CTNF 18/306,610 CTNF 79126 DETAILED ACTION Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA. In response to the restriction requirement, Applicant elected claims 5-15 for further examination. As a result, claims 1-4 are withdrawn from further prosecution. Claim Rejections - 35 USC § 103 07-20-aia AIA 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. 07-21-aia AIA Claim (s) 5-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Lanza et al. (WO 2006/065950) in view of Roy (EP 3968333) and Jadon (US 2023/0031889) . Regarding to claims 5 and 13: Lanza et al. discloses a computer-implemented method of training a chemical super learner model, comprising: obtaining a plurality of training examples ( FIG. 2, elements 225: Training examples (pagraph [0040]) ) from a training database ( FIG. 3: Training data 220 ) wherein each training example comprises: a molecule ( FIG. 2, elements MOL_A, MOL_B, MOL_C ), and a first property ( paragraph [0021]: The training examples include a description for a molecule and data regarding a property of interest for the molecule ); processing the plurality of training examples, with a molecular descriptor generator to produce a plurality of molecular descriptors; pre-processing, with a pre-processor, the plurality of molecular descriptors ( FIG. 5, step 510: Generate representation of molecule for single target models ); training one or more machine-learned models ( FIG. 3: The plurality of single target activity models 205 ) using the pre-processed plurality of molecular descriptors and the training database ( paragraph [0023]: The training set is used to train a molecule properties model ), wherein each of the one or more machine-learned models are configured to accept a pre-processed molecular descriptor and return a first property prediction ( FIG. 5, step 515: Obtain prediction for single target activity models ); scoring the one or more machine learned models, wherein upon scoring each of the one or more machine-learned models has a score; selecting a subset of the one or more machine learned models, wherein each of the machine-learned models in the subset has a better score than the machine-learned models outside of the subset ( FIG. 4, step 405: Select single target activity models. Paragraph [0048]: The selection is based on the measured (or predicted) accuracy of the models, wherein such accuracy reads on the claimed score ); and forming the trained chemical super learner model ( FIG. 3, element 310: Metal Model ). Lanza et al. however does not teach wherein the molecule description is a simplified molecular-input line-entry system (SMILES) description, and the comprising of tuning hyperparameters of each of the machine-learned models in the subset; determining a weight for each machine-learned model in the subset; and forming the trained chemical super learner model as a weighted average of each machine-learned model in the subset, wherein each machine-learned model in the subset is weighted in the weighted average according to its weight. Roy et al. teaches that a trained predictive model, after extensive hyperparameter tuning, could effectively predict a docking score for any given SMILES string ( paragraph [0049] ). Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to modify Lanza’s method to include hyperparameter tuning the trained model to gain the effectiveness of the model as taught by Roy ( paragraph [0049] ). In addition, Jadon discloses a method in a machine learning system comprising training a plurality of machine learning models, selecting a machine learning models in the plurality of machine learning models to form a selected machine learning model to generate a prediction ( Abstract ), wherein the selection is based on evaluation metrics such as scores and weighted average ( paragraph [0145] ). Therefore, it would have been obvious for one having ordinary skill in the art before the effective filing date of the claimed invention to modify Lanza’s method to include selecting the plurality of trained models based on the evaluation metrics such as their scores and weighted average to ensure the selected model having high performance to gain the prediction accuracy as taught by Jadon ( Abstract ). Regarding to claims 6-7, 14-15: wherein each of the training examples in the plurality of training examples further comprises a second property, wherein each of one or more machine-learned models are trained jointly with the first property and the second property and configured to return a first property prediction and a second property prediction ( Lanza: FIGs. 2-3 show the single model 205 is trained with different properties (activity scores) from the training data, and the learned model predicts a plurality of predictions 345 ). Regarding to claims 8, 16: wherein the pre-processor comprises a set of pre-processor parameters (Lanza: FIG. 5, step 510: Parameters that are used for generating the representation of molecule for single target models ). Regarding to claims 9, 17: wherein the molecular descriptor generator accepts the SMILES of each training example in the plurality of training examples and returns a vector for each training example, the vector comprising: a Morgan fingerprint representation of the training example; a Mordred representation of the training example; and an embedding representation of the training example ( In the technique of training a machine learning model, a Morgan fingerprint representation, a Mordred representation, and an embedding representation are well known vectors. Please see Wiltschko et al. (WO 2020/163860), paragraph [0077] ). Regarding to claims 10, 18: wherein the hyperparameters of each machine-learned model in the subset are tuned independently using a genetic algorithm (Lanza : paragraph [0012] ). Regarding to claims 11-12, 19-20: wherein the weight of each of the machine-learned models in the subset is determined using a sequential least-squares programming meta learner, and further comprising estimating a generalization error of the trained chemical super learner model (Jadon et al.: paragraphs [0126] , [0116]). Conclusion 07-100 Any inquiry concerning this communication or earlier communications from the examiner should be directed to LAM S NGUYEN whose telephone number is (571)272-2151. 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, DOUGLAS RODRIGUEZ, can be reached on 571-431-0716. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300. Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000. /LAM S NGUYEN/ Primary Examiner, Art Unit 2853 Application/Control Number: 18/306,610 Page 2 Art Unit: 2853 Application/Control Number: 18/306,610 Page 3 Art Unit: 2853 Application/Control Number: 18/306,610 Page 4 Art Unit: 2853 Application/Control Number: 18/306,610 Page 5 Art Unit: 2853 Application/Control Number: 18/306,610 Page 6 Art Unit: 2853
Prosecution Insights