Natera, Inc.

Adobe, Inc.

Joshua Sweetkind-Singer

Matthew Rabinowitz

Milena Banjevic

David Walter Johnson

Zachary Demko

Dusan Kijacic

Dimitri Petrov

Jing Xu

Deepak Pai

Debraj Basu

George Gemelos

Allison Ryan

Shankar Venkitachalam

Vijay Srivastava

Kenneth Hahn

Disclosed herein is a system and method for increasing the fidelity of measured genetic data, for making allele calls, and for determining the state of aneuploidy, in one or a small set of cells, or from fragmentary DNA, where a limited quantity of genetic data is available. Poorly or incorrectly measured base pairs, missing alleles and missing regions are reconstructed using expected similarities between the target genome and the genome of genetically related individuals. In accordance with one embodiment, incomplete genetic data from an embryonic cell are reconstructed at a plurality of loci using the more complete genetic data from a larger sample of diploid cells from one or both parents, with or without haploid genetic data from one or both parents. In another embodiment, the chromosome copy number can be determined from the measured genetic data, with or without genetic information from one or both parents.

System and method for cleaning noisy genetic data and determining chromosome copy number

In some examples, a prototype model that includes a representative subset of data points (e.g., inputs and output classifications) of a machine learning model is analyzed to efficiently interpret the machine learning model's behavior. Performance metrics such as a critic fraction, local explanation scores, and global explanation scores are determined. A local explanation score capture an importance of a feature of a test point to the machine learning model determining a particular class for the test point and is computed by comparing a value of a feature of a test point to values for prototypes of the prototype model. Using a similar approach, global explanation scores may be computed for features by combining local explanation scores for data points. A critic fraction may be computed to quantify a misclassification rate of the prototype model, indicating the interpretability of the model.

Prototype-based machine learning reasoning interpretation

Methods, systems, and non-transitory computer readable storage media are disclosed for analyzing feature impact of a machine-learning model using prototypes across analytical spaces. For example, the disclosed system can identify features of data points used to generate outputs via a machine-learning model and then map the features to a feature space and the outputs to a label space. The disclosed system can then utilize an iterative process to determine prototypes from the data points based on distances between the data points in the feature space and the label space. Furthermore, the disclosed system can then use the prototypes to determine the impact of the features within the machine-learning model based on locally sensitive directions; region variability; or mean, range, and variance of features of the prototypes.

Determining feature impact within machine learning models using prototypes across analytical spaces

In various embodiments of the present disclosure, output data generated by a deployed machine learning model may be received. An input data anomaly may be detected based at least in part on analyzing input data of the deployed machine learning model. An output data anomaly may further be detected based at least in part on analyzing the output data of the deployed machine learning model. A determination may be made that the input data anomaly contributed to the output data anomaly based at least in part on comparing the input data anomaly to the output data anomaly. A report may be generated that is indicative of the input data anomaly and the output data anomaly, and the report may be transmitted to a client device.

Anomaly detection and reporting for machine learning models

The present disclosure relates to systems, non-transitory computer-readable media, and methods for accurately, efficiently, and flexibly generating digital graphical representations reflecting multiple data series in-scale utilizing dynamic y-axes. In particular, in one or more embodiments, the disclosed systems generate a normalized graphical representation portraying multiple data series in a common scale with a dynamic y-axis that portrays individualized data values based on user selection of various data series. Specifically, the presently disclosed systems and methods can generate normalized values for each of the included data series, plot the normalized values along a normalized y-axis, and include a dynamic y-axis that reflects the initial values of any of the included data series.

Generating digital graphical representations reflecting multiple data series utilizing dynamic y-axes

Disclosed herein is a system and method for making allele calls, and for determining the ploidy state, in one or a small set of cells, or where a limited quantity of genetic data is available. Poorly or incorrectly measured base pairs, missing alleles and missing regions are reconstructed and the haplotypes are determined using expected similarities between the target genome and the knowledge of the genomes of genetically related individuals. In one embodiment, incomplete genetic data from an embryonic cell are reconstructed at a plurality of loci using the genetic data from both parents, and possibly one or more sperm and/or sibling embryos. In another embodiment, the chromosome copy number can be determined using the same input data. In another embodiment, these determinations are made for embryo selection during IVF, for non-invasive prenatal diagnosis, or for making phenotypic predictions.

Growing community of inventors

San Jose, CA, United States of America

Joshua Sweetkind-Singer