HC StratoMineR Introduction

Core Life Analytics' StratoMineR™ software helps biologists analyze the complex data derived from high-content image analysis. A powerful, intuitive workflow allows users to port numeric data produced by IN Carta® Image Analysis Software directly into StratoMineR. Here the data is used to generate rich phenotypic profiles using advanced data mining methods, and explore them with interactive visualizations.

.

Introduction to Stratominer

IN Carta Image Analysis Software provides robust, quantitative results from complex biological images and datasets utilizing advanced AI technology. Directly import this data into StratoMineR, an intuitive web-based platform which guides users through a typical workflow in analysis of high-content, multi-parametric data.

https://share.vidyard.com/watch/vpastC7BEbrRQhkqAHEu4V

Hit Selection

Calculate phenotypic distance scores, based on the principal components, and compare the phenotypes of your samples to those associated with your controls, for example. Alternatively, build machine Learning models to predict a compound’s probability of being in a certain reagent class.

https://share.vidyard.com/watch/iPgeLFchWDHMitRUwBGcms

Data Reduction

Data reduction, usingPrincipal Component Analysis (PCA) for example, is essential when your dataset contains a large number of features. PCA transforms a large set of features into a smaller set of variables (components) that still contains most of the original information. Projecting thousands of features into a smaller number of components makes it easier to explore and visualize your data.

https://share.vidyard.com/watch/NydP7t6Kgbp5g2wHbLMQa1

Feature Selection

Image analysis software can generate hundreds or thousands of phenotypic measurements per cell, so-called features. In this step, StratoMineR helps you to curate the features in your dataset by identifying and removing irrelevant or redundant features. StratoMineR uses several criteria for this, and removes for example features with a standard deviation of 0, or a maximum Spearman’s rho of > 0.99.

https://share.vidyard.com/watch/pYHef2LQn5n3fk3GLa948C

Quality Control

The first step of QC is what we call “visual data mining”: using interactive visualizations, with filters and tiling options, users can really dig into their raw data. Quality metrics, such as the Z’, can be calculated here as well.

https://share.vidyard.com/watch/wmi12nhXoJ8yNuXrvbp1Vw