RECURSION PHARMACEUTICALS, INC.

The present disclosure relates to systems, non-transitory computer-readable media, and methods for efficiently acquiring biological data via active machine learning. In particular, in some embodiments, the disclosed systems generate, utilizing a map prediction machine learning model, similarity prediction confidence scores for a plurality of perturbation pairs. In addition, in some embodiments, the disclosed systems determine, from the similarity prediction confidence scores utilizing an acquisition function, a perturbation pair for developing a ground truth similarity. Moreover, in some embodiments, the disclosed systems generate a tuned map prediction model by comparing a similarity prediction for the perturbation pair generated by the map prediction machine learning model with the ground truth similarity. Furthermore, in some embodiments, the disclosed systems utilize, in response to determining that a measure of confidence for the tuned map prediction model satisfies a stopping criterion, the tuned map prediction model to generate a blended map of biology.

A method for designing a cell-based screening assay experiment includes, at a graphical user interface, receiving a first input identifying a template plate definition and receiving a second input defining a first group of disease reagents and at least one destination concentration of the disease reagents. The method also includes receiving a third input defining a first group of treatment reagents, at least one destination concentration for the treatment reagents, a number of replicates of the treatment reagents, and a selection to associate the first group of treatment reagents with the first group of disease reagents. The method further includes generating, based at least partially on the inputs, a design for the cell-based screening assay experiment, including determining a number of test wells required for performing the experiment, and distributing the test wells over a one or more test plates that are instances of the template plate definition.

The present disclosure relates to systems, non-transitory computer-readable media, and methods for embedding perturbation data via a machine learning model and filtering, aligning, and aggregating the embeddings to generate a genome-wide perturbation database for real-time generation of perturbation heatmaps. In particular, in one or more embodiments, the disclosed systems can receive a plurality of perturbation images portraying cells from a plurality of wells corresponding to a plurality of cell perturbations. Further, the systems can generate, utilizing a machine learning model, a plurality of well-level image embeddings from the plurality of perturbation images. Moreover, the systems can align, utilizing an alignment model, the plurality of well-level image embeddings to generate aligned well-level image embeddings. Additionally, the systems can aggregate, according to perturbations of one or more perturbation experiments, the well-level image embeddings to generate perturbation-level image embeddings. Furthermore, the systems can generate perturbation comparisons utilizing the perturbation-level image embeddings.

The present disclosure relates to systems, non-transitory computer-readable media, and methods for embedding perturbation data via a machine learning model and filtering, aligning, and aggregating the embeddings to generate a genome-wide perturbation database for real-time generation of perturbation heatmaps. In particular, in one or more embodiments, the disclosed systems can receive a plurality of perturbation images portraying cells from a plurality of wells corresponding to a plurality of cell perturbations. Further, the systems can generate, utilizing a machine learning model, a plurality of well-level image embeddings from the plurality of perturbation images. Moreover, the systems can align, utilizing an alignment model, the plurality of well-level image embeddings to generate aligned well-level image embeddings. Additionally, the systems can aggregate, according to perturbations of one or more perturbation experiments, the well-level image embeddings to generate perturbation-level image embeddings. Furthermore, the systems can generate perturbation comparisons utilizing the perturbation-level image embeddings.

The present disclosure relates to systems, non-transitory computer-readable media, and methods for utilizing language machine learning model (LLM) as autonomous reasoners to navigate and execute multiple layers of a computerized bio-activity discovery pipeline of a tech-bio exploration system. In particular, the disclosed systems can utilize an LLM that learns to access one or more tech-bio exploration tools to execute one or more processes and/or tasks in a bio-activity discovery pipeline. For instance, the disclosed systems can provide an interactive query prompt interface to enable users to provide tech-bio queries (as prompts) and utilize the LLM with the prompts to execute one or more tasks in the bio-activity discovery pipeline to generate and/or retrieve bio-activity data for the query. Moreover, the disclosed systems can utilize one or more LLMs to autonomously utilize and/or interact with one or more tech-bio tools in the bio-activity discovery pipeline to generate and/or obtain bio-activity data.

The present disclosure relates to systems, non-transitory computer-readable media, and methods that a implement a framework for active learning to discover pairwise interactions via representation learning. Indeed, in one or more implementations, the disclosed systems generate a first individual perturbation embedding from a first representation of a first cell exposed to a first perturbation and a second individual perturbation embedding, from a second representation of a second cell exposed to a second perturbation. For instance, the disclosed systems combine the first individual perturbation embedding and the second individual perturbation embedding to determine a predicted pairwise embedding. Moreover, in some instances, the disclosed systems generate a pairwise embedding from a representation of a cell exposed to both the first and second perturbation. Additionally, from comparing the predicted pairwise embedding with the pairwise embedding, the disclosed systems generate a measure of biological interaction of the first and second perturbation.

Disclosed are novel compounds according to formula (I), their pharmaceutically acceptable salts, and pharmaceutical compositions thereof. Also disclosed are methods of using such compounds and compositions to treat various diseases, disorders and conditions.

The present disclosure relates to systems, non-transitory computer-readable media, and methods for training and utilizing neural network potential models to generate quantum mechanics property predictions of different classes from training compound geometries and/or query compound geometries. For example, the disclosed systems can utilize a backbone neural network to generate a latent feature representation of a training compound geometry and/or a query compound geometry. Moreover, the disclosed systems can utilize different task heads to generate a quantum mechanics property prediction from different classes from the latent feature representation. Further, the disclosed systems can train neural network potential models based on ground truth quantum mechanics property predictions from the different classes to improve functionality of implementing computers.

The present disclosure relates to systems, non-transitory computer-readable media, and methods for utilizing a contrastive molecular-phenomic embedding model that learns joint latent space embeddings between molecular structures and phenomic images to generate molecular-phenomic embeddings that represent molecular impacts on cellular functions. Indeed, the disclosed systems can utilize phenomic image embeddings generated from a pretrained phenomic image encoder model and corresponding molecular structural embeddings with a contrastive molecular-phenomic embedding model to learn a joint latent space between molecular structures and phenomic images utilizing a modified rank-n-contrast loss with a learnable temperature parameter. In addition, the disclosed systems can utilize molecular structures and/or phenomic images with the contrastive molecular-phenomic embedding model to generate molecular-phenomic embeddings that enable a variety of molecular inferences (e.g., similar molecule determinations, similar phenomic image determinations, phenotypic impact determinations from particular molecules, molecular activity classifications, and/or inactive region filtering).

The present disclosure relates to systems, non-transitory computer-readable media, and methods for utilizing a contrastive molecular-phenomic embedding model that learns joint latent space embeddings between molecular structures and phenomic images to generate molecular-phenomic embeddings that represent molecular impacts on cellular functions. Indeed, the disclosed systems can utilize phenomic image embeddings generated from a pretrained phenomic image encoder model and corresponding molecular structural embeddings with a contrastive molecular-phenomic embedding model to learn a joint latent space between molecular structures and phenomic images (with an inter-sample similarity aware loss (S2L), explicit and implicit concentration encoding, and under sampling of inactive molecular data). In addition, the disclosed systems can utilize molecular structures and/or phenomic images with the contrastive molecular-phenomic embedding model to generate molecular-phenomic embeddings that enable a variety of molecular inferences (e.g., similar molecule determinations, similar phenomic image determinations, phenotypic impact determinations from particular molecules, and/or molecular activity classifications).