DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Herein, “the previous Office action” refers to the Final Rejection filed 4/17/2025.
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 10/17/2025 has been entered.
Priority
As detailed on the Filing Receipt filed 4/5/2023, the instant application claims priority to as early as 3/31/2021. At this point in prosecution, all claims are accorded the earliest claimed priority date.
Information Disclosure Statement
The Information Disclosure Statement filed on 10/17/2025 is in compliance with the provisions of 37 CFR 1.97 and has been considered in full. A signed copy of the IDS is included with this Office Action.
Oath/Declaration
A declaration under 37 CFR § 1.132 (hereafter, “the Wilcox Declaration”) was received on 10/17/2025.
The Wilcox Declaration includes statements and data indicating that trained model achieves cited performance metrics based on a minimal number of features, and particularly notes that the model achieves improved mean ROC and specificity when trained on both protein and lipid features relative to when trained on just protein features (pp. 1-2, point 4; pp. 3-6, Exhibits A-D). The claims require that the recited classifier utilizes both protein and lipid features, comprising fewer than 100 features, and achieves an AUC of at least 0.84. The presented statements and data affirm that Applicant has possession of a model that performs as claimed.
Evidence of possession does not rebut obviousness of claimed subject matter over prior art, when the prior art predates the effective filing date of the claimed subject matter. The Wilcox Declaration fails to outweigh the evidence of obviousness, and is insufficient to overcome the rejections under 35 USC § 103 as set forth in the previous Office action.
Claim Status
Claims 2, 5-6, 12, 18-19, 25-30 and 32 are canceled.
Claims 1, 3-4, 7-11, 13-17, 20-24 and 31 are pending.
Claims 13-15 stand withdrawn pursuant to 37 CFR 1.142(b) as being directed to a nonelected invention, there being no currently allowable generic or linking claim. Election without traverse was made in the reply filed 5/26/2023.
Claims 1, 3-4, 7-11, 16-17, 20-24 and 31 are under examination.
Withdrawn Rejections
The rejection of claims 10-11 and 32 under 35 USC §112(b), as being indefinite, is hereby withdrawn in view of Applicant’s amendment of claims 10-11 to resolve indefinite language and cancelation of claim 32.
The rejection of claims 1, 4, 7-11, 16-17, 20-24 and 31-32 under 35 USC § 103, as being unpatentable over Drake, in view of Kobold, is hereby withdrawn in view of Applicant’s amendment of claim 1, persuasive argument the cited references do not teach or suggest a classifier that uses fewer than 100 features from multi-omic data (remarks filed 10/17/2025 at pg. 7, paras. 1-2) and cancelation of claim 32.
The rejection of claim 3 under 35 USC § 103, as being unpatentable over Drake, in view of Kobold and Wang, is hereby withdrawn in view of Applicant’s amendment of claim 1, and resulting necessity of applying additional art to claim 1 and dependents thereof.
Response to Arguments - Claim Rejections Under 35 USC § 103
In the remarks filed 10/17/2025, Applicant traverses the rejections under 35 USC § 103 and presents supporting arguments.
Applicant alleges that the classifier having a performance characteristic including an AUC value of at least 0.84, as claimed, is unexpected as evidenced by the Wilcox Declaration (pg. 7, para. 3). The Wilcox Declaration notes that the claimed AUC was achieved by training the model on both protein and lipid features, an improved performance relative to when trained on just protein features, but does not demonstrate that such performance is unexpected (i.e., would not be expected by one of ordinary skill in the art).
Drake discloses assaying a plurality of classes of molecules, identifying a set of features corresponding to each of the plurality of assayed classes of molecules, and training a classifier based on the features (para. 0010 and 0022), wherein assayed classes can include proteins and lipids (paras. 0013-14). Drake also presents restriction to analysis of a single class of molecules, e.g., only circulating proteins, as a relative deficiency of prior methods (para. 0003). Drake thus presents combined utilization of protein features and lipid features by a trained classifier as an embodiment of the disclosed multi-omic method, and teaches that multi-omic analysis is functionally superior to utilization of protein features alone.
Drake further discloses selection of model features and validation of performance based on AUC values (paras. 0251 and 0485), and exemplifies performance thresholding according to a minimum AUC value of 0.84 (para. 0231). Any classifier passing this performance threshold would necessarily exhibit an average or median AUC of at least 0.84.
Drake does not expressly disclose a set of protein and lipid features that allow a classifier trained thereon to achieve the exemplified AUC threshold of 0.84. However, the presence of express guidance within the four corners of one or more given prior art references is not necessary to conclude that one of ordinary skill in the art would have possessed the knowledge or ability to achieve particular embodiments disclosed therein. As the courts have explained, one of ordinary skill “would, of necessity have the capability of understanding the scientific and engineering principles applicable to the pertinent art” (Ex parte Hiyamizu, 10 USPQ2d 1393, 1394 (Bd. Pat. App. & Inter. 1988)).
The previously-cited book chapter by Wang, Y (Chapter 2 in: Advances in Experimental Medicine and Biology, vol 1316, Springer, pp. 25-39; published 3/20/2021) reviews application of lipodomics to cancer diagnosis, and discusses numerous lipid markers that are diagnostic for various forms of cancer. The predictive utility of lipids for classification of cancer is viewed as conventional knowledge within the field of the invention, as is evidenced by Wang. One of ordinary skill in the art would reasonably expect that a classifier trained on both protein features (i.e., a first class of marker data) and lipid features (i.e., an additional class of marker data known in the art to have predictive utility for cancer) would exhibit improved performance in classifying cancer relative to a classifier trained on only the first class. This improvement is not considered an unexpected result, and the argument of unexpected results is found unpersuasive.
However, Applicant’s amendment of claim 1 introduced new limitations (e.g., “using fewer than 100 features”) to claim 1 and dependents thereof that are not taught by the references applied in the previous Office action. Hence, the previous rejections under § 103 have been withdrawn.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 USC §§ 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 USC § 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 USC § 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 USC § 102(b)(2)(C) for any potential 35 USC § 102(a)(2) prior art against the later invention.
Claims 1, 3-4, 7-11, 16-17, 20-24 and 31 are rejected under 35 USC § 103 as being unpatentable over Drake et al (WO 2019/200410; published 10/17/2019; on IDS filed 2/7/2023; previously cited), in view of Kobold et al (US 2009/0090855; published 4/9/2009; on IDS filed 2/7/2023; previously cited), El-Khoury et al (Cancers 12(6): 1629, 15 pages; published 6/19/2020) and Chen et al (Oncotarget 7(24): 36622-36631; published 5/2/2016). The new grounds of rejection presented herein were necessitated by Applicant’s amendment of the claims (filed 10/17/2025).
Claim 1 discloses a multi-omic method comprising steps of: dividing a biofluid sample, from a subject suspected of having cancer, into at least first and second portions; forming biomolecule coronas, comprising adsorbed proteins, by contacting the first portion with nanoparticles; assaying the adsorbed proteins or fragments thereof, using mass spectrometry or an immunoassay, to generate proteomic measurements; assaying the second portion, using mass spectrometry or an immunoassay, to detect lipids and generate lipid measurements; obtaining multi-omic data comprising the proteomic and lipid measurements; and processing the multi-omic data, using a classifier, to assign a likelihood of cancer.
The claim further requires that: the classifier assigns the classification using fewer than 100 features from the multi-omic data; performance of the classifier is characterized by a receiver operating characteristic (ROC) curve having an average or median area under the curve (AUC) of at least 0.84; and the classifier is trained using samples from subjects with cancer and control samples from subjects without cancer.
With respect to claim 1, Drake discloses dividing a biofluid sample, from a subject, into aliquots for analysis of each aliquot (paras. 0080, 0121 and 0135-36); assaying proteins, using mass spectrometry or an immunoassay, to generate proteomic measurements (paras. 0011-0013, 0018, 0107 and 0196); assaying lipids, using mass spectrometry or an immunoassay, to generate lipid measurements (paras. 0014, 0018 and 0115); preparing a feature vector, i.e., multi-omic data, comprising feature values obtained using each set of measurements (para. 0010); and processing the data using classifiers (paras. 0051-56, 0063-64, 0069-70, 0122, 0185 and 0249) to generate a classification comprising likelihood of cancer (paras. 0029, 0091, 0295 and 0389).
Drake further discloses evaluation of classifier performance via cross-validation to calculate mean, i.e., average, AUC values (para. 0485) and exemplifies evaluation of various models exhibiting mean AUC values of at least 0.84 (paras. 0485-86, 0509 and Table 11). Drake also discloses application of thresholds for model performance to select features (para. 0251), and exemplifies thresholding based on a minimum AUC value of 0.84 (para. 0231). Any classifier passing this performance threshold would exhibit an average or median AUC of at least 0.84.
Additionally, Drake discloses that a classifier is trained using a first subset of training samples identified as having a specified property and a second subset identified as not having the specified property (para. 0010), wherein the specified property may be cancer (para. 0021).
Drake does not disclose adsorbing proteins to nanoparticles; or embodiments wherein the classifier assigns the classification using fewer than 100 features from the multi-omic data.
Kobold teaches analysis of biological analytes comprising contacting a biological sample with magnetic nanoparticles; incubating the sample to adsorb analytes to the particles; fractionating the sample by eluting the liquid and/or washing the particles; and detecting the compounds in sample fractions via mass spectrometry (paras. 0037 and 0085). Kobold exemplifies analytes including proteins and lipids (paras. 0027, 0033-34, 0076 and 0115), and biological sample types including plasma, blood and serum (paras. 0034-35). In this way, Kobold teaches adsorption of proteins to nanoparticles.
Kobold further teaches that the combined application of these techniques allows for effective separation and analysis of compounds having a low molecular weight (paras. 0099 and 0109-0112). Kobold does not teach embodiments wherein a classifier assigns a classification using fewer than 100 features from the multi-omic data.
El-Khoury presents a detection method for lung cancer including a model that uses 6 protein features, quantified by mass spectrometry of plasma samples, and distinguishes lung cancer with an AUC value of 0.999 (pg. 1, Abstract; pg. 5, para. 1; pg. 6, Table 1).
Chen presents a predictive model for breast cancer that uses 15 lipid features, quantified by mass spectrometry of plasma samples, and distinguishes early-stage breast cancer with mean AUC values of 0.926 observed in training and 0.938 observed in validation (pg. 36622, Results; pg. 36623, r. column; pg. 36627, Fig. 3).
The combined teachings of El-Khoury and Chen indicate the feasibility of developing a classifier that assigns a cancer classification using well under 100 total protein and lipid features and has a performance characterized by an average AUC of well over 0.84. Therefore, the combination of Drake, Kobold, Chen and El-Khoury is considered to teach or suggest every limitation of the claim.
With respect to claim 3, El-Khoury presents a detection method for lung cancer including a model that uses 6 protein features (pg. 1, Abstract; pg. 5, para. 1; pg. 6, Table 1) and Chen presents a predictive model for breast cancer that uses 15 lipid features (pg. 36622).
With respect to claim 4, Drake discloses identifying features corresponding to each of the assayed classes of molecules (para. 0010), wherein classes of molecules may include lipids (paras. 0014 and 0115). In this way, Drake discloses lipid features.
With respect to claim 7, Drake discloses classification based on age features (paras. 0072 and 0490).
With respect to claim 8, Drake discloses analyzing a different analyte in each of a plurality of sample aliquots (para. 0121), and further discloses generating nucleic acid measurements (paras. 0015 and 0034) via whole transcriptome shotgun sequencing, i.e., a transcriptomic assay (para. 0187).
With respect to claim 9, Drake discloses analysis of mRNA, microRNA, and CpG methylation (paras. 0012 and 0018).
With respect to claim 10, Drake discloses analysis of a combination of analytes (para. 0118) assayed via a combination of sequencing techniques (para. 0129). Drake is considered to disclose combined analysis of the exemplified mRNA, microRNA and methylation features.
With respect to claim 11, Drake discloses identifying features corresponding to each of the assays performed and training a classifier based on features (paras. 0010 and 0022).
With respect to claim 16, Drake discloses generating a subset of features from lipid measurements; generating a subset of features from proteomic measurements; preparing a feature vector of pooled features; and generating a classification based on pooled features (paras. 0010-0020).
With respect to claim 17, Drake discloses assessment of univariate difference to identify an optimal predictive set of features, i.e., top features (paras. 0230, 0240-41 and 0492-93).
With respect to claim 20, Drake discloses SVMs, random forest classifiers and naïve Bayes classifiers (para. 0020). Drake also exemplifies a hidden Markov model (para. 0489).
With respect to claim 21, Drake discloses prediction of subject response to chemotherapy, drug therapy, radiation therapy, and surgical intervention based on classification, and treatment based on predicted response (paras. 0091, 0103, 0107 and 0314).
With respect to claim 22, Drake exemplifies cancers including lung cancer, pancreatic cancer, colon cancer, liver cancer, breast cancer and ovarian cancer (paras. 0204 and 0326).
With respect to claim 23, Drake exemplifies detection of early-stage cancer (para. 0305).
With respect to claim 24, Drake discloses that biofluid samples can include blood, serum or plasma (para. 0080).
With respect to claim 31, Drake discloses classifier performance thresholding based on a minimum AUC value of 0.9 (para. 0231). Any classifier passing this performance threshold would exhibit an average or median AUC of at least 0.9.
An invention would have been obvious to one of ordinary skill in the art if some teaching in the prior art would have led that person to combine prior art reference teachings to arrive at the claimed invention. Before the effective filing date of the claimed invention, said practitioner would have implemented steps of adsorbing proteins to nanoparticles and assaying using mass spectrometry, as taught by Kobold, to enhance the cancer diagnostic methods taught by Drake, because Kobold teaches that the combined application of these techniques allows for effective separation and analysis of compounds having a low molecular weight (paras. 0099 and 0109-0112). Said practitioner would have had a reasonable expectation of success because Drake and Kobold both discuss mass spectral analysis of proteins, lipids and metabolites in plasma samples.
An invention would have been obvious to one of ordinary skill in the art if some teaching in the prior art would have led that person to combine prior art reference teachings to arrive at the claimed invention. Before the effective filing date of the claimed invention, said practitioner would have employed the cancer diagnostic methodology of Drake to generate a classifier that uses less than 100 features from multi-omic data as claimed, because Drake discloses development of a classifier that uses protein and lipid features and development of classifiers that achieve an AUC of at least 0.84 while El-Khoury and Chen disclose cancer classifiers that collectively use less than 100 protein and lipid features and achieve AUCs of over 0.84. Thus, El-Khoury and Chen indicate the feasibility of generating a cancer classifier that uses a very small set of features (21, i.e., less than 100), of a combined form (i.e. protein and lipid features) disclosed by Drake, to achieve a performance threshold (i.e., an AUC of at least 0.84) disclosed by Drake. As one of ordinary skill in the art would be aware, minimizing required input features is desirable to reduce the complexity (costs, turnaround time, etc.) of clinical testing necessary to produce said features.
Said practitioner would have had a reasonable expectation of success because Drake, El-Khoury and Chen all discuss classification of cancer via high-performing models trained on molecular features measured by mass spectral analysis of subject plasma samples.
In this way the disclosure of Drake, in view of Kobold, El-Khoury and Chen, makes obvious the limitations of claims 1, 3-4, 7-11, 16-17, 20-24 and 31. Thus, the invention is prima facie obvious.
Response to Arguments - Double Patenting
In the remarks filed 10/17/2025, Applicant traverses the rejection on grounds of nonstatutory double patenting and presents supporting arguments.
Applicant alleges that claim 1 and dependents thereof, as amended, are patentably distinct from the claims in the ‘190 application, now US Patent No 12,334,190 (pg. 8, para. 5). No particular points of distinction are highlighted. Thus, the argument of patentable distinction is found unpersuasive and the rejection is maintained.
Double Patenting
The nonstatutory double patenting rejection is based on a judicially created doctrine grounded in public policy (a policy reflected in the statute) so as to prevent the unjustified or improper timewise extension of the “right to exclude” granted by a patent and to prevent possible harassment by multiple assignees. A nonstatutory double patenting rejection is appropriate where the conflicting claims are not identical, but at least one examined application claim is not patentably distinct from the reference claim(s) because the examined application claim is either anticipated by, or would have been obvious over, the reference claim(s). See, e.g., In re Berg, 140 F.3d 1428, 46 USPQ2d 1226 (Fed. Cir. 1998); In re Goodman, 11 F.3d 1046, 29 USPQ2d 2010 (Fed. Cir. 1993); In re Longi, 759 F.2d 887, 225 USPQ 645 (Fed. Cir. 1985); In re Van Ornum, 686 F.2d 937, 214 USPQ 761 (CCPA 1982); In re Vogel, 422 F.2d 438, 164 USPQ 619 (CCPA 1970); In re Thorington, 418 F.2d 528, 163 USPQ 644 (CCPA 1969).
A timely filed terminal disclaimer in compliance with 37 CFR 1.321(c) or 1.321(d) may be used to overcome an actual or provisional rejection based on nonstatutory double patenting provided the reference application or patent either is shown to be commonly owned with the examined application, or claims an invention made as a result of activities undertaken within the scope of a joint research agreement. See MPEP § 717.02 for applications subject to examination under the first inventor to file provisions of the AIA as explained in MPEP § 2159. See MPEP § 2146 et seq. for applications not subject to examination under the first inventor to file provisions of the AIA . A terminal disclaimer must be signed in compliance with 37 CFR 1.321(b).
The filing of a terminal disclaimer by itself is not a complete reply to a nonstatutory double patenting (NSDP) rejection. A complete reply requires that the terminal disclaimer be accompanied by a reply requesting reconsideration of the prior Office action. Even where the NSDP rejection is provisional the reply must be complete. See MPEP § 804, subsection I.B.1. For a reply to a non-final Office action, see 37 CFR 1.111(a). For a reply to final Office action, see 37 CFR 1.113(c). A request for reconsideration while not provided for in 37 CFR 1.113(c) may be filed after final for consideration. See MPEP §§ 706.07(e) and 714.13.
The USPTO Internet website contains terminal disclaimer forms which may be used. Please visit www.uspto.gov/patent/patents-forms. The actual filing date of the application in which the form is filed determines what form (e.g., PTO/SB/25, PTO/SB/26, PTO/AIA /25, or PTO/AIA /26) should be used. A web-based eTerminal Disclaimer may be filled out completely online using web-screens. An eTerminal Disclaimer that meets all requirements is auto-processed and approved immediately upon submission. For more information about eTerminal Disclaimers, refer to www.uspto.gov/patents/apply/applying-online/eterminal-disclaimer.
Instant claims 1, 8-9, 16-17, 20-22 and 24 are rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1-3, 6, 12-15, 17 and 19 of U.S. Patent No. 12,334,190 (issued 6/17/2025; on IDS filed 10/17/2025; hereafter, “‘190”), in view of Drake, El-Khoury and Chen. ‘190 shares joint inventors (Ma, Philip; Wilcox, Bruce; Collin, Francois; Belthangady, Chinmay; Yang, Mi; Khadka, Manoj; Liu, Manway; Blume, John; Langer, Robert S. JR.; Khaledian, Ehdieh) and a common assignee (PrognomIQ, Inc.) with the instant application.
A substantially similar nonstatutory double patenting rejection was previously presented as a provisional rejection over claims of Application No. 17/709,185, which has issued as US 12,334,190 since the previous Office action. The new grounds of rejection presented herein, relative to the previous provisional rejection, were necessitated by Applicant’s amendment of the claims (filed 10/17/2025).
Instant claim 1 discloses a multi-omic method comprising steps of: dividing a biofluid sample, from a subject suspected of having cancer, into at least first and second portions; forming biomolecule coronas, comprising adsorbed proteins, by contacting the first portion with nanoparticles; assaying the adsorbed proteins or fragments thereof, using mass spectrometry or an immunoassay, to generate proteomic measurements; assaying the second portion, using mass spectrometry or an immunoassay, to detect lipids and generate lipid measurements; obtaining multi-omic data comprising the proteomic and lipid measurements; and processing the multi-omic data, using a classifier, to assign a likelihood of cancer.
The claim further requires that: the classifier assigns the classification using fewer than 100 features from the multi-omic data; performance of the classifier is characterized by a receiver operating characteristic (ROC) curve having an average or median area under the curve (AUC) of at least 0.84; and the classifier is trained using samples from subjects with cancer and control samples from subjects without cancer.
With respect to instant claim 1, ‘190 discloses obtaining proteomic data from a biofluid of a subject suspected of having cancer (claim 1) comprising forming biomolecule coronas by protein adsorption to nanoparticles (claim 2), and assaying using mass spectrometry (claim 3).
‘190 further discloses inputting combined data into a classifier to identify a likelihood of cancer, wherein the classifier is trained on cancer and non-cancer samples (claim 1) and has a performance characterized by an average or median AUC of at least 0.85 (claim 21).
‘190 does not disclose measurement of lipids; dividing the sample into portions; or embodiments wherein the classifier assigns the classification using fewer than 100 features from the multi-omic data.
Drake discloses applying classifiers to multi-omic data (paras. 0051-56, 0063-64, 0069-70, 0122, 0185 and 0249) to evaluate cancer (paras. 0051, 0057 and 0066), and exemplifies analysis of protein and lipid features (paras. 011-14). Drake further exemplifies analysis of sample portions (paras. 0135-36).
Drake additionally discusses restriction to analysis of a single class of molecules, e.g., only circulating proteins, as a relative deficiency of prior methods (para. 0003). Drake thus teaches that multi-omic analysis is functionally superior to utilization of protein features alone.
Drake does not teach embodiments wherein the classifier assigns the classification using fewer than 100 features from the multi-omic data.
El-Khoury presents a detection method for lung cancer including a model that uses 6 protein features, quantified by mass spectrometry of plasma samples, and distinguishes lung cancer with an AUC value of 0.999 (pg. 1, Abstract; pg. 5, para. 1; pg. 6, Table 1).
Chen presents a predictive model for breast cancer that uses 15 lipid features, quantified by mass spectrometry of plasma samples, and distinguishes early-stage breast cancer with mean AUC values of 0.926 observed in training and 0.938 observed in validation (pg. 36622, Results; pg. 36623, r. column; pg. 36627, Fig. 3).
The combined teachings of El-Khoury and Chen indicate the feasibility of developing a classifier that assigns a cancer classification using well under 100 total protein and lipid features and has a performance characterized by an average AUC of well over 0.84. Therefore, the combination of ‘190, Drake, Chen and El-Khoury is considered to teach or suggest every limitation of the claim.
With respect to instant claim 8, ‘190 discloses analysis of combined data that includes proteomic data and nucleic acid measurements (claim 1).
With respect to instant claim 9, ‘190 discloses that the nucleic acids comprise mRNA and/or miRNA (claim 6).
With respect to instant claim 16, ‘190 discloses identifying and pooling subsets of features among the proteomic data and nucleic acid data (claims 12) having improved classification performance (claim 13).
With respect to instant claim 17, ‘190 discloses identifying top features having improved classification performance by analyzing univariate data (claims 12-13).
With respect to instant claim 20, ‘190 discloses that the classifier is trained using deep
learning, a random forest classification analysis, a support vector machine analysis, a naive Bayes analysis, or a hidden Markov analysis (claim 14).
With respect to instant claim 21, ‘190 discloses identifying, via the classifier, a biological state as indicative of cancer and administering a chemotherapy, radiation or surgical cancer treatment to the subject (claim 17).
With respect to instant claim 22, ‘190 discloses that the cancer is selected from
the group consisting of: lung cancer, pancreatic cancer, colon cancer, liver cancer, breast cancer and ovarian cancer (claim 15).
With respect to instant claim 24, ‘190 discloses that the biofluid sample comprises a blood, serum, or plasma sample (claim 19).
An invention would have been obvious to one of ordinary skill in the art if some teaching in the prior art would have led that person to combine prior art reference teachings to arrive at the claimed invention. Before the effective filing date of the claimed invention, said practitioner would have implemented analysis of lipid features in combination with protein features, as taught by Drake, to enhance the cancer diagnostic methods taught by ‘190, because Drake discusses restriction to analysis of a single class of molecules, e.g., only circulating proteins, as a relative deficiency of prior methods (para. 0003) and thus teaches that such multi-omic analysis is functionally superior to utilization of protein features alone. Said practitioner would have had a reasonable expectation of success because ‘190 and Drake both discuss classification of cancer via models trained on protein features measured by mass spectral analysis of subject plasma samples.
An invention would have been obvious to one of ordinary skill in the art if some teaching in the prior art would have led that person to combine prior art reference teachings to arrive at the claimed invention. Before the effective filing date of the claimed invention, said practitioner would have employed the cancer diagnostic methodology of ‘190, in view of Drake, to generate a classifier that uses less than 100 features from multi-omic data as claimed, because ‘190 discloses development of a classifier that uses protein features and achieves an AUC of at least 0.85 while El-Khoury and Chen disclose cancer classifiers that collectively use less than 100 protein and lipid features and achieve higher AUC values (respectively, 0.999 and 0.926/0.938). Thus, El-Khoury and Chen indicate the feasibility of generating a cancer classifier that uses a very small set of features (21, i.e., less than 100), of a combined form (i.e. protein and lipid features) taught advantageously by Drake, to achieve well above a performance threshold disclosed by ‘190. As one of ordinary skill in the art would be aware, minimizing required input features is desirable to reduce the complexity (costs, turnaround time, etc.) of clinical testing necessary to produce said features.
In this way the claims of ‘190, in view of Drake, El-Khoury and Chen are patentably indistinct from the limitations of instant claims 1, 8-9, 16-17, 20-22 and 24.
Instant claim 1 is rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1, 9, 11-12 and 14-18 of U.S. Patent No. 12,387,508 (issued 8/12/2025; on IDS filed 10/17/2025; hereafter, “‘508”), in view of Drake, El-Khoury and Chen. ‘508 shares joint inventors (Ma, Philip; Wilcox, Bruce; Belthangady, Chinmay; Liu, Manway) and a common assignee (PrognomIQ, Inc.) with the instant application.
Instant claim 1 discloses a multi-omic method comprising steps of: dividing a biofluid sample, from a subject suspected of having cancer, into at least first and second portions; forming biomolecule coronas, comprising adsorbed proteins, by contacting the first portion with nanoparticles; assaying the adsorbed proteins or fragments thereof, using mass spectrometry or an immunoassay, to generate proteomic measurements; assaying the second portion, using mass spectrometry or an immunoassay, to detect lipids and generate lipid measurements; obtaining multi-omic data comprising the proteomic and lipid measurements; and processing the multi-omic data, using a classifier, to assign a likelihood of cancer.
The claim further requires that: the classifier assigns the classification using fewer than 100 features from the multi-omic data; performance of the classifier is characterized by a receiver operating characteristic (ROC) curve having an average or median area under the curve (AUC) of at least 0.84; and the classifier is trained using samples from subjects with cancer and control samples from subjects without cancer.
With respect to instant claim 1, ‘508 discloses a computer system comprising computer processor(s) and a computer readable medium comprising machine-executable code that, upon execution by the processor(s), implements a method comprising the function of: combining arrays of intensity values to generate a multi-dimensional dataset by aligning the arrays based on identified mass spectrometry features (claim 1), wherein the intensity values relate to abundances of distinct groups of biological species of biological sample(s) (claim 11) that are generated by mass spectrometry (claim 12), the distinct groups of biological species comprise proteins (claim 14) which have been adsorbed to nanoparticles (claim 15), and the biological sample(s) comprise a comprise a biofluid sample (claim 16).
‘508 further discloses the code-implemented function of: applying a classifier to the dataset to generate a label corresponding to a biological state (claim 1), wherein the features are identified from training datasets obtained from subjects having the biological state and from subjects not having the biological state (claim 9) and the biological state comprises a healthy state or a disease state (claim 17) that is a cancer state (claim 18).
‘508 does not disclose dividing the sample into portions; measurement of lipids; embodiments wherein the classifier assigns the classification using fewer than 100 features from multi-omic data; or embodiments wherein performance of the classifier is characterized by an AUC of at least 0.84.
Drake discloses applying classifiers to multi-omic data (paras. 0051-56, 0063-64, 0069-70, 0122, 0185 and 0249) to evaluate cancer (paras. 0051, 0057 and 0066), and exemplifies analysis of protein and lipid features (paras. 011-14) in sample portions (paras. 0135-36).
Drake additionally discusses restriction to analysis of a single class of molecules, e.g., only circulating proteins, as a relative deficiency of prior methods (para. 0003). Drake thus teaches that multi-omic analysis is functionally superior to utilization of protein features alone.
Drake also teaches performance thresholding of classifiers based on a minimum AUC value of 0.84 (para. 0231), but does not teach embodiments wherein the classifier assigns the classification using fewer than 100 features from the multi-omic data.
El-Khoury presents a detection method for lung cancer including a model that uses 6 protein features, quantified by mass spectrometry of plasma samples, and distinguishes lung cancer with an AUC value of 0.999 (pg. 1, Abstract; pg. 5, para. 1; pg. 6, Table 1).
Chen presents a predictive model for breast cancer that uses 15 lipid features, quantified by mass spectrometry of plasma samples, and distinguishes early-stage breast cancer with mean AUC values of 0.926 observed in training and 0.938 observed in validation (pg. 36622, Results; pg. 36623, r. column; pg. 36627, Fig. 3).
The combined teachings of El-Khoury and Chen indicate the feasibility of developing a classifier that assigns a cancer classification using well under 100 total protein and lipid features and has a performance characterized by an average AUC of well over 0.84. Therefore, the combination of ‘508, Drake, Chen and El-Khoury is considered to teach or suggest every limitation of the claim.
An invention would have been obvious to one of ordinary skill in the art if some teaching in the prior art would have led that person to combine prior art reference teachings to arrive at the claimed invention. Before the effective filing date of the claimed invention, said practitioner would have implemented analysis of lipid features in combination with protein features, as taught by Drake, to enhance the cancer diagnostic methods taught by ‘508, because Drake discusses restriction to analysis of a single class of molecules, e.g., only circulating proteins, as a relative deficiency of prior methods (para. 0003) and thus teaches that such multi-omic analysis is functionally superior to utilization of protein features alone. Said practitioner would have had a reasonable expectation of success because ‘508 and Drake both discuss classification of cancer via models trained on protein features measured by mass spectral analysis of subject plasma samples.
An invention would have been obvious to one of ordinary skill in the art if some teaching in the prior art would have led that person to combine prior art reference teachings to arrive at the claimed invention. Before the effective filing date of the claimed invention, said practitioner would have employed the cancer diagnostic methodology of ‘508, in view of Drake, to generate a classifier that uses less than 100 features from multi-omic data as claimed, because Drake teaches that multi-omic analysis is functionally superior to utilization of protein features alone (para. 003) while El-Khoury and Chen disclose cancer classifiers that collectively use less than 100 protein and lipid features and achieve high AUC values (respectively, 0.999 and 0.926/0.938). Thus, El-Khoury and Chen indicate the feasibility of generating a cancer classifier that uses a very small set of features (21, i.e., less than 100), of a combined form (i.e. protein and lipid features) taught advantageously by Drake, to achieve high performance in classifying cancer state as disclosed by ‘508. As one of ordinary skill in the art would be aware, minimizing required input features is desirable to reduce the complexity (costs, turnaround time, etc.) of clinical testing necessary to produce said features.
In this way the claims of ‘508, in view of Drake, El-Khoury and Chen are patentably indistinct from the limitations of instant claim 1.
Instant claim 1 is provisionally rejected on the ground of nonstatutory double patenting as being unpatentable over claims 99 and 103-104 of co-pending Application No.19/039,712 (hereafter, “‘712”), in view of Kobold, Drake, El-Khoury and Chen. ‘712 shares joint inventors (Ma, Philip; Wilcox, Bruce; Collin, Francois; Belthangady, Chinmay; Yang, Mi; Khadka, Manoj; Liu, Manway; Blume, John; Khaledian, Ehdieh) and a common assignee (PrognomIQ, Inc.) with the instant application.
Instant claim 1 discloses a multi-omic method comprising steps of: dividing a biofluid sample, from a subject suspected of having cancer, into at least first and second portions; forming biomolecule coronas, comprising adsorbed proteins, by contacting the first portion with nanoparticles; assaying the adsorbed proteins or fragments thereof, using mass spectrometry or an immunoassay, to generate proteomic measurements; assaying the second portion, using mass spectrometry or an immunoassay, to detect lipids and generate lipid measurements; obtaining multi-omic data comprising the proteomic and lipid measurements; and processing the multi-omic data, using a classifier, to assign a likelihood of cancer.
The claim further requires that: the classifier assigns the classification using fewer than 100 features from the multi-omic data; performance of the classifier is characterized by a receiver operating characteristic (ROC) curve having an average or median area under the curve (AUC) of at least 0.84; and the classifier is trained using samples from subjects with cancer and control samples from subjects without cancer.
With respect to instant claim 1, ‘712 discloses a method comprising steps of: isolating plasma from a blood sample and producing a fraction of the plasma sample (claim 99); quantifying proteins of the plasma sample to obtain a plurality of proteomic measurements (claim 99), wherein the proteins are quantified with a combination of mass spectrometry and an immunosorbent assay (claims 103-104); and analyzing the proteomic measurements using a classifier to assign a degree of risk of a lung cancer in the subject, wherein the classifier is trained on lung cancer samples and non-cancer samples (claim 99).
‘712 does not disclose adsorbing proteins with nanoparticles; measurement of lipids; embodiments wherein the classifier assigns the classification using fewer than 100 features from multi-omic data; or embodiments wherein performance of the classifier is characterized by an AUC of at least 0.84.
Kobold teaches analysis of biological analytes comprising contacting a biological sample with magnetic nanoparticles; incubating the sample to adsorb analytes to the particles; fractionating the sample by eluting the liquid and/or washing the particles; and detecting the compounds in sample fractions via mass spectrometry (paras. 0037 and 0085). Kobold exemplifies analytes including proteins and lipids (paras. 0027, 0033-34, 0076 and 0115), and biological sample types including plasma, blood and serum (paras. 0034-35). In this way, Kobold teaches adsorption of proteins to nanoparticles.
Kobold further teaches that the combined application of these techniques allows for effective separation and analysis of compounds having a low molecular weight (paras. 0099 and 0109-0112). Kobold does not teach embodiments wherein the classifier assigns the classification using fewer than 100 features from multi-omic data; or embodiments wherein performance of the classifier is characterized by an AUC of at least 0.84.
Drake discloses applying classifiers to multi-omic data (paras. 0051-56, 0063-64, 0069-70, 0122, 0185 and 0249) to evaluate cancer (paras. 0051, 0057 and 0066), and exemplifies analysis of protein and lipid features (paras. 011-14) in sample portions (paras. 0135-36).
Drake additionally discusses restriction to analysis of a single class of molecules, e.g., only circulating proteins, as a relative deficiency of prior methods (para. 0003). Drake thus teaches that multi-omic analysis is functionally superior to utilization of protein features alone.
Drake also teaches performance thresholding of classifiers based on a minimum AUC value of 0.84 (para. 0231), but does not teach embodiments wherein the classifier assigns the classification using fewer than 100 features from the multi-omic data.
El-Khoury presents a detection method for lung cancer including a model that uses 6 protein features, quantified by mass spectrometry of plasma samples, and distinguishes lung cancer with an AUC value of 0.999 (pg. 1, Abstract; pg. 5, para. 1; pg. 6, Table 1).
Chen presents a predictive model for breast cancer that uses 15 lipid features, quantified by mass spectrometry of plasma samples, and distinguishes early-stage breast cancer with mean AUC values of 0.926 observed in training and 0.938 observed in validation (pg. 36622, Results; pg. 36623, r. column; pg. 36627, Fig. 3).
The combined teachings of El-Khoury and Chen indicate the feasibility of developing a classifier that assigns a cancer classification using well under 100 total protein and lipid features and has a performance characterized by an average AUC of well over 0.84. Therefore, the combination of ‘712, Kobold, Drake, Chen and El-Khoury is considered to teach or suggest every limitation of the claim.
An invention would have been obvious to one of ordinary skill in the art if some teaching in the prior art would have led that person to combine prior art reference teachings to arrive at the claimed invention. Before the effective filing date of the claimed invention, said practitioner would have implemented steps of adsorbing proteins to nanoparticles and assaying using mass spectrometry, as taught by Kobold, to enhance the cancer diagnostic methods taught by ‘712, because Kobold teaches that the combined application of these techniques allows for effective separation and analysis of compounds having a low molecular weight (paras. 0099 and 0109-0112). Said practitioner would have had a reasonable expectation of success because ‘712 and Kobold both discuss mass spectral analysis of proteins in plasma samples.
An invention would have been obvious to one of ordinary skill in the art if some teaching in the prior art would have led that person to combine prior art reference teachings to arrive at the claimed invention. Before the effective filing date of the claimed invention, said practitioner would have implemented analysis of lipid features in combination with protein features, as taught by Drake, to enhance the cancer diagnostic methods taught by ‘712, because Drake discusses restriction to analysis of a single class of molecules, e.g., only circulating proteins, as a relative deficiency of prior methods (para. 0003) and thus teaches that such multi-omic analysis is functionally superior to utilization of protein features alone. Said practitioner would have had a reasonable expectation of success because ‘712 and Drake both discuss classification of cancer via models trained on protein features measured by mass spectral analysis of subject plasma samples.
An invention would have been obvious to one of ordinary skill in the art if some teaching in the prior art would have led that person to combine prior art reference teachings to arrive at the claimed invention. Before the effective filing date of the claimed invention, said practitioner would have employed the cancer diagnostic methodology of ‘712, in view of Kobold and Drake, to generate a classifier that uses less than 100 features from multi-omic data as claimed, because Drake teaches that multi-omic analysis is functionally superior to utilization of protein features alone (para. 003) while El-Khoury and Chen disclose cancer classifiers that collectively use less than 100 protein and lipid features and achieve high AUC values (respectively, 0.999 and 0.926/0.938). Thus, El-Khoury and Chen indicate the feasibility of generating a cancer classifier that uses a very small set of features (21, i.e., less than 100), of a combined form (i.e. protein and lipid features) taught advantageously by Drake, to achieve high performance in classifying cancer state as disclosed by ‘712. As one of ordinary skill in the art would be aware, minimizing required input features is desirable to reduce the complexity (costs, turnaround time, etc.) of clinical testing necessary to produce said features.
In this way the claims of ‘712, in view of Kobold, Drake, El-Khoury and Chen are patentably indistinct from the limitations of instant claim 1.
Instant claim 1 is rejected on the ground of nonstatutory double patenting as being unpatentable over claims 1, 10, 12-13 and 15-18 of co-pending Application No. 19/263,039 (hereafter, “‘039”), in view of Drake, El-Khoury and Chen. ‘039 shares joint inventors (Ma, Philip; Wilcox, Bruce; Belthangady, Chinmay; Liu, Manway) and a common assignee (PrognomIQ, Inc.) with the instant application.
Instant claim 1 discloses a multi-omic method comprising steps of: dividing a biofluid sample, from a subject suspected of having cancer, into at least first and second portions; forming biomolecule coronas, comprising adsorbed proteins, by contacting the first portion with nanoparticles; assaying the adsorbed proteins or fragments thereof, using mass spectrometry or an immunoassay, to generate proteomic measurements; assaying the second portion, using mass spectrometry or an immunoassay, to detect lipids and generate lipid measurements; obtaining multi-omic data comprising the proteomic and lipid measurements; and processing the multi-omic data, using a classifier, to assign a likelihood of cancer.
The claim further requires that: the classifier assigns the classification using fewer than 100 features from the multi-omic data; performance of the classifier is characterized by a receiver operating characteristic (ROC) curve having an average or median area under the curve (AUC) of at least 0.84; and the classifier is trained using samples from subjects with cancer and control samples from subjects without cancer.
With respect to instant claim 1, ‘039 discloses a computer comprising computer processor(s) and a computer readable medium comprising machine-executable code that, upon execution by the processor(s), implements a method comprising the function of: comb