law of nature. [Step 2A, Prong 2: NO]
Step 2B: In the second step it is determined whether the claimed subject matter includes additional elements that amount to significantly more than the judicial exception. See MPEP § 2106.05.
The claims do not include any additional steps appended to the judicial exception that are sufficient to amount to significantly more than the judicial exception for the following reasons.
Claims 2-7, 9-11, and 13-21 do not recite any elements in addition to the judicial exception and thus are part of the judicial exception.
The additional elements of claims 1 and 12 include:
a computer.
The additional elements of claims 8 and 22-23 include:
receiving gene data comprising variant data from a human subject, wherein the variant data was obtained from cells of a blood or plasma sample from the human subject (i.e. receiving data) (claims 8 and 22);
wherein the variant data comprises at least one of: (1) a position of a variant in a gene; (2) information about a domain region of a protein which is modified in a variant; (3) information about a structural motif of a protein which is modified in a variant; or (4) functional region of a protein which is modified in a variant (i.e. receiving data) (claim 23).
The additional elements of claims 1, 8, 12, and 22-23 encompass a computer used to carry out the abstract idea and receiving particular variant data by the computer. The courts have found the use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not provide significantly more. See Affinity Labs v. DirecTV, 838 F.3d 1253, 1262, 120 USPQ2d 1201, 1207 (Fed. Cir. 2016) (cellular telephone); TLI Communications LLC v. AV Auto, LLC, 823 F.3d 607, 613, 118 USPQ2d 1744, 1748 (Fed. Cir. 2016) (computer server and telephone unit).
Taken alone, the additional elements do not amount to significantly more than the above-identified judicial exception(s). Even when viewed as a combination, the additional elements fail to transform the exception into a patent-eligible application of that exception. Thus, the claims as a whole do not amount to significantly more than the exception itself. [Step 2B: NO]
Therefore, the instantly rejected claims are not drawn to eligible subject matter as they are directed to an abstract idea and law of nature without significantly more. For additional guidance, applicant is directed generally the MPEP § 2106.
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.
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 1-5, 7-15, 18-19, and 21-23 are rejected under 35 U.S.C. 103 as being unpatentable over Scherer (2016) in view of Kimmerling (2019) as evidenced by Chen (2016). This rejection is newly recited.
Cited references:
Scherer et al., Distinct biological subtypes and patterns of genome evolution in lymphoma revealed by circulating tumor DNA, 2016, Science Translational Medicine, 8, pg. 1-11 and supplementary (newly cited); and
Kimmerling et al., US 2020/0227168 A1, effectively filed 10 Jan. 2019 based on priority to U.S. Provisional App. No. 62/790,804 (newly cited); and
Chen et al., XGBoost: A scalable Tree Boosting System, 2016, KDD’16: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pg. 785-794 (newly cited).
Regarding claims 1 and 12, Scherer discloses a method for determining a cell of origin (COO) of diffuse large B cell lymphoma (DLBCL) (i.e. a type of lymphoma) (Abstract), wherein the method comprises the following steps.
Scherer discloses a collection of genomic features incorporated into a classifier for predicting a COO of DLBCL (Fig. S12; pg. 5, col. 1, para. 3 to col. 2, para. 2), wherein the collection of features comprise genes including mutations (Fig. S12). Scherer further discloses mutations in genes are detected from circulating tumor DNA obtained from plasma in humans with DLBCL (Abstract; Fig. 1).
Scherer discloses training a classifier (i.e. a machine learning model) for classifying COO of DLBCL (i.e. to classify the type of lymphoma) using genetic features (i.e. cell gene data) determined from patients in a training cohort (pg. 5, col. 1, para. 3; Fig. S12).
Further regarding claims 1 and 12¸ Scherer does not disclose the following limitations:
Regarding claims 1 and 12, Scherer discloses using a probabilistic classifier (i.e. a machine learning model) for COO classification, but does not disclose using the specific machine learning architectures of claims 1 and 12, as described below.
Regarding claim 1, Scherer does not disclose construing a plurality of decision trees based on the collection of features, wherein each decision tree comprises a plurality of nodes corresponding to a random subset of features from the collection of features. Scherer further does not disclose the machine learning model Is trained by iterating through a set of training data using the plurality of decision trees.
Regarding claim 12, Scherer similarly does not disclose constructing an ensemble decision tree machine learning model, the ensemble decision tree comprising a plurality of individual decision tree machine learning models, each individual machine learning model comprising an initial decision tree having a plurality of nodes corresponding to features of the collection of features. Scherer further does not disclose iteratively training each individual decision tree machine learning model to generate successive decision trees that are added to each individual decision tree machine learning model, wherein each successive decision tree is configured to correct for an error in a previous decision tree, and wherein each decision tree comprises a subset of the collection of features.
However, these limitations were known in the art, before the effective filing date of the claimed invention, as shown by Kimmerling as evidenced by Chen.
Regarding claims 1 and 12, Kimmerling discloses a method for using machine learning to predict a cancer status from clinical data comprising genetic data of cells (Abstract; [0004]-[0005], e.g. biomarker measurements include genetic data). Kimmerling discloses any machine learning model may be used to analyze the data for classification including a random forest or extreme gradient boosting methods (XGBoost) ([0045]).
Further regarding claim 1, Kimmerling discloses that random forests use decision tree learning and include multiple decision trees, wherein decision trees each include a series of nodes corresponding to input variables (i.e. a plurality of decision trees, each decision tree comprising a plurality of nodes corresponding to features) ([0047]). Kimmerling further discloses that in random forests, a random subset of features is selected at each split (i.e. node) in the learning process (i.e. the nodes correspond to a random subset of features) ([0047]), and furthermore, that during training, random forests use bootstrap aggregation to average predictions by multiple trees that are given different sets of training data (i.e. training iterates through a set of training data using the plurality of decision trees) ([0047]). Kimmerling discloses random forest classifiers are a stacking type of machine learning algorithm which improves predictive force ([0046]).
Further regarding claim 12, Kimmerling further discloses the machine learning algorithm may include an extreme gradient boost method (XGBoost) ([0045]), and discloses boosting is focused on turning weak learners into strong learner, which consists of iteratively learning weak classifiers with respect to a distribution and adding them to a final strong classifier (e.g. successive decision trees are configured to correct for an error in a previous decision tree) ([0049]). While Kimmerling does not explicitly describe the XGBoost algorithm involves constructing an ensemble decision tree model comprising initial decision trees with successive decision trees generated during training, Kimmerling further incorporates Chen by reference, which further describes the XGBoost algorithm ([0049]). Chen discloses XGBoost uses a tree ensemble model comprising multiple decision trees, each decision tree comprising a plurality of nodes corresponding to a different subset of features (pg. 786, col. 1, para. 4 to col. 2, para. 1; Figure 1, e.g. tree ensemble with multiple decision trees, wherein each node includes a feature). Chen further discloses the model is trained in an additive manner and involves learning each tree structure in the ensemble, such that each tree is learned iteratively to construct a stronger tree classifier (pg. 786, col. 2, para. 2, to pg. 787, col. 1, para. 3; pg. 787, col. 2, para. 2). Kimmerling further discloses that boosting algorithms are machine learning ensemble meta-algorithms for reducing bias and variance ([0049]).
It would have been prima facie obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to have modified the machine learning model of Scherer to have used a machine learning model of a random forest and/or an XGBoost algorithm, as described by Kimmerling above, thus arriving at the inventions of claims 1 and 12. One of ordinary skill in the art would have been motivated to combine the methods of Scherer and XGBoost in order to utilize a machine learning model that results in a strong predictive force and reduces bias and variance in predictions, as shown by Kimmerling ([0046]; [0049]). This modification would have had a reasonable expectation of success given both Scherer and Kimmerling disclose using machine learning to make cancer predictions based on genetic data, and Kimmerling discloses any of a variety of machine learning algorithms may be used to analyze the data ([0045]). Accordingly, one of ordinary skill in the art would have a reasonable expectation of success in using one of the already known machine learning algorithms described in Kimmerling.
Regarding the dependent claims:
Regarding claim 2, Scherer discloses the type of lymphoma is DLBCL (Abstract).
Regarding claims 3-5, Scherer discloses the collection of features include BCL2 translocation, BCL2 mutation, CREBBP, SGK1, GNA13, TNFRSF14, BCL6 mutation, BCL6 translocation, MYC translocation, MYC/BCL2 DH translocation, IRF8, CD83, EZH2, STAT3, STAT6, CD79B, PIM1, MLL2 (i.e. KMT2D), MYD88, EP300, CD58, CARD11, KLHL6, and CCND3 (FIG. S12).
Regarding claim 7, Scherer discloses the identified genes each (1) have about 2% of subjects from a GCB-like cell origin or a ABC-like cell origin having a variant in the gene (FIG. S12, e.g. each variant frequency of a gene in GCB or ABC column is >=0.02).
Regarding claim 8, Scherer discloses mutations in genes are detected from circulating tumor DNA obtained from plasma in humans with DLBCL (Abstract; Fig. 1).
Scherer discloses testing the trained COO classifier (i.e. using the machine learning model) with gene mutation data from pretreatment plasma ctDNA of a human subject (i.e. receiving gene variant data from a plasma sample) to determine whether the subject has GCB or non-GCB DLBCL (i.e. to determine the type of lymphoma) (pg. 5, col.1, para. 3 to col. 2, para. 2, e.g. classifier built on mutations enriched in DLBCL GCB or non-GCB subtypes; FIG. 4; FIG. S12, e.g. gene variant data used for classification).
Regarding claim 9, Scherer discloses the features include which gene a variant is located (Fig. S1 and Fig. S12).
Regarding claim 10, Scherer discloses the features include whether the variant is a missense, splice, or stop-gain mutation, an indel frameshift mutation, or a fusion or translocation (Fig. S1; FIG. S12, e.g. each feature is a frequency of a translocation or mutation), wherein the type of variants represents information about a structural motif modified by the variant (e.g. missense, splice, stop-gain, indel frameshift represent information about changes in amino acid sequence/protein structure).
Regarding claim 11, Scherer further discloses the collection of features comprises variant allele fraction data (Fig. S12, e.g. the frequency of each feature incorporated into COO; Table S5, e.g. mutation frequencies used for subtype determination).
Regarding claim 13, Scherer in view of Kimmerling, as evidenced by Chen, as applied to claims 1 and 12 above disclose each tree has a random subset of features.
Regarding claims 14-15, Scherer does not disclose the training each individual decision tree is limited to a selected number of iterations through the training data, wherein the selected number of iterations is 25.
However, Chen, incorporated by reference into Kimmerling, further discloses XGBoost uses a set number of training iterations, and provides in example in which 20 or 30 training iterations are used (Figure 3). Chen discloses the accuracy of the model improves with the number of iterations (Figure 3, e.g. increased AUC with increased iterations).
It would have been prima facie obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to have limited a number of training iterations to 25 through routine experimentation of the number of iterations within the prior art conditions of increasing the iterations to increase model accuracy and decreasing the number of iterations to improve model runtime, as shown by Chen (Figure 3; pg. 787, col. 2, para. 3 to pg. 788, col. 1, para. 1). See MPEP 2144.05 II. A.
Regarding claim 18, Scherer discloses the type of lymphoma is DLBCL (Abstract).
Regarding claim 19, Scherer discloses the features include EZH2 mutation, GNA13 mutation, BCL2 translocation, BCL2 mutation, CD79B mutation, PIM1 mutation, SGK1 mutation, STAT6 mutation, TNFRSF14 mutation, CD58 mutation (Suppl., pg. 4, para. 2, e.g. mutation includes SNV or indel; Fig. S12 and Table S5; pg. 2, col. 1, 2 to col. 1, para. 2, e.g. variants include SNVs, indels, and breakpoints involving genes that participate in fusions).
Regarding claim 21, Scherer discloses benchmarking the classifier performance using the training cohort by comparing the classification results to a immunohistochemical classification of COO and determining a concordance rate (i.e. an accuracy) of 80% (pg. 5, col. 1, para. 3 to col. 2, para. 2; Fig. 4).
Regarding claim 22, Scherer discloses mutations in genes are detected from circulating tumor DNA obtained from plasma in humans with DLBCL (Abstract; Fig. 1).
Scherer discloses testing the trained COO classifier (i.e. using the machine learning model) with gene mutation data from pretreatment plasma ctDNA of a human subject (i.e. receiving gene variant data from a plasma sample) to determine whether the subject has GCB or non-GCB DLBCL (i.e. to determine the type of lymphoma) (pg. 5, col.1, para. 3 to col. 2, para. 2, e.g. classifier built on mutations enriched in DLBCL GCB or non-GCB subtypes; FIG. 4; FIG. S12, e.g. gene variant data used for classification).
Regarding claim 23, Scherer discloses the variant data comprises information about a structural motif of a protein which is modified in a variant (Fig. S1 and Fig. S12, e.g. missense, splice mutation of feature genes; Fig. 1, e.g. mutations detected in tumor or plasma).
Therefore, the invention is prima facie obvious.
Claim 6 is rejected under 35 U.S.C. 103 as being unpatentable over Scherer in view of Kimmerling, as evidenced by Chen, as applied to claim 1 above, and further in view of Reddy (2017). This rejection is newly recited.
Cited reference: Reddy et al., 2017, Genetic and Functional Drivers of Diffuse Large B Cell Lymphoma, Cell, 171, pg. 481-494 (newly cited).
Regarding claim 6, Scherer in view of Kimmerling, as evidenced by Chen, disclose the methods of claim 1 above
Further regarding claim 6, Scherer discloses the collection of features include BCL2 translocation, BCL2 mutation, CREBBP, SGK1, GNA13, TNFRSF14, BCL6 mutation, BCL6 translocation, MYC translocation, MYC/BCL2 DH translocation, IRF8, CD83, EZH2, STAT3, STAT6, CD79B, PIM1, MLL2 (i.e. KMT2D), MYD88, EP300, CD58, CARD11, KLHL6, and CCND3 (i.e. 21 genes) (FIG. S12).
Scherer in view of Kimmerling, as evidenced by Chen, as applied to claim 1 above, does not disclose the following limitations:
Regarding claim 6, Scherer in view of Kimmerling, as evidenced by Chen, as applied to claims 1 and 12 above, does not disclose at least 30 of the genes from the collection of features is selected from the group of genes consisting of Scherer discloses the collection of features include BCL2 translocation, BCL2 mutation, CREBBP, SGK1, GNA13, TNFRSF14, BCL6 mutation, BCL6 translocation, MYC translocation, MYC/BCL2 DH translocation, IRF8, CD83, EZH2, STAT3, STAT6, CD79B, PIM1, MLL2 (i.e. KMT2D), MYD88, EP300, CD58, CARD11, KLHL6, and CCND3 (FIG. S12).
However, regarding claim 6 Reddy analyzes whole-exome sequencing data of DLBCL patients to define the landscape of genetic drivers of the disease (Abstract), and discloses the top 60 most frequently mutated genes include FOX01, IRF8, IRF4, TBLXR1, TET2, PAX5, ZNF608, DUSP2, EP300, ETS1, and CKN2A (pg. 482, col. 2, para. 1; Figure 1).
It would have been prima facie obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention to have modified the collection of features of Scherer in view of Kimmerling, as applied to claim 1 above, to have further included the additional 11 frequency mutated genes in DLBCL patients of Reddy, discussed above, thus arriving at the invention of claim 6. One of ordinary skill in the art would have been motivated to combine the methods of Scherer in view of Kimmerling with Reddy, in order to incorporate additional driver mutations of DLBCL in the model of Scherer, given Scherer uses genomic features that are frequently mutated in DLBCL (Fig. S12) and Reddy discloses the driver genes include subnetworks associated with cell of origin in DLBCL (pg. 482, col. 2, para. 6). This modification would have had a reasonable expectation of success given Scherer uses gene features based on the presence of mutations in genes, and thus the driver mutations that are frequency mutated in DLBCL of Reddy are applicable to the method of Scherer.
Therefore, the invention is prima facie obvious.
Conclusion
No claims are allowed.
Claim 20 is free of the prior art. Claim 20 requires the collection of features comprise at least 20 genes that are selected from the group of genes (interpreted to mean gene-variant pairs in the 112(b) rejection above) consisting of EZH2_SNV, GNA13_SNV, BCL2_Fusion, CD79B_SNV, PIM1_SNV, IGLL5_Indel, PIM1_Indel, SGK1_SNV, MYD88_L273P_SNV, STAT6_SNV, TNFRSF14_SNV, P2RY8_SNV, CIITA_Indel, EGR1_SNV, ATG5_Indel, IRF4_SNV, S1PR2_SNV, SOCS1_SNV, CD58_Indel, and CNTNAP2_SNV, such that all of the recited gene-variant pairs are required. The closest prior art of record Scherer and Reddy each disclose the features include genes from the group of genes, as applied to claims 3-6 and 19 in the above rejection. However, Scherer and Reddy do not disclose or make obvious using features comprising all 20 of the recited gene-variant pairs, including the specific mutation in MYD88_L273P_SNV in combination with various other gene-variant pairs including at least P2RY8_SNV, CIITA_Indel, EGR1_SNV, and ATG5_Indel.
Inquiries
Any inquiry concerning this communication or earlier communications from the examiner should be directed to KAITLYN L MINCHELLA whose telephone number is (571)272-6485. The examiner can normally be reached 7:00 - 4:00 M-Th.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Olivia Wise can be reached 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.
/KAITLYN L MINCHELLA/Primary Examiner, Art Unit 1685