This talk will focus on broad trends in AI development and deployment, using high content screening as a case study. Specifically, we aim to examine questions in data quality, data provenance, model scorecard and communication with a perspective informed from  both development and deployment challenges. Rubrics that might aid the potential auditing of AI methods in this space will also be discussed.

In this talk I will present an overview of different AI approaches applied in terms of target identification, as well as some literature examples and applications. The talk will also provide some context on the evolution of this space for drug discovery and development.

Charting an organs’ biological atlas requires us to spatially-resolve entire single cell transcriptomes. scRNA-seq can profile cells comprehensively, but lose spatial information. Spatial transcriptomics allows for spatial measurements, but either at lower resolution or with limited gene throughput. To overcome these limitations we present Tangram, a method that aligns sc/snRNA-seq data to various forms of spatial data collected from the same region.

The goal of the Illuminating the Druggable Genome (IDG) Program is to catalyze research in areas of biology that are currently understudied but that have high potential to impact human health by identifying biochemical, cellular, or animal model phenotypes for understudied proteins from druggable gene families (kinases, ion channels, GPCRs), enabling further investigation of those proteins by providing reagents and tools, and generating, maintaining, and facilitating the use of a minable knowledgebase.

Chemical probes are potent and selective small molecule inhibitors that can be used in cellular assays to induce a phenotype.  Screening a library of epigenetic chemical probes in cellular assay from patient tissue, we have discovered novel targets and lead compounds for developing treatments for two rare diseases: Dupuytren's contracture and Friedreich’s ataxia.

The AbbVie Functional Genomics team has established a robust flow cytometry-based cell sorting platform for genome-wide CRISPR screening to identify regulators of cell function relevant to autoimmunity and cancer. Using knockout and activation screens we have defined cell-intrinsic and stimuli-specific regulators of Programmed cell death ligand-1 (PD-L1) expression in mouse and human tumor lines. These screens provide insight into PD-1/PD-L pathway regulation and identify potential targets for immune-oncology therapies.

Systematic approaches to identify new regulators of T cell functions in vivo can provide potentially orthogonal or complementary opportunities. We recently performed in vivo CRISPR screens in CD8 cytotoxic T cells, both in a genome-scale and in a focused manner, in tumor models of immunotherapy. Our screen re-discovered prime immunotherapy targets such as PD-1, TIM-3 and LAG3, as well as previously undocumented targets. We characterized novel targets for their activity in mouse and human CD8 T cells including CAR Ts.

Single cell applications provide increased resolution that can be used to elucidate possible drug target pathways and improve characterization of drug response. However, each sample type is unique and there are a number of technical considerations that influence data quality of single cell genomics. Learn about key sample preparation considerations and resources for single cell applications and how these can be leveraged for drug discovery research.

Measurements of changes in gene activation and expression provide a basis to understand genetic changes causing biological responses. Cell-to-cell gene disruption induced by CRISPR and other gene-perturbation technologies help tease out the drivers of these responses. Adaptations of these two screening approaches can be used to discover the genetic drivers responsible for phenotypic variabilities such as drug sensitivities and disease variation across tissue microenvironments, and between single cells.

In vivo somatic genome editing and the generation of Somatically Engineered Mouse Models (SEMMs) has opened the possibility to perform functional screens directly in vivo (DIVoS: Direct In Vivo Screen). In this talk the speaker will provide an introduction on the approaches of in vivo somatic editing and a couple of examples on how the technology can be used for in vivo target identification.

TNIK is amplified in approximately 50% of LSCC cases. TNIK genetic depletion or pharmacological inhibition reduces the growth of LSCC cells in vitro and in vivo. In addition, TNIK inhibition showed antitumor activity and increased apoptosis in established LSCC patient-derived xenografts. Mechanistically, we identified the tumor suppressor Merlin/NF2 as a novel TNIK substrate, and show that TNIK and Merlin are required for the activation of focal adhesion kinase.  

This talk will discuss the application of 3D spheroid-based high content screening for target identification and validation studies.