Single sample genetic classification via tensor motifs

A computer program product for classifying subject genetic data, the computer program product comprising:

• a computer readable storage medium readable by a processing circuit and storing program instructions for execution by the processing circuit for performing a method comprising:
• generating a set of training data for each phenotype in a database comprising a set of subjects, the set of training data generated by dividing genomic information of N subjects selected with or without repetition into windows with or without repetition, computing a first distribution of genomic events in the windows for each of the N subjects, and extracting, for each window, a first tensor that represents the first distribution of genomic events for each of the N subjects;
• training a classifier on the set of training data;
• generating a set of test data for each phenotype in the database comprising the set of subjects, wherein generating the set of test data comprises generating a random number x between 0 and 1, extracting N*x subjects in the database, and extracting, with repetition, N*(1−x) subjects in the database;
• computing a second distribution of genomic events in the windows for each phenotype of the set of test data;
• extracting, for each window, a second tensor that represents the second distribution of genomic events in the windows for each phenotype of the set of test data;
• classifying each phenotype of the set of test data with the classifier;
• updating the classifier based on a comparison of a curve of the first tensor and a curve of the second tensor;
• assigning a phenotype to a patient not found in the database using the classifier;
• adding the patient and assigned phenotype to the database;
• and
• selecting and implementing a treatment plan to the patient based on the phenotype assigned to the patient.

A processing system for classifying subject genetic data, comprising:

• a processor in communication with one or more types of memory, the processor configured to perform a method, wherein the method comprises:
• generating a set of training data for each phenotype in a database comprising a set of subjects, the set of training data generated by dividing genomic information of N subjects selected with or without repetition into windows with or without repetition, computing a first distribution of genomic events in the windows for each of the N subjects, and extracting, for each window, a first tensor that represents the first distribution of genomic events for each of the N subjects;
• training a classifier on the set of training data;
• generating a set of test data for each phenotype in the database comprising the set of subjects, wherein generating the set of test data comprises generating a random number x between 0 and 1, extracting N*x subjects in the database, and extracting, with repetition, N*(1−x) subjects in the database;
• computing a second distribution of genomic events in the windows for each phenotype of the set of test data;
• extracting, for each window, a second tensor that represents the second distribution of genomic events in the windows for each phenotype of the set of test data;
• classifying each phenotype of the set of test data with the classifier;
• updating the classifier based on a comparison of a curve of the first tensor and a curve of the second tensor;
• assigning a phenotype to a patient not found in the database using the classifier;
• adding the patient and assigned phenotype to the database;
• and
• selecting and implementing a treatment plan to the patient based on the phenotype assigned to the patient.