Systems, methods, and computer-readable media for gene and genetic variant prioritization

Inventors

Gelbman, Dekel • Gurovich, Yaron

Assignees

G2i Assignment For Benefit Of Creditors LLC • FDNA Inc

Abstract

Systems, methods, and computer-readable media are disclosed for performing image processing in connection with phenotypic analysis. For example, at least one processor may be configured to receive electronic numerical information corresponding to pixels reflective of at least one external soft tissue image of an individual and access geographically dispersed genetic information stored in a database. The geographically dispersed genetic information may include numerical data that correlates anomalies in pixels in soft tissue images of a plurality of geographically dispersed individuals to specific genes or to specific genetic variants. The at least one processor may also be configured to compare the electronic numerical information for the individual with the numerical data of the geographically dispersed genetic information stored in a database, to determine at least a likelihood that the individual has a specific genetic variant, and prioritize, based on the comparison, one or more genetic variants according to likelihood of pathogenicity.

Core Innovation

The invention relates to an electronic system and computer-implemented method for performing image processing in connection with phenotypic analysis. The system receives electronic numerical information corresponding to pixels reflective of at least one external soft tissue image of an individual and performs phenotypic analysis based on the pixel-derived electronic numerical information. The approach uses image processing to generate and use representation-level data, including feature vectors for subsequent comparison.

The invention accesses geographically dispersed genetic information stored in a database. The geographically dispersed genetic information includes numerical data that correlates anomalies in pixels in soft tissue images of a plurality of geographically dispersed individuals to specific genes or to specific genetic variants. The system compares the electronic numerical information for the individual with the numerical data stored in the database to determine at least a likelihood that the individual has at least one pathogenic genetic variant.

Based on the comparison, the system prioritizes one or more genetic variants according to a likelihood of pathogenicity. The document also describes de-identification of image representations using convolutional neural networks and pooling operations to generate de-identified feature vectors derived from the external soft tissue image pixels. The document further describes use of textual phenotypic data and genetic annotations in conjunction with the pixel-derived and database-stored genetic information.

The document describes additional use cases including patient matching using phenotypic similarity via comparator scoring and distance between representations, and predicting or discovering shared genetic disorders by clustering similar patients. In these contexts, the document describes use of region-based neural networks and training, testing, and operation workflows for machine learning models, and optionally integrating genetic data to support the grouping and comparison.

Claims Coverage

The patent document includes three independent claim types: an electronic system, a computer-implemented method, and a non-transitory computer-readable medium. Each independent claim contains the same core inventive workflow: pixel-reflective external soft tissue image numerical information is compared to geographically dispersed genetic information to determine a likelihood of pathogenic genetic variants and prioritize genetic variants based on likelihood of pathogenicity.

Across the independent claims, the invention is characterized by comparing pixel-derived electronic numerical information from an external soft tissue image against geographically dispersed genetic information that correlates pixel anomalies to genes or genetic variants, determining a likelihood of at least one pathogenic genetic variant, and prioritizing one or more genetic variants according to the likelihood of pathogenicity.

Stated Advantages

Documented Applications

No documented applications found