{
  "abstract": "Background High attrition rates and rising costs in oncology drug development highlight the need for scalable, biologically informative screening platforms. Traditional high-throughput screening (HTS) methods often rely on viability-based readouts that offer limited insight into mechanism of action (MOA). DRUG-seq (Digital RNA with perturbation of genes) combines transcriptome profiling with compound screening to address this gap, enabling rapid and scalable analysis of gene expression signatures linked to pharmacologic activity.Methods We applied DRUG-seq to both cancer cell lines and patient-derived tumor organoids. In cell lines, the protocol yielded detection of over 13,000 genes and 1 million unique molecular identifiers (UMIs) from <1,000 cells. We next adapted DRUG-seq to Champions Oncology’s CTG3D platform—a diverse biobank of patient-derived organoids (PDOs) and xenograft organoids (PDX-Os) that preserve tumor heterogeneity and 3D architecture. Organoids were seeded in 384-well plates and treated with a curated panel of standard-of-care (SOC) compounds. Transcriptomic responses were analyzed using dimensionality reduction and unsupervised clustering.Results UMAP analysis of SOC-treated CTG3D organoids revealed tightly clustered transcriptional signatures consistent with known compound MOAs and public datasets. The assay demonstrated high sensitivity and reproducibility across replicates. siRNA-mediated knockdowns in CTG3D organoids showed consistent downregulation of target gene sets, validating the platform’s ability to detect functionally relevant expression changes. These datasets were further integrated with AI/ML tools to model transcriptional responses, infer compound MOA, and predict the effects of unscreened compounds.Conclusions DRUG-seq applied to CTG3D organoids enables scalable, transcriptomics-driven drug screening in clinically relevant tumor models. By integrating rich expression data with AI-based analytics, this platform supports mechanistic insight, compound prioritization, and toxicity prediction. The combination of high-throughput transcriptomics and patient-derived 3D biology offers a powerful tool for accelerating preclinical discovery and advancing personalized therapeutic strategies.",
  "authors": [
    {
      "affiliations": [
        "Champions Oncology, Rockville, MD, USA"
      ],
      "name": "Devin Porter"
    },
    {
      "affiliations": [
        "Champions Oncology, Hackensack, NJ, USA"
      ],
      "name": "Christine Baer"
    },
    {
      "affiliations": [
        "Champions Oncology, Hackensack, NJ, USA"
      ],
      "name": "Kurtis Cruz"
    },
    {
      "affiliations": [
        "Champions Oncology, Rockville, MD, USA"
      ],
      "name": "Henry Gervaise"
    },
    {
      "affiliations": [
        "Champions Oncology, Rockville, MD, USA"
      ],
      "name": "BanuPriya Sridharan"
    },
    {
      "affiliations": [
        "Champions Oncology, Hackensack, NJ, USA"
      ],
      "name": "Brandon Walling"
    }
  ],
  "title": "148 DRUG-seq in CTG3D organoid models: a high-throughput, transcriptomics-driven platform for AI-enhanced drug discovery",
  "uid": "63bdfc69-3466-5161-87d5-029dd7943cdd"
}