DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 01/06/2026 has been entered.
Status of Amendments
Claims 1-22 are currently pending in this case and have been examined and
addressed below. This communication is a Non-Final Rejection in response to the
Amendment to the Claims and Remarks filed on 01/06/2026.
Claims 1, 12, and 17 are amended claims.
Claims 2-3, 5, 13-14, and 18-19 are original claims.
Claims 4, 6, 7-11, 15-16, and 20-22 are previously presented.
Claims 23 and 24 have been cancelled and will not be considered at this time.
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 1-22 are rejected under 35 U.S.C. § 101 because the claimed
invention is directed to a judicial exception (i.e. an abstract idea) without
significantly more.
Step 1 – Statutory Categories of Invention:
Claims 1-22 are drawn to a method, system, and an article of manufacture, which
are statutory categories of invention.
Step 2A – Judicial Exception Analysis, Prong 1:
Independent claim 1 recites a method comprising predicting drug interaction outcomes for pathogens; using drug interaction outcome data for a plurality of pathogens; predicting drug interaction outcome data based on genetic information of a pathogen of interest; wherein the drug interaction outcome data includes, for each respective pathogen of the plurality of pathogens, an outcome of one or more drug treatments applied to the respective pathogen; and wherein (i) predicting at least one classification of outcomes for drug treatments applied to one or more pathogens of interest and (ii) generating a matrix including at least one of joint sensitivity or resistance profiles corresponding to at least one combination of the drug treatment based on omics profiles of each of the drug treatments in combination, and wherein generating the matrix includes: transforming omics profiles of each of the drug treatments into binary resistance profiles or response profiles, wherein for chemogenomics data the binary conversion determines deletion strains sensitive or resistant to a drug, and wherein for transcriptomics data the binary conversion determines genes as downregulated or upregulated by the drug; obtaining, genetic information of a pathogen of interest not in the plurality of pathogens; generating predicted drug interaction outcome data for one or more drug treatments of interest applied to the pathogen of interest, based on the genetic information of the pathogen of interest; and indicating the predicted drug interaction outcome data for the one or more drug treatments of interest applied to the pathogen of interest.
Independent claim 12 recites a system comprising predicting drug interaction outcomes for pathogens; using drug interaction outcome data for a plurality of pathogens; predicting drug interaction outcome data based on genetic information of a pathogen of interest; wherein the drug interaction outcome data includes, for each respective pathogen of the plurality of pathogens, an outcome of one or more drug treatments applied to the respective pathogen; and wherein (i) predicting at least one classification of outcomes for drug treatments applied to one or more pathogens of interest and (ii) generating a matrix including at least one of joint sensitivity or resistance profiles corresponding to at least one combination of the drug treatment based on omics profiles of each of the drug treatments in combination, and wherein generating the matrix includes: transforming omics profiles of each of the drug treatments into binary resistance profiles or response profiles, wherein for chemogenomics data the binary conversion determines deletion strains sensitive or resistant to a drug, and wherein for transcriptomics data the binary conversion determines genes as downregulated or upregulated by the drug; obtain, genetic information of a pathogen of interest not in the plurality of pathogens; generate predicted drug interaction outcome data for one or more drug treatments of interest applied to the pathogen of interest, based on the genetic information of the pathogen of interest; and indicate the predicted drug interaction outcome data for the one or more drug treatments of interest applied to
the pathogen of interest.
Independent claim 17 recites a non-transitory computer-readable medium comprising predicting drug interaction outcomes for pathogens; using drug interaction outcome data for a plurality of pathogens; predicting drug interaction outcome data based on genetic information of a pathogen of interest; wherein the drug interaction outcome data includes, for each respective pathogen of the plurality of pathogens, an outcome of one or more drug treatments applied to the respective pathogen; and wherein (i) predicting at least one classification of outcomes for drug treatments applied to one or more pathogens of interest and (ii) generating a matrix including at least one of joint sensitivity or resistance profiles corresponding to at least one combination of the drug treatment based on omics profiles of each of the drug treatments in combination, and wherein generating the matrix includes: transforming omics profiles of each of the drug treatments into binary resistance profiles or response profiles, wherein for chemogenomics data the binary conversion determines deletion strains sensitive or resistant to a drug, and wherein for transcriptomics data the binary conversion determines genes as downregulated or upregulated by the drug; obtain,
genetic information of a pathogen of interest not in the plurality of pathogens; generate
predicted drug interaction outcome data for one or more drug treatments of interest applied to the pathogen of interest, based on the genetic information of the pathogen of
interest; and indicate the predicted drug interaction outcome data for the one or more
drug treatments of interest applied to the pathogen of interest.
Independent claim 21 recites a method comprising obtaining a set of training data for a plurality of pathogens including actual outcomes of one or more drug treatments applied to each of the plurality of pathogens; classifying the set of training data into a plurality of subsets each corresponding to a different actual outcome or a range of actual outcomes; and predicting an outcome of applying a drug treatment of interest to a pathogen of interest using the classified subsets of training data, by (i) predicting at least one classification of outcomes for drug treatments applied to one or more pathogens of interest and (ii) generating a matrix including at least one of joint
sensitivity or resistance profiles corresponding to at least one combination of the drug
treatment based on omics profiles of each of the drug treatments in combination.
These steps amount to certain methods of organizing human activity which
includes functions relating to managing personal behavior or relationships or
interactions between people (including social activities, teaching, and following rules or
instructions) (MPEP § 2106.04(a)(2)(II)(C) citing the abstract idea grouping for methods
of organizing human activity for managing personal behavior or relationships or
interactions between people – also note MPEP § 2106.04(a)(2)(II) stating certain activity
between a person and a computer may fall within the “certain methods of organizing
human activity” grouping).
Step 2A – Judicial Exception Analysis, Prong 2:
This judicial exception is not integrated into a practical application because the additional elements within the claims only amount to instructions to implement the judicial exception using a computer [MPEP 2106.05(f)].
The claims recite the additional elements of by one or more processors, a computer system, memory, and a tangible non-transitory computer readable medium.
These elements are recited at a high-level of generality such that it amounts to mere instructions to apply the exception because this is an example of applying the abstract idea by use of general-purpose computer which does not integrate the abstract idea into a practical application.
Claims 1, 12, and 17 recite obtaining a base multispecies transfer machine learning model; wherein the base multispecies transfer machine learning model is trained; wherein the base multispecies transfer machine learning model is used to train a fine-tuned multispecies transfer machine learning model; the base multispecies transfer machine learning model is trained; generating ortholog predictions by determining, using reciprocal-best- BLAST hits (RBBH), gene orthology between the plurality of pathogens; obtaining the fine-tuned multispecies transfer machine learning model; wherein the fine-tuned model is trained using a transfer learning process by modifying at least one of an underlying weight or structure of the base multispecies transfer learning model based on the pathogen of interest; and using the fine-tuned multispecies transfer machine learning model. Claim 21 recites generating the statistical model and wherein the statistical model is a base multispecies transfer machine learning model trained. These limitations are recited as a tool to apply data to an algorithm and report the results (MPEP § 2106.05(f)(2) see case involving a commonplace business method or mathematical algorithm being applied on a general purpose computer within the “Other examples.. i.”) amounting to instruction to implement the abstract idea using a general purpose computer. Alice Corp. Pty. Ltd. V. CLS Bank Int’l, 134 S. Ct. 2347, 1357 (2014).
The above claims, as a whole, are therefore directed to an abstract idea.
Step 2B – Additional Elements that Amount to Significantly More:
The present claims do not include additional elements that are sufficient to amount to more than the abstract idea because the additional elements or combination of elements amount to no more than a recitation of instructions to implement the abstract idea on a computer.
As discussed above with respect to integration of the abstract idea into a
practical application, the claims recite the additional elements of one or more processors, a computer system, memory, and a tangible non-transitory computer readable medium.
Thus, taken alone, the additional elements do not amount to significantly more than the above-identified judicial exception. Looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. Their collective functions merely provide conventional computer implementation.
Claims 1, 12, 17, and 21 recite a base multispecies transfer machine learning model, wherein the base multispecies transfer machine learning model is used to train a fine-tuned multispecies transfer machine learning model, the fine-tuned model is trained using a transfer learning process, and using reciprocal-best- BLAST hits (RBBH). These limitations are only recited as a tool for performing steps of the abstract idea, therefore these limitations amount to mere instructions to perform the abstract idea using a computer and are not sufficient to amount to significantly more than the abstract idea (MPEP 2106.05(f) see for additional guidance on the “mere instructions to apply an exception”).
For the reasons stated, these claims are consequently rejected under 35 U.S.C. § 101.
Analysis of Dependent Claims
Dependent claims 2, 13, and 18 recite the drug interaction outcome data further includes one or both of genetic information or clinical information of each of a plurality of living subjects and each of the plurality of living subjects has at least one of the plurality of pathogens.
Dependent claims 3,14, and 19 recite obtaining, by the one or more processors, one or both of genetic information or clinical information of a living subject of interest having the pathogen of interest; and wherein generating, by the one or more processors using the machine learning model, the predicted drug interaction outcome data for the one or more drug treatments of interest applied to the pathogen of interest, is further based on one or both of the genetic information or the clinical information of the living subject of interest.
Dependent claims 4, 15, and 20 recite wherein one or both of: (i) each drug treatment of the one or more drug treatments included in the drug interaction outcome data includes a plurality of individual drugs, and, each respective outcome of the one or more drug treatments includes an indication of a measure of synergistic interaction of the plurality of individual drugs or a measure of antagonistic interaction of the plurality of individual drugs, or (ii) each drug treatment of interest of the one or more drug treatments of interest includes a plurality of individual drugs of interest, and, each respective outcome of the one or more drug treatments of interest includes an indication of a measure of synergistic interaction of the plurality of individual drugs of interest or a measure of antagonistic interaction of the plurality of individual drugs of interest.
Dependent claims 6 and 16 recite wherein the plurality of pathogens of the drug interaction outcome data include at least two different pathogen strains, and, for each of the at least two different pathogen strains, the drug interaction outcome data includes one or more of chemogenomics data, transcriptomics data or gene orthology data.
Dependent claim 7 recites obtaining, by the one or more processors, drug information for a plurality of individual drugs of interest, wherein each of the one or more drug treatments of interest includes two or more individual drugs of interest of the plurality of individual drugs of interest.
Dependent claim 8 recites wherein one or both of: (i) the pathogen of interest is not one of the plurality of pathogens of the drug interaction outcome data, or (ii) at least one of the one or more drug treatments of interest is not one of the one or more drug treatments of the drug interaction outcome data.
Dependent claim 10 recites identifying, by the one or more processors, a recommended drug treatment out of the one or more drug treatments of interest based on the predicted drug interaction outcome data for each of the one or more drug treatments of interest; and indicating, by the one or more processors, the recommended drug treatment.
Each of these steps of the preceding dependent claims 2-4, 6-8, 10, 13-16, 18-20, and 22 only serve to further limit or specify the features of independent claims 1, 12, 17, or 21 accordingly, and hence are nonetheless directed towards fundamentally the same abstract idea as the independent claim and utilize the additional elements analyzed below in the expected manner.
Dependent claim 5 recites wherein either the measure of synergistic interaction or the measure of antagonistic interaction, for one or both of (i) each of the individual drugs of the drug treatments, (ii) or each of the individual drugs of interest of the drug treatments of interest, includes one or more scores which are determined using one or both of the Loewe Additivity model or the Bliss Independence model. The Loewe Additivity model and the Bliss Independence model are additional elements, which is mere instructions to apply the exception and does not provide a practical application or significantly more for the same reasons.
Dependent claim 9 recites the machine learning model is a random forest model. The random forest model is an additional element, which is mere instructions to apply the exception and does not provide a practical application or significantly more for the same reasons.
Dependent claim 11 recites analyzing, by the one or more processors, feedback from a user following the indication of the predicted drug interaction outcome data for the one or more drug treatments of interest applied to the pathogen of interest, the feedback regarding the predicted drug interaction outcome data and including user input; and updating, by the one or more processors, the machine learning model based on the feedback from the user. The machine learning updating techniques are additional elements, which is mere instructions to apply the exception and does not provide a practical application or significantly more for the same reasons.
Dependent claim 22 recites wherein the base multispecies transfer machine learning model is used to train a fine-tuned multispecies transfer machine learning model for predicting drug interaction outcome data based on genetic information of the pathogen of interest, and wherein the fine-tuned multispecies transfer machine learning model is trained using a transfer learning process by modifying at least one of an underlying weight or structure of the base multispecies transfer learning model based on the pathogen of interest. This limitation is recited to carry out the steps of the abstract idea is mere instructions to apply the exception because a mathematical algorithm applied on a general-purpose computer has been found by the courts to be mere instructions to apply as in MPEP 2106.05(f)(2).
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1-3, 6, 9-10, 12-14, and 16-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Eltoukhy (US 20200395100 A1) in view of Altman (US 20230402127 A1) in view of Larder (US 20030190603 A1) in view of Wei (Geptop: A Gene Essentiality Prediction Tool for Sequenced Bacterial Genomes Based on Orthology and Phylogeny (2013)) in view of Chandrasekaran (Chemogenomics and orthology‐based design of antibiotic combination therapies (2016)).
As per Claim 1, Eltoukhy discloses a computer-implemented method of using transfer machine learning for predicting drug interaction outcomes for pathogens, comprising:
obtaining, by one or more processors, a base multispecies transfer machine learning model for predicting drug interaction outcomes for pathogens, wherein the base multispecies transfer machine learning model is trained using drug interaction outcome data for a plurality of pathogens, ([Para. 0006] receiving into computer memory a training dataset comprising, for each of a plurality of individuals having a disease, (1) genetic information from the individual generated at first time point and (2) treatment response of the individual to one or more therapeutic interventions determined at a second, later, time point; and implementing a machine learning algorithm using the dataset to generate at least one computer implemented classification algorithm. [Para. 0142] used to monitor systemic infections themselves, as may be caused by a pathogen such as a bacteria or virus. Copy number variation or even mutation detection may be used to determine how a population of pathogens are changing during the course of infection. This may be particularly important during chronic infections, such as HIV/AIDs or Hepatitis infections, whereby viruses may change life cycle state and/or mutate into more virulent forms during the course of infection. [Para. 0159] performed by a programmable processor)
wherein the drug interaction outcome data includes, for each respective pathogen of the plurality of pathogens, an outcome of one or more drug treatments applied to the respective pathogen, ([Para. 0006] receiving into computer memory a training dataset comprising, for each of a plurality of individuals having a disease, (1) genetic information from the individual generated at first time point and (2) treatment response of the individual to one or more therapeutic interventions determined at a second, later, time point; and implementing a machine learning algorithm using the dataset to generate at least one computer implemented classification algorithm. [Para. 0142] used to monitor systemic infections themselves, as may be caused by a pathogen such as a bacteria or virus. Copy number variation or even mutation detection may be used to determine how a population of pathogens are changing during the course of infection. This may be particularly important during chronic infections, such as HIV/AIDs or Hepatitis infections, whereby viruses may change life cycle state and/or mutate into more virulent forms during the course of infection. [Para. 0159] performed by a programmable processor)
Eltoukhy does not explicitly disclose, however Altman teaches
wherein the base multispecies transfer machine learning model is used to train a fine-tuned multispecies transfer machine learning model for predicting drug interaction outcome data based on genetic information of a pathogen of interest, ([Para. 0009] the model is trained by: a fine-tuning task, comprising: 1) receiving a fine-tuning dataset comprising a plurality of batches of naturally occurring sequences, wherein the fine-tuning dataset is a subset of the pre-training dataset, or a set of sequences that are related to the pre-training dataset by common ancestry, homology, or multiple sequence alignment; 2) inputting each batch of sequences into the language model, wherein the model is configured to output a fine-tuning set of semantic features.)
and wherein the base multispecies transfer machine learning model is trained by (i) predicting at least one classification of outcomes for drug treatments applied to one or more pathogens of interest and (ii) generating a matrix including at least one of joint sensitivity or resistance profiles corresponding to at least one combination of the drug treatment based on omics profiles of each of the drug treatments in combination and wherein generating the matrix includes: ([Para. 0009] The model is trained by a transfer learning task, comprising: 1) receiving a final training dataset comprising labeled sequences mapped to effects; and 2) training a neural network model based on the final training dataset, wherein the neural network model is configured to receive data corresponding to the pre-training set of semantic features and/or the fine-tuning set of semantic features, and output one or more effect scores. [Para. 0071] The term “semantic feature” refers to a representation of how the elements relate to or connect with each other in the input sequence data. In some embodiments, the representation is mathematical or numerical. In some embodiments, the semantic features may be a human and/or machine interpretable representation of the state of the input sequence. The output semantic features may be presented in a vector or a matrix (i.e. generating a matrix), and may be used as input for a downstream task, such as in transfer learning. [Para. 0010] the model is trained by: a) receiving a training dataset of sequences, comprising a training reference sequence and a training primary genetic variant, wherein the training primary genetic variant has an effect on the reference sequence with respect to a metric of interest; b) inputting the training dataset into a generative procedure configured to generate one or more training secondary genetic variants according to a random seed; c) calculating a loss function, wherein the loss function maps the combined effect of the primary and secondary genetic variants and the effect of the reference sequence onto a quantitative error score; d) accepting or rejecting the one or more training secondary genetic variants according to one or more predetermined acceptance criteria on the loss function; e) updating the generative procedure by incorporating the accepted one or more training secondary genetic variants in a new round of additional training secondary genetic variants; and f) repeating steps b) to e) until the loss converges to a minimum.)
obtaining, by the one or more processors (I) genetic information of a pathogen of interest not in the plurality of pathogens and (ii) the fine-tuned multispecies transfer machine learning model, ([Para. 0010] The model is trained by: a) receiving a training dataset of sequences, comprising a training reference sequence and a training primary genetic variant, wherein the training primary genetic variant has an effect on the reference sequence with respect to a metric of interest. [Para. 0073] a pre-trained model developed for a task may be used as the starting point for a model on a second task. In some embodiments of the present disclosure, the semantic representation learned from the language model in the pre-training task and/or the fine-tuning task may be transferred to use in the neural network model.)
wherein the fine-tuned model is trained using a transfer learning process by modifying at least one of an underlying weight or structure of the base multispecies transfer learning model based on the pathogen of interest; ([Para. 0073] a pre-trained model developed for a task may be used as the starting point for a model on a second task. In some embodiments of the present disclosure, the semantic representation learned from the language model in the pre-training task and/or the fine-tuning task may be transferred to use in the neural network model.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy and incorporate machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, with the motivation of using machine learning to assess the effects of genetic variants (Altman Para. 0003).
Eltoukhy/ Altman does not explicitly disclose, however Larder discloses
generating, by the one or more processors using the fine-tuned multispecies transfer machine learning model, predicted drug interaction outcome data for one or more drug treatments of interest applied to the pathogen of interest, based on the genetic information of the pathogen of interest; ([Para. 0018] designing a therapeutic agent treatment regimen for a patient afflicted with a disease by predicting resistance of the disease to a therapeutic agent using the determined genetic sequence and the trained neural network)
and indicating, by the one or more processors, the predicted drug interaction outcome data for the one or more drug treatments of interest applied to the pathogen of interest. ([Para. 0018] using the predicted drug resistance to design a therapeutic drug treatment regimen to treat the patient afflicted with the disease. [Para. 0046] accurately prescribe a therapeutic agent or combination of therapeutic agents based upon the pathogen's or malignant cell's existing or developed therapeutic agent resistance, and thereby most effectively treat the patient's disease state.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy, machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, and incorporate the collection and use of the genetic sequence of the pathogen to determine treatment outcomes as taught by Larder, with the motivation of to accurately predict the development of therapeutic agent resistance or sensitivity based upon genotypic and phenotypic information and to accurately define the genetic basis of therapeutic agent resistance in order to design a therapeutic drug treatment regimen to treat the patient afflicted with the disease ( Larder Para. 0013 and Para. 0018).
Eltoukhy/ Altman/ Larder do not explicitly teach, however Wei teaches
generating ortholog predictions by determining, using reciprocal-best- BLAST hits (RBBH), gene orthology between the plurality of pathogens, ([Pg. 2 Introduction] For estimating orthology, we used the reciprocal best hit (RBH) method (i.e. reciprocal-best blast hits), which was widely and effectively applied to map orthologs. [Pg. 3 Results Homology mapping of essential E.coli genes to other organisms] We used a RBH (i.e. reciprocal-best blast hits)method to search for orthologs of essential E. coli genes in 18 bacterial species.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy, machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, the collection and use of the genetic sequence of the pathogen to determine treatment outcomes as taught by Larder, and incorporate the creation and utilization of a prediction tool for sequencing bacterial genomes based on orthology as taught by Wei, with the motivation of predicting bacterial essential genes (Wei Abstract).
Eltoukhy/ Altman/ Larder/ Wei do not explicitly teach, however Chandrasekaran teaches
and transforming omics profiles of each of the drug treatments into binary resistance profiles or response profiles, wherein for chemogenomics data the binary conversion determines deletion strains sensitive or resistant to a drug, and wherein for transcriptomics data the binary conversion determines genes as downregulated or upregulated by the drug ([Pg. 1 Introduction] Present an approach entitled Inferring Drug Interactions using chemo‐Genomics and Orthology (INDIGO), which predicts antibiotic combinations that interact synergistically (i.e. upregulated) or antagonistically (i.e. downregulated) in inhibiting bacterial growth based on the chemogenomic profiles of the individual antibiotics. [Pg. 3 Framework for predicting drug-drug interactions using chemogenomic profiles] The input for INDIGO consists of (i) chemogenomic profiles of individual drugs of interest and (ii) experimental interaction scores for the combinations of drugs. A chemogenomic profile of a drug is an array of fitness scores for gene‐deletion strains treated with the drug of interest compared to the wild‐type strain. We transformed the chemogenomic profile of a drug into a binary sensitivity profile by identifying the deletion strains that are significantly sensitive to a drug. The sensitivity profiles of individual drugs in a drug combination are combined by INDIGO using Boolean operations to create a joint profile.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy, machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, the collection and use of the genetic sequence of the pathogen to determine treatment outcomes as taught by Larder, the creation and utilization of a prediction tool for sequencing bacterial genomes based on orthology as taught by Wei, and incorporate a computational approach called INDIGO, which uses chemogenomics data to predict antibiotic combinations as taught by Chandrasekaran, with the motivation of enables the discovery of effective antibiotic combinations in less‐studied pathogens by leveraging chemogenomics data in model organisms (Chandrasekaran Pg. 1 Abstract).
As per Claim 2, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran disclose the computer-implemented method of claim 1, Eltoukhy further discloses wherein the drug interaction outcome data further includes one or both of genetic information or clinical information of each of a plurality of living subjects and each of the plurality of living subjects has at least one of the plurality of pathogens. ([Para. 0006] receiving into computer memory a training dataset comprising, for each of a plurality of individuals having a disease, 1) genetic information from the individual generated at first time point and (2) treatment response of the individual to one or more therapeutic interventions determined at a second, later, time point. [Para. 0092] used to monitor systemic infections themselves, as may be caused by a pathogen such as a bacteria or virus. Copy number variation or even mutation detection may be used to determine how a population of pathogens are changing during the course of infection. This may be particularly important during chronic infections, such as HIV/AIDs or Hepatitis infections, whereby viruses may change life cycle state and/or mutate into more virulent forms during the course of infection.)
As per Claim 3, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran disclose the computer-implemented method of claim 2, Eltoukhy discloses further comprising:
obtaining, by the one or more processors, one or both of genetic information or clinical information of a living subject of interest having the pathogen of interest; ([Para. 0006] receiving into computer memory a training dataset comprising, for each of a plurality of individuals having a disease, (1) genetic information from the individual generated at first time point. [Para. 0159] data processing method steps can be performed by a programmable processor)
and wherein generating, by the one or more processors using the machine learning model, the predicted drug interaction outcome data for the one or more drug treatments of interest applied to the pathogen of interest, is further based on one or both of the genetic information or the clinical information of the living subject of interest. ([Para. 0006] implementing a machine learning algorithm using the dataset to generate at least one computer implemented classification algorithm, wherein the classification algorithm, based on genetic information from a subject, predicts therapeutic response of the subject to a therapeutic intervention. As used herein, a therapeutic response is a treatment response to a particular therapeutic intervention. . [Para. 0092] used to monitor systemic infections themselves, as may be caused by a pathogen such as a bacteria or virus. Copy number variation or even mutation detection may be used to determine how a population of pathogens are changing during the course of infection. This may be particularly important during chronic infections, such as HIV/AIDs or Hepatitis infections, whereby viruses may change life cycle state and/or mutate into more virulent forms during the course of infection. [Para. 0159] data processing method steps can be performed by a programmable processor)
As per Claim 6, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran disclose the computer-implemented method of claim 1, Eltoukhy further discloses wherein the plurality of pathogens of the drug interaction outcome data include at least two different pathogen strains, and, for each of the at least two different pathogen strains, the drug interaction outcome data includes one or more of chemogenomics data, transcriptomics data or gene orthology data. ([Para. 0006] receiving into computer memory a training dataset comprising, for each of a plurality of individuals having a disease, (1) genetic information from the individual generated at first time point and (2) treatment response of the individual to one or more therapeutic interventions determined at a second, later, time point; and implementing a machine learning algorithm using the dataset to generate at least one computer implemented classification algorithm. Examiner interprets more therapeutic interventions to be indicative of at least two different pathogen strains. [Para. 0017] The genetic information comprises sequence or abundance data from one or more genetic loci in cell-free DNA from the individuals. In some embodiments, the treatment response includes genetic information (i.e. gene orthology data) from the individual generated at a second, later, time point. [Para. 0142] The systems and methods may be used to monitor systemic infections themselves, as may be caused by a pathogen such as a bacteria or virus. Copy number variation or even mutation detection may be used to determine how a population of pathogens are changing during the course of infection. This may be particularly important during chronic infections, such as HIV/AIDs or Hepatitis infections (i.e. plurality of types of diseases/ infections are indicative of strains), whereby viruses may change life cycle state and/or mutate into more virulent forms during the course of infection. [Para. 0159] performed by a programmable processor)
As per Claim 9, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran disclose the computer-implemented method of claim 1, Eltoukhy further discloses wherein the machine learning model is a random forest model. ([Para. 0042] a trained classifier may use a learning algorithm selected from the group consisting of: a random forest. [Para. 0073] a machine learning algorithm is selected from the group consisting of: a supervised or unsupervised learning algorithm selected from support vector machine, random forest)
As per Claim 10, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran disclose the computer-implemented method of claim 1, Eltoukhy discloses further comprising: identifying, by the one or more processors, a recommended drug treatment out of the one or more drug treatments of interest based on the predicted drug interaction outcome data for each of the one or more drug treatments of interest; and indicating, by the one or more processors, the recommended drug treatment. ([Para. 0036] Fig. 1A showcases that the system mines historical cell-free DNA (cfDNA) from a population of cancer subjects or patients (2). The mining is done using genetic data captured from patients undergoing treatment or from healthy people. Once the data mining has been done, the system can recommend treatments based on prior successes and by matching the treatment to the subject/patient genetic characteristics. First, the system obtains subject criteria with genetic characteristics (4). Next, the system identifies similar subjects with similar genetic characteristics (6). The system then identifies successful treatments from these similar subjects (8). Based on prior treatments and outcomes for the similar subjects, the system identifies treatments to be recommended for the current subject (10). [Para. 0037] the system iteratively monitors the treatment process. This is done through subsequent genetic readings (12). Based on the readings, the system identifies the best matching treatment and recommends the treatment based on the success and the subsequent genetic analysis (14).)
As per Claim 12, Eltoukhy discloses a computer system for using transfer machine learning for predicting drug interaction outcomes for pathogens, comprising:
one or more processors; ([Para. 0159] The data processing can be implemented advantageously in one or more computer programs that are executable on a programmable system including at least one programmable processor)
a program memory coupled to the one or more processors and storing executable instructions that, when executed by the one or more processors, cause the computer system to: ([Para. 0159] a processor will receive instructions and data from a read-only memory and/or a random-access memory. )
obtain a base multispecies transfer machine learning model for predicting drug interaction outcomes for pathogens, wherein the base multispecies transfer machine learning model is trained using drug interaction outcome data for a plurality of pathogens ([Para. 0006] receiving into computer memory a training dataset comprising, for each of a plurality of individuals having a disease, (1) genetic information from the individual generated at first time point and (2) treatment response of the individual to one or more therapeutic interventions determined at a second, later, time point; and implementing a machine learning algorithm using the dataset to generate at least one computer implemented classification algorithm. [Para. 0142] used to monitor systemic infections themselves, as may be caused by a pathogen such as a bacteria or virus. Copy number variation or even mutation detection may be used to determine how a population of pathogens are changing during the course of infection. This may be particularly important during chronic infections, such as HIV/AIDs or Hepatitis infections, whereby viruses may change life cycle state and/or mutate into more virulent forms during the course of infection. [Para. 0159] performed by a programmable processor)
wherein the drug interaction outcome data includes, for each respective pathogen of the plurality of pathogens, an outcome of one or more drug treatments applied to the respective pathogen, ([Para. 0006] receiving into computer memory a training dataset comprising, for each of a plurality of individuals having a disease, (1) genetic information from the individual generated at first time point and (2) treatment response of the individual to one or more therapeutic interventions determined at a second, later, time point; and implementing a machine learning algorithm using the dataset to generate at least one computer implemented classification algorithm. [Para. 0142] used to monitor systemic infections themselves, as may be caused by a pathogen such as a bacteria or virus. Copy number variation or even mutation detection may be used to determine how a population of pathogens are changing during the course of infection. This may be particularly important during chronic infections, such as HIV/AIDs or Hepatitis infections, whereby viruses may change life cycle state and/or mutate into more virulent forms during the course of infection. [Para. 0159] performed by a programmable processor)
Eltoukhy does not explicitly disclose, however Altman discloses
wherein the base multispecies transfer machine learning model is used to train a fine-tuned multispecies transfer machine learning model for predicting drug interaction outcome data based on genetic information of a pathogen of interest, ([Para. 0009] the model is trained by: a fine-tuning task, comprising: 1) receiving a fine-tuning dataset comprising a plurality of batches of naturally occurring sequences, wherein the fine-tuning dataset is a subset of the pre-training dataset, or a set of sequences that are related to the pre-training dataset by common ancestry, homology, or multiple sequence alignment; 2) inputting each batch of sequences into the language model, wherein the model is configured to output a fine-tuning set of semantic features.)
and wherein the base multispecies transfer machine learning model is trained by (i) predicting at least one classification of outcomes for drug treatments applied to one or more pathogens of interest and (ii) generating a matrix including at least one of joint sensitivity or resistance profiles corresponding to at least one combination of the drug treatment based on omics profiles of each of the drug treatments in combination and wherein generating the matrix includes:; ([Para. 0009] The model is trained by a transfer learning task, comprising: 1) receiving a final training dataset comprising labeled sequences mapped to effects; and 2) training a neural network model based on the final training dataset, wherein the neural network model is configured to receive data corresponding to the pre-training set of semantic features and/or the fine-tuning set of semantic features, and output one or more effect scores. [Para. 0071] The term “semantic feature” refers to a representation of how the elements relate to or connect with each other in the input sequence data. In some embodiments, the representation is mathematical or numerical. In some embodiments, the semantic features may be a human and/or machine interpretable representation of the state of the input sequence. The output semantic features may be presented in a vector or a matrix (i.e. generating a matrix), and may be used as input for a downstream task, such as in transfer learning. [Para. 0010] the model is trained by: a) receiving a training dataset of sequences, comprising a training reference sequence and a training primary genetic variant, wherein the training primary genetic variant has an effect on the reference sequence with respect to a metric of interest; b) inputting the training dataset into a generative procedure configured to generate one or more training secondary genetic variants according to a random seed; c) calculating a loss function, wherein the loss function maps the combined effect of the primary and secondary genetic variants and the effect of the reference sequence onto a quantitative error score; d) accepting or rejecting the one or more training secondary genetic variants according to one or more predetermined acceptance criteria on the loss function; e) updating the generative procedure by incorporating the accepted one or more training secondary genetic variants in a new round of additional training secondary genetic variants; and f) repeating steps b) to e) until the loss converges to a minimum.)
obtain (i)genetic information of a pathogen of interest not in the plurality of pathogens and (ii) the fine-tuned multispecies transfer machine learning model, ([Para. 0010] The model is trained by: a) receiving a training dataset of sequences, comprising a training reference sequence and a training primary genetic variant, wherein the training primary genetic variant has an effect on the reference sequence with respect to a metric of interest. [Para. 0073] a pre-trained model developed for a task may be used as the starting point for a model on a second task. In some embodiments of the present disclosure, the semantic representation learned from the language model in the pre-training task and/or the fine-tuning task may be transferred to use in the neural network model.)
wherein the fine-tuned multispecies transfer machine learning model is trained using a transfer learning process by modifying at least one of an underlying weight or structure of the base multispecies transfer learning model based on the pathogen of interest; ([Para. 0073] a pre-trained model developed for a task may be used as the starting point for a model on a second task. In some embodiments of the present disclosure, the semantic representation learned from the language model in the pre-training task and/or the fine-tuning task may be transferred to use in the neural network model.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy and incorporate machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, with the motivation of using machine learning to assess the effects of genetic variants (Altman Para. 0003).
Eltoukhy/ Altman do not explicitly disclose, however Larder discloses
generate, using the fine-tuned multispecies transfer machine learning model, predicted drug interaction outcome data for one or more drug treatments of interest applied to the pathogen of interest, based on the genetic information of the pathogen of interest; ([Para. 0018] designing a therapeutic agent treatment regimen for a patient afflicted with a disease by predicting resistance of the disease to a therapeutic agent using the determined genetic sequence and the trained neural network)
and indicate the predicted drug interaction outcome data for the one or more drug treatments of interest applied to the pathogen of interest. ([Para. 0018] using the predicted drug resistance to design a therapeutic drug treatment regimen to treat the patient afflicted with the disease. [Para. 0046] accurately prescribe a therapeutic agent or combination of therapeutic agents based upon the pathogen's or malignant cell's existing or developed therapeutic agent resistance, and thereby most effectively treat the patient's disease state.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy, machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, and incorporate the collection and use of the genetic sequence of the pathogen to determine treatment outcomes as taught by Larder, with the motivation of to accurately predict the development of therapeutic agent resistance or sensitivity based upon genotypic and phenotypic information and to accurately define the genetic basis of therapeutic agent resistance in order to design a therapeutic drug treatment regimen to treat the patient afflicted with the disease ( Larder Para. 0013 and Para. 0018).
Eltoukhy/ Altman/ Larder do not explicitly teach, however Wei teaches
generating ortholog predictions by determining, using reciprocal-best- BLAST hits (RBBH), gene orthology between the plurality of pathogens,([Pg. 2 Introduction] For estimating orthology, we used the reciprocal best hit (RBH) method (i.e. reciprocal-best blast hits), which was widely and effectively applied to map orthologs. [Pg. 3 Results Homology mapping of essential E.coli genes to other organisms] We used a RBH (i.e. reciprocal-best blast hits)method to search for orthologs of essential E. coli genes in 18 bacterial species.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy, machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, the collection and use of the genetic sequence of the pathogen to determine treatment outcomes as taught by Larder, and incorporate the creation and utilization of a prediction tool for sequencing bacterial genomes based on orthology as taught by Wei, with the motivation of predicting bacterial essential genes (Wei Abstract).
Eltoukhy/ Altman/ Larder/ Wei do not explicitly teach, however Chandrasekaran teaches
and transforming omics profiles of each of the drug treatments into binary resistance profiles or response profiles, wherein for chemogenomics data the binary conversion determines deletion strains sensitive or resistant to a drug, and wherein for transcriptomics data the binary conversion determines genes as downregulated or upregulated by the drug ([Pg. 1 Introduction] Present an approach entitled Inferring Drug Interactions using chemo‐Genomics and Orthology (INDIGO), which predicts antibiotic combinations that interact synergistically (i.e. upregulated) or antagonistically (i.e. downregulated) in inhibiting bacterial growth based on the chemogenomic profiles of the individual antibiotics. [Pg. 3 Framework for predicting drug-drug interactions using chemogenomic profiles] The input for INDIGO consists of (i) chemogenomic profiles of individual drugs of interest and (ii) experimental interaction scores for the combinations of drugs. A chemogenomic profile of a drug is an array of fitness scores for gene‐deletion strains treated with the drug of interest compared to the wild‐type strain. We transformed the chemogenomic profile of a drug into a binary sensitivity profile by identifying the deletion strains that are significantly sensitive to a drug. The sensitivity profiles of individual drugs in a drug combination are combined by INDIGO using Boolean operations to create a joint profile.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy, machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, the collection and use of the genetic sequence of the pathogen to determine treatment outcomes as taught by Larder, the creation and utilization of a prediction tool for sequencing bacterial genomes based on orthology as taught by Wei, and incorporate a computational approach called INDIGO, which uses chemogenomics data to predict antibiotic combinations as taught by Chandrasekaran, with the motivation of enables the discovery of effective antibiotic combinations in less‐studied pathogens by leveraging chemogenomics data in model organisms (Chandrasekaran Pg. 1 Abstract).
As per Claim 13, Claim(s) 13 is/are analogous to Claim(s) 2, thus Claim(s) 13 is/are similarly analyzed and rejected in a manner consistent with the rejection of Claim(s) 2.
As per Claim 14, Claim(s) 14 is/are analogous to Claim(s) 3, thus Claim(s) 14 is/are similarly analyzed and rejected in a manner consistent with the rejection of Claim(s) 3.
As per Claim 16, Claim(s) 16 is/are analogous to Claim(s) 6, thus Claim(s) 16 is/are similarly analyzed and rejected in a manner consistent with the rejection of Claim(s) 6.
As per Claim 17, Eltoukhy teaches a tangible, non-transitory computer-readable medium storing executable instructions for using transfer machine learning for predicting drug interaction outcomes for pathogens, when executed by one or more processors of a computer system, cause the computer system to:
obtain a base multispecies transfer machine learning model for predicting drug interaction outcomes for pathogens, wherein the base multispecies transfer machine learning model is trained using drug interaction outcome data for a plurality of pathogens or model organisms,([Para. 0006] receiving into computer memory a training dataset comprising, for each of a plurality of individuals having a disease, (1) genetic information from the individual generated at first time point and (2) treatment response of the individual to one or more therapeutic interventions determined at a second, later, time point; and implementing a machine learning algorithm using the dataset to generate at least one computer implemented classification algorithm. [Para. 0142] used to monitor systemic infections themselves, as may be caused by a pathogen such as a bacteria or virus. Copy number variation or even mutation detection may be used to determine how a population of pathogens are changing during the course of infection. This may be particularly important during chronic infections, such as HIV/AIDs or Hepatitis infections, whereby viruses may change life cycle state and/or mutate into more virulent forms during the course of infection. [Para. 0159] performed by a programmable processor)
Eltoukhy does not explicitly teach, however Altman teaches
wherein the base multispecies transfer machine learning model is used to train a fine-tuned multispecies transfer machine learning model for predicting drug interaction outcome data based on genetic information of a pathogen of interest, ([Para. 0009] the model is trained by: a fine-tuning task, comprising: 1) receiving a fine-tuning dataset comprising a plurality of batches of naturally occurring sequences, wherein the fine-tuning dataset is a subset of the pre-training dataset, or a set of sequences that are related to the pre-training dataset by common ancestry, homology, or multiple sequence alignment; 2) inputting each batch of sequences into the language model, wherein the model is configured to output a fine-tuning set of semantic features.)
and wherein the base multispecies transfer machine learning model is trained by (i) predicting at least one classification of outcomes for drug treatments applied to one or more pathogens of interest and (ii) generating a matrix including at least one of joint sensitivity or resistance profiles corresponding to at least one combination of the drug treatment based on omics profiles of each of the drug treatments in combination and wherein generating the matrix includes:; ([Para. 0009] The model is trained by a transfer learning task, comprising: 1) receiving a final training dataset comprising labeled sequences mapped to effects; and 2) training a neural network model based on the final training dataset, wherein the neural network model is configured to receive data corresponding to the pre-training set of semantic features and/or the fine-tuning set of semantic features, and output one or more effect scores. [Para. 0071] The term “semantic feature” refers to a representation of how the elements relate to or connect with each other in the input sequence data. In some embodiments, the representation is mathematical or numerical. In some embodiments, the semantic features may be a human and/or machine interpretable representation of the state of the input sequence. The output semantic features may be presented in a vector or a matrix (i.e. generating a matrix), and may be used as input for a downstream task, such as in transfer learning. [Para. 0010] the model is trained by: a) receiving a training dataset of sequences, comprising a training reference sequence and a training primary genetic variant, wherein the training primary genetic variant has an effect on the reference sequence with respect to a metric of interest; b) inputting the training dataset into a generative procedure configured to generate one or more training secondary genetic variants according to a random seed; c) calculating a loss function, wherein the loss function maps the combined effect of the primary and secondary genetic variants and the effect of the reference sequence onto a quantitative error score; d) accepting or rejecting the one or more training secondary genetic variants according to one or more predetermined acceptance criteria on the loss function; e) updating the generative procedure by incorporating the accepted one or more training secondary genetic variants in a new round of additional training secondary genetic variants; and f) repeating steps b) to e) until the loss converges to a minimum.)
obtain (i)genetic information of a pathogen of interest not in the plurality of pathogens and (ii) the fine-tuned multispecies transfer machine learning model, ([Para. 0010] The model is trained by: a) receiving a training dataset of sequences, comprising a training reference sequence and a training primary genetic variant, wherein the training primary genetic variant has an effect on the reference sequence with respect to a metric of interest. [Para. 0073] a pre-trained model developed for a task may be used as the starting point for a model on a second task. In some embodiments of the present disclosure, the semantic representation learned from the language model in the pre-training task and/or the fine-tuning task may be transferred to use in the neural network model.)
wherein the fine-tuned multispecies transfer machine learning model is trained using a transfer learning process by modifying at least one of an underlying weight or structure of the base multispecies transfer learning model based on the pathogen of interest; ([Para. 0073] a pre-trained model developed for a task may be used as the starting point for a model on a second task. In some embodiments of the present disclosure, the semantic representation learned from the language model in the pre-training task and/or the fine-tuning task may be transferred to use in the neural network model.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy and incorporate machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, with the motivation of using machine learning to assess the effects of genetic variants (Altman Para. 0003).
Eltoukhy/ Altman do not explicitly disclose, however Larder discloses
generate, using the fine-tuned multispecies transfer machine learning model, predicted drug interaction outcome data for one or more drug treatments of interest applied to the pathogen of interest, based on the genetic information of the pathogen of interest; ([Para. 0018] designing a therapeutic agent treatment regimen for a patient afflicted with a disease by predicting resistance of the disease to a therapeutic agent using the determined genetic sequence and the trained neural network) Larder
and indicate the predicted drug interaction outcome data for the one or more drug treatments of interest applied to the pathogen of interest. ([Para. 0018] using the predicted drug resistance to design a therapeutic drug treatment regimen to treat the patient afflicted with the disease. [Para. 0046] accurately prescribe a therapeutic agent or combination of therapeutic agents based upon the pathogen's or malignant cell's existing or developed therapeutic agent resistance, and thereby most effectively treat the patient's disease state.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy, machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, and incorporate the collection and use of the genetic sequence of the pathogen to determine treatment outcomes as taught by Larder, with the motivation of to accurately predict the development of therapeutic agent resistance or sensitivity based upon genotypic and phenotypic information and to accurately define the genetic basis of therapeutic agent resistance in order to design a therapeutic drug treatment regimen to treat the patient afflicted with the disease ( Larder Para. 0013 and Para. 0018).
Eltoukhy/ Altman/ Larder do not explicitly teach, however Wei teaches
generating ortholog predictions by determining, using reciprocal-best- BLAST hits (RBBH), gene orthology between the plurality of pathogens, ([Pg. 2 Introduction] For estimating orthology, we used the reciprocal best hit (RBH) method (i.e. reciprocal-best blast hits), which was widely and effectively applied to map orthologs. [Pg. 3 Results Homology mapping of essential E.coli genes to other organisms] We used a RBH (i.e. reciprocal-best blast hits)method to search for orthologs of essential E. coli genes in 18 bacterial species.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy, machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, the collection and use of the genetic sequence of the pathogen to determine treatment outcomes as taught by Larder, and incorporate the creation and utilization of a prediction tool for sequencing bacterial genomes based on orthology as taught by Wei, with the motivation of predicting bacterial essential genes (Wei Abstract).
Eltoukhy/ Altman/ Larder/ Wei do not explicitly teach, however Chandrasekaran teaches
and transforming omics profiles of each of the drug treatments into binary resistance profiles or response profiles, wherein for chemogenomics data the binary conversion determines deletion strains sensitive or resistant to a drug, and wherein for transcriptomics data the binary conversion determines genes as downregulated or upregulated by the drug ([Pg. 1 Introduction] Present an approach entitled Inferring Drug Interactions using chemo‐Genomics and Orthology (INDIGO), which predicts antibiotic combinations that interact synergistically (i.e. upregulated) or antagonistically (i.e. downregulated) in inhibiting bacterial growth based on the chemogenomic profiles of the individual antibiotics. [Pg. 3 Framework for predicting drug-drug interactions using chemogenomic profiles] The input for INDIGO consists of (i) chemogenomic profiles of individual drugs of interest and (ii) experimental interaction scores for the combinations of drugs. A chemogenomic profile of a drug is an array of fitness scores for gene‐deletion strains treated with the drug of interest compared to the wild‐type strain. We transformed the chemogenomic profile of a drug into a binary sensitivity profile by identifying the deletion strains that are significantly sensitive to a drug. The sensitivity profiles of individual drugs in a drug combination are combined by INDIGO using Boolean operations to create a joint profile.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy, machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, the collection and use of the genetic sequence of the pathogen to determine treatment outcomes as taught by Larder, the creation and utilization of a prediction tool for sequencing bacterial genomes based on orthology as taught by Wei, and incorporate a computational approach called INDIGO, which uses chemogenomics data to predict antibiotic combinations as taught by Chandrasekaran, with the motivation of enables the discovery of effective antibiotic combinations in less‐studied pathogens by leveraging chemogenomics data in model organisms (Chandrasekaran Pg. 1 Abstract).
As per Claim 18, Claim(s) 18 is/are analogous to Claim(s) 2, thus Claim(s) 18 is/are similarly analyzed and rejected in a manner consistent with the rejection of Claim(s) 2.
As per Claim 19, Claim(s) 19 is/are analogous to Claim(s) 3, thus Claim(s) 19 is/are similarly analyzed and rejected in a manner consistent with the rejection of Claim(s) 3.
Claim(s) 4-5, 15, and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Eltoukhy (US 20200395100 A1) in view of Altman (US 20230402127 A1) in view of Larder (US 20030190603 A1) in view of Wei (Geptop: A Gene Essentiality Prediction Tool for Sequenced Bacterial Genomes Based on Orthology and Phylogeny (2013)) in view of Chandrasekaran (Chemogenomics and orthology‐based design of antibiotic combination therapies (2016)) in view of White (US 20240127967 A1).
As per Claim 4, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran discloses the computer-implemented method of claim 1, however Eltoukhy/ Altman/ Larder/ Wei Chandrasekaran do not explicitly disclose, however White discloses wherein one or both of:
(i) each drug treatment of the one or more drug treatments included in the drug interaction outcome data includes a plurality of individual drugs, and, each respective outcome of the one or more drug treatments includes an indication of a measure of synergistic interaction of the plurality of individual drugs or a measure of antagonistic interaction of the plurality of individual drugs, or (ii) each drug treatment of interest of the one or more drug treatments of interest includes a plurality of individual drugs of interest, and, each respective outcome of the one or more drug treatments of interest includes an indication of a measure of synergistic interaction of the plurality of individual drugs of interest or a measure of antagonistic interaction of the plurality of individual drugs of interest. ([Para. 0070] High throughput multi-dose matrix assays are used to screen drug combinations with antiviral activity against Zika, arenaviruses, Ebola, and SARS-CoV-2. The results of those assays are reported in dose-response matrixes with efficacy of each combination reported within each matrix locus. To screen combinations based on the dose-response matrixes, there are available tools, such as SynergyFinder 2.0 and MacSynergy II, which calculate synergy scores based on deviation in observed combined effects from expected efficacy of a drug combination given by different reference models including Highest Single Agent (HSA), Loewe additivity, Bliss independence and Zero interaction potency (ZIP). Based on which reference model best explains the data, drug-drug interactions can be classified as antagonistic, Loewe additive, intermediate, Bliss independent or synergistic, among which synergy has the highest increase in combined efficacy, exceeding that of multiplicative effects. The quantified synergy score demonstrates the strength of drug-drug interaction and measured biological effect.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the effective filing date, to modify the method of predicting drug interaction outcomes for pathogens as taught by Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran and incorporate wherein one or both of: (i) each drug treatment of the one or more drug treatments included in the drug interaction outcome data includes a plurality of individual drugs, and, each respective outcome of the one or more drug treatments includes an indication of a measure of synergistic interaction of the plurality of individual drugs or a measure of antagonistic interaction of the plurality of individual drugs, or (ii) each drug treatment of interest of the one or more drug treatments of interest includes a plurality of individual drugs of interest, and, each respective outcome of the one or more drug treatments of interest includes an indication of a measure of synergistic interaction of the plurality of individual drugs of interest or a measure of antagonistic interaction of the plurality of individual drugs of interest as taught by White, with the motivation of providing predictions can be used to more accurately select drugs and drug treatment regimens that can be successful in controlling viral infection in animal studies, clinical trials and in medical or veterinary interventions (White Abstract).
As per Claim 5, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran/ White discloses the computer-implemented method of claim 4, White further discloses wherein either the measure of synergistic interaction or the measure of antagonistic interaction, for one or both of (i) each of the individual drugs of the drug treatments, (ii) or each of the individual drugs of interest of the drug treatments of interest, includes one or more scores which are determined using one or both of the Loewe Additivity model or the Bliss Independence model. ([Para. 0070] High throughput multi-dose matrix assays are used to screen drug combinations with antiviral activity against Zika, arenaviruses, Ebola, and SARS-CoV-2. The results of those assays are reported in dose-response matrixes with efficacy of each combination reported within each matrix locus. To screen combinations based on the dose-response matrixes, there are available tools, such as SynergyFinder 2.0 and MacSynergy II, which calculate synergy scores based on deviation in observed combined effects from expected efficacy of a drug combination given by different reference models including Highest Single Agent (HSA), Loewe additivity, Bliss independence and Zero interaction potency (ZIP). Based on which reference model best explains the data, drug-drug interactions can be classified as antagonistic, Loewe additive, intermediate, Bliss independent or synergistic, among which synergy has the highest increase in combined efficacy, exceeding that of multiplicative effects. The quantified synergy score demonstrates the strength of drug-drug interaction and measured biological effect.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the effective filing date, to modify the method of predicting drug interaction outcomes for pathogens as taught by Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran and incorporate wherein either the measure of synergistic interaction or the measure of antagonistic interaction, using one or both of the Loewe Additivity model or the Bliss Independence model taught by White, with the motivation of providing predictions can be used to more accurately select drugs and drug treatment regimens that can be successful in controlling viral infection in animal studies, clinical trials and in medical or veterinary interventions (White Abstract).
As per Claim 15, Claim(s) 15 is/are analogous to Claim(s) 4, thus Claim(s) 15 is/are similarly analyzed and rejected in a manner consistent with the rejection of Claim(s) 4.
As per Claim 20, Claim(s) 20 is/are analogous to Claim(s) 4, thus Claim(s) 20 is/are similarly analyzed and rejected in a manner consistent with the rejection of Claim(s) 4.
Claim(s) 7 and 8 is/are rejected under 35 U.S.C. 103 as being unpatentable over Eltoukhy (US 20200395100 A1) in view of Altman (US 20230402127 A1) in view of Larder (US 20030190603 A1) in view of Wei (Geptop: A Gene Essentiality Prediction Tool for Sequenced Bacterial Genomes Based on Orthology and Phylogeny (2013)) in view of Chandrasekaran (Chemogenomics and orthology‐based design of antibiotic combination therapies (2016)) in view of Bobrowski (“Discovery of Synergistic and Antagonistic Drug Combinations against SARS-CoV-2 In Vitro”)
REGARDING CLAIM 7
As per Claim 7, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran disclose the computer-implemented method of claim 1, Eltoukhy further discloses, by the one or more processors ([Para. 0159] performed by a programmable processor)
Eltoukhy does not explicitly disclose, however Bobrowski discloses
further comprising: obtaining, drug information for the plurality of individual drugs of interest, wherein each of the one or more drug treatments of interest includes two or more individual drugs of interest of the plurality of individual drugs of interest. ([Pg. 5 Introduction] First identified a list of 76 individual drug candidates for repurposing in combination therapy against COVID-19 using a combination of text mining, knowledge mining, and machine learning. Here we selected some of the hits from virtual screening of DrugBank33 and NPC34 collections by our QSAR models of SARS-CoV Mpro inhibition35 and all the drugs found in Chemotext and ROBOKOP searching for “SARS”, “Coronaviridae”, etc. This resulted in 76 individual drugs that may potentially create 2580 binary and 70300 ternary combinations (i.e. two or more individual drugs of interest).)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the effective filing date, to modify the method of predicting drug interaction outcomes for pathogens as taught by Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran and incorporate obtaining, drug information for the plurality of individual drugs of interest, wherein each of the one or more drug treatments of interest includes two or more individual drugs of interest of the plurality of individual drugs of interest taught by Bobrowski, with the motivation of emphasizing the importance of drug repurposing and preclinical testing of drug combinations for potential therapeutic use against SARS-CoV-2 infections (Bobrowski Pg. 2 Abstract).
As per Claim 8, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran disclose the computer-implemented method of claim 1, Eltoukhy/ Larder/ Wei/ Chandrasekaran do not explicitly disclose, however Bobrowski discloses wherein one or both of: (i) the pathogen of interest is not one of the plurality of pathogens of the drug interaction outcome data, or (ii) at least one of the one or more drug treatments of interest is not one of the one or more drug treatments of the drug interaction outcome data. ([Pg. 4 Introduction] While there are many ongoing or upcoming clinical trials testing combinations to treat COVID-19, few have undergone extensive preclinical studies prior to their combination in patients. Due to a lack of such studies, more information is needed on the combinatorial use of antivirals and other drugs against SARS-CoV-2 in order to (1) more efficiently prioritize synergistic combinations for translation into clinical use; and (2) flag antagonistic combinations prior to their evaluation in the preclinical stage. To this point, we have recently used data and text mining approaches to propose drug combinations for repurposing against SARS-CoV-2, operating on the assumption that combinations of drugs with differing mechanisms might exhibit synergistic activity. )
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the effective filing date, to modify the method of predicting drug interaction outcomes for pathogens as taught by Eltoukhy/ Altman/ Larder / Wei/ Chandrasekaran and incorporate wherein one or both of: (i) the pathogen of interest is not one of the plurality of pathogens of the drug interaction outcome data, or (ii) at least one of the one or more drug treatments of interest is not one of the one or more drug treatments of the drug interaction outcome data taught by Bobrowski, with the motivation of emphasizing the importance of drug repurposing and preclinical testing of drug combinations for potential therapeutic use against SARS-CoV-2 infections (Bobrowski Pg. 2 Abstract).
Claim(s) 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Eltoukhy (US 20200395100 A1) in view of Altman (US 20230402127 A1) in view of Larder (US 20030190603 A1) in view of Wei (Geptop: A Gene Essentiality Prediction Tool for Sequenced Bacterial Genomes Based on Orthology and Phylogeny (2013)) in view of Chandrasekaran (Chemogenomics and orthology‐based design of antibiotic combination therapies (2016)) in view of Ambrose (US 11227680 B2).
As per Claim 11, Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran disclose the computer-implemented method of claim 1, Eltoukhy/ Larder/ Wei/ Chandrasekaran do not explicitly disclose, Ambrose discloses further comprising: analyzing, by the one or more processors, feedback from a user following the indication of the predicted drug interaction outcome data for the one or more drug treatments of interest applied to the pathogen of interest, the feedback regarding the predicted drug interaction outcome data and including user input; and updating, by the one or more processors, the machine learning model based on the feedback from the user. ([Col. 2, Lines 25- 33] obtaining, by the one or more processors, a third indication comprising designation of a drug therapy from the one or more drug therapies displayed; retaining by the one or more processors, the designation on a memory device; prompting, by the one or more processors, through a user interface, a user to provide data indicating an actual efficacy of the drug therapy as utilized by the patient with the infection at one or more predetermined intervals after obtaining the designation. [Col. 15, Lines 19-30] Once the user has selected the drug therapy from the drug therapies returned as options, the user can select whether he or she would like to be prompted to follow up with the patient. Should the user opt to follow up, in an embodiment of the present invention, the program code will display a reminder to the user to follow up regarding a given patient. FIG. 18 is an example of a possible display for this follow up activity and additionally may collect information regarding efficacy, including but not limited to, requesting that a user enter information and/or importing information from an external data repository. Examiner interprets information collected through follow-up to be indicative of user input. [Col. 21, Lines 60-63] he program code combines the obtained additional indications with one or more of patient demographic data, clinical data, and laboratory data, to generate or update a base model (2620).)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, at the time of filing, to modify the method of predicting drug interaction outcomes for pathogens as taught by Eltoukhy/ Altman/ Larder/ Wei/ Chandrasekaran and incorporate analyzing, by the one or more processors, feedback from a user following the indication of the predicted drug interaction outcome data for the one or more drug treatments of interest applied to the pathogen of interest, the feedback regarding the predicted drug interaction outcome data and including user input; and updating, by the one or more processors, the machine learning model based on the feedback from the user as taught by Ambrose, with the motivation of selecting therapies that optimize the probability of positive outcomes for patients suffering from an infection (Ambrose Col. 1, Lines 23-25).
Claim(s) 21 and 22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Eltoukhy (US 20200395100 A1) in view of Altman (US 20230402127 A1).
As per Claim 21, Eltoukhy discloses a computer-implemented method for training a statistical model to predict drug interaction outcomes for pathogens, comprising:
obtaining, by one or more processors, a set of training data for a plurality of pathogens including actual outcomes of one or more drug treatments applied to each of the plurality of pathogens; ([Para. 0006] receiving into computer memory a training dataset comprising, for each of a plurality of individuals having a disease, (1) genetic information from the individual generated at first time point and (2) treatment response of the individual to one or more therapeutic interventions determined at a second, later, time point)
classifying, by the one or more processors, the set of training data into a plurality of subsets each corresponding to a different actual outcome or a range of actual outcomes;([Para. 0040] provides a way to classify treatment responses to therapeutic interventions, and subsequently determine whether a given individual falls into a particular classification (e.g., responsive to treatment, nonresponsive to treatment, or a particular level of responsiveness such as fully responsive or partially responsive). [Para. 0041] creating a trained classifier, comprising the steps of: (a) providing a plurality of different classes, wherein each class represents a set of subjects with a shared characteristic (e.g. from one or more cohorts); (b) providing a multi-parametric model representative of the cell-free DNA molecules from each of a plurality of samples belonging to each of the classes, thereby providing a training data set; and (c) training a learning algorithm on the training data set to create one or more trained classifiers, wherein each trained classifier classifies a test sample into one or more of the plurality of classes.)
and generating, by the one or more processors, the statistical model for predicting an outcome of applying a drug treatment of interest to a pathogen of interest using the classified subsets of training data.([Para. 0006] implementing a machine learning algorithm using the dataset to generate at least one computer implemented classification algorithm, wherein the classification algorithm, based on genetic information from a subject, predicts therapeutic response of the subject to a therapeutic intervention.)
Eltoukhy does not explicitly disclose, however Altman discloses
wherein the statistical model is a base multispecies transfer machine learning model trained by (i) predicting at least one classification of outcomes for drug treatments applied to one or more pathogens of interest and (ii) generating a matrix including at least one of joint sensitivity or resistance profiles corresponding to at least one combination of the drug treatment based on omics profiles of each of the drug treatments in combination. ([Para. 0010] the model is trained by: a) receiving a training dataset of sequences, comprising a training reference sequence and a training primary genetic variant, wherein the training primary genetic variant has an effect on the reference sequence with respect to a metric of interest; b) inputting the training dataset into a generative procedure configured to generate one or more training secondary genetic variants according to a random seed; c) calculating a loss function, wherein the loss function maps the combined effect of the primary and secondary genetic variants and the effect of the reference sequence onto a quantitative error score; d) accepting or rejecting the one or more training secondary genetic variants according to one or more predetermined acceptance criteria on the loss function; e) updating the generative procedure by incorporating the accepted one or more training secondary genetic variants in a new round of additional training secondary genetic variants; and f) repeating steps b) to e) until the loss converges to a minimum.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy and incorporate machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, with the motivation of using machine learning to assess the effects of genetic variants (Altman Para. 0003).
As per Claim 22, Eltoukhy/ Altman disclose the computer-implemented method of Claim 21, Altman further discloses the computer-implemented method of wherein the base multispecies transfer machine learning model is used to train a fine-tuned multispecies transfer machine learning model for predicting drug interaction outcome data based on genetic information of the pathogen of interest, ([Para. 0009] the model is trained by: a fine-tuning task, comprising: 1) receiving a fine-tuning dataset comprising a plurality of batches of naturally occurring sequences, wherein the fine-tuning dataset is a subset of the pre-training dataset, or a set of sequences that are related to the pre-training dataset by common ancestry, homology, or multiple sequence alignment; 2) inputting each batch of sequences into the language model, wherein the model is configured to output a fine-tuning set of semantic features.)
and wherein the fine-tuned multispecies transfer machine learning model is trained using a transfer learning process by modifying at least one of an underlying weight or structure of the base multispecies transfer learning model based on the pathogen of interest. ([Para. 0073] a pre-trained model developed for a task may be used as the starting point for a model on a second task. In some embodiments of the present disclosure, the semantic representation learned from the language model in the pre-training task and/or the fine-tuning task may be transferred to use in the neural network model.)
Therefore, it would be prima facie obvious to one of ordinary skill in the art, before the time of filing, to modify the neural network being trained specifically for predicting treatment outcomes as taught by Eltoukhy and incorporate machine learning-based methods for assessing the combined impact of multiple genetic variants as taught by Altman, with the motivation of using machine learning to assess the effects of genetic variants (Altman Para. 0003).
Response to Arguments
Applicant's arguments, see pgs. 12-13, filed 01/06/2026 have been fully considered are persuasive regarding the newly added limitations. Therefore, the rejection has been withdrawn. However, upon further consideration, a new grounds of rejection is made in view of Wei and Chandrasekaran, as per the rejection above.
Applicant's arguments, see pg. 13, filed 01/06/2026 have been fully considered but they are not persuasive.
Applicant submits that the amended claim 1 is integrated into a practical application because the specific technical steps of RBBH gene orthology determination and binary conversion of omics data are particular implementations that improve cross-species drug interaction prediction, not merely generic computer implementation of an abstract idea. Examiner respectfully disagrees. As Examiner has identified the binary conversion is part of the abstract idea, it is not considered under Step 2A, Prong One. The use of RBBH gene orthology determination amounts to a tool to apply data to an algorithm and report the results (MPEP § 2106.05(f)(2) see case involving a commonplace business method or mathematical algorithm being applied on a general purpose computer within the “Other examples.. i.”) amounting to instruction to implement the abstract idea using a general purpose computer. Alice Corp. Pty. Ltd. V. CLS Bank Int’l, 134 S. Ct. 2347, 1357 (2014).
An improvement to the abstract idea of cross-species drug interaction prediction does not amount to an improvement to technology or a technical field (see MPEP § 2106.05(a)(III) stating “it is important to keep in mind that an improvement in the abstract idea itself (e.g. a recited fundamental economic concept) is not an improvement in technology. For example, in Trading Technologies Int’l v. IBG,921 F.3d 1084, 1093-94, 2019 USPQ2d 138290 (Fed. Cir. 2019), the court determined that the claimed user interface simply provided a trader with more information to facilitate market trades, which improved the business process of market trading but did not improve computers or technology.”). There is no indication in the instant disclosure that the involvement of a computer assists in improving the technology for the outlined problem statement. Here, the improvement is to drug interaction prediction. The instant application and claim language fail to detail how a computer aids the method, the extent to which the computer aids the method, or the significance of a computer to the performance of the method. Merely adding generic computer components to perform the method is not sufficient.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Patricia K Edouard whose telephone number is (571)272-6084. The examiner can normally be reached Monday - Friday 7:30 AM - 5:00 PM.
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, Peter H Choi can be reached at 469-295-9171. 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.
/P.K.E./Examiner, Art Unit 3681
/PETER H CHOI/Supervisory Patent Examiner, Art Unit 3681