US 12,170,135 B2
Systems and methods for evaluating query perturbations
Mason Victors, Riverton, UT (US); Berton Earnshaw, Salt Lake City, UT (US); Renat Khaliullin, Salt Lake City, UT (US); Blake Borgeson, Salt Lake City, UT (US); Peter McLean, Salt Lake City, UT (US); Nathan Lazar, Salt Lake City, UT (US); and Katie-Rose Skelly, Salt Lake City, UT (US)
Assigned to Recursion Pharmaceuticals, Inc., Salt Lake City, UT (US)
Filed by Recursion Pharmaceuticals, Inc., Salt Lake City, UT (US)
Filed on Jul. 27, 2023, as Appl. No. 18/226,917.
Application 18/226,917 is a continuation of application No. 16/905,162, filed on Jun. 18, 2020, granted, now 11,715,551.
Claims priority of provisional application 62/863,700, filed on Jun. 19, 2019.
Claims priority of provisional application 62/863,711, filed on Jun. 19, 2019.
Claims priority of provisional application 62/863,414, filed on Jun. 19, 2019.
Claims priority of provisional application 62/863,696, filed on Jun. 19, 2019.
Prior Publication US 2023/0386632 A1, Nov. 30, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G16H 20/10 (2018.01); G16H 10/40 (2018.01)
CPC G16H 20/10 (2018.01) [G16H 10/40 (2018.01)] 22 Claims
OG exemplary drawing
 
1. A computer system for evaluating a query perturbation, in a cell based assay representing a test state, the cell based assay comprising a plurality of wells across one or more multiwell plates, the computer system comprising:
one or more processors;
a memory;
an imaging device configured for capturing images via epifluorescence microscopy; and
one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, the one or more programs including instructions for:
capturing, by the imaging device under control of the one or more processors, images of wells of the one or more multiwell plates via high throughput screening using epifluorescence microscopy, such that a corresponding well of the wells is represented by a corresponding two-dimensional pixelated image having a corresponding plurality of native pixel values;
obtaining, for each respective control perturbation in a set of control perturbations from the native pixel values of the images, a corresponding control data point, thereby obtaining a plurality of control data points, wherein each corresponding control data point comprises a plurality of dimensions, each dimension in the plurality of dimensions representing a measure of central tendency of a different feature, in a plurality of features, determined across a corresponding plurality of control aliquots of cells in corresponding wells, in the plurality of wells, representing the respective control perturbation;
obtaining, for each respective test perturbation in a set of one or more test perturbations from the images, a corresponding test data point, thereby obtaining a plurality of test data points, wherein each corresponding test data point comprises the plurality of dimensions, each dimension in the plurality of dimensions comprising a measurement of central tendency of a different feature, in the plurality of features, determined across a corresponding plurality of test aliquots of the cells representing the respective test perturbation in corresponding wells in the plurality of wells;
computing a composite test vector, the composite test vector between a first point defined by a respective measure of central tendency across the plurality of control data points for each dimension in the plurality of dimensions and a second point defined by a respective measure of central tendency across the plurality of test data points for each dimension in the plurality of dimensions;
obtaining a plurality of query perturbation data points, wherein each corresponding query perturbation data point comprises the plurality of dimensions, each dimension in the plurality of dimensions comprising a measure of central tendency of a different feature, in the plurality of features, determined across a plurality of instances of query perturbation aliquots of the cells representing a respective test perturbation, in the set of test perturbations, and a first amount of the query perturbation in a corresponding subset of the plurality of wells; and
computing a query perturbation vector, between the first point and a respective measure of central tendency across the plurality of query perturbation data points for each dimension in the plurality of dimensions, such that the images captured during the high throughput screening are utilized to both uncover and differentiate highly translatable drug candidates by facilitating drug screening without having a predetermined target hypothesis.