At Fulcrum Therapeutics, we aim to develop treatments for rare genetic diseases by modulating gene expression at the source. In our drug discovery platform, we probe our sophisticated cell-based disease models with small molecule and genetic perturbagens at scale to swiftly identify and deliver candidate drug targets for clinical development. Here we present FulcrumSeek, an automated transcriptomic and phenotypic profiling platform for high-throughput drug discovery and functional genomics screens.
The FulcrumSeek platform has been custom-built to maximize flexibility and data quality. The transcriptomic profiling branch of FulcrumSeek features a mini-bulk RNA-sequencing workflow optimized for universal compatibility with a wide range of disease-relevant cell models. A direct lysis and 3’ barcoding scheme allow for up-front sample pooling, minimizing reagent usage and labware handling in downstream steps. The barcoded reverse transcription product can be used for full transcriptome and/or custom targeted sequencing library generation. The 3’-based method, along with UMI integration for identifying and eliminating PCR duplicates, creates a robust and cost-effective platform to assess the expression of thousands of genes in each well.
The high content imaging branch of FulcrumSeek utilizes a fully automated cell staining workflow to profile markers of differentiation, cell structure, and organelles. A machine learning approach to image analysis allows us to identify intracellular features and cluster phenotypic responses to chemical and genetic perturbagens. Considered together with companion transcriptomic data, the FulcrumSeek platform provides a holistic view of cellular response to perturbation for a broad range of cell models.
The true power of FulcrumSeek lies in the ever-expanding high dimensional data matrix, thus far derived from over 10 highly specialized disease models, thousands of small molecule and CRISPR perturbagens, multiple treatment timecourses, and tens of thousands of genetic and phenotypic features captured from each condition. We have verified our platform by confirming identification of known targets, and now use artificial intelligence tools to classify cellular response to perturbagens and identify novel targets for rare disease. As our data matrix grows, so does our ability to draw associations between perturbation and outcome, accelerating our power to predict paths forward to treat rare diseases.