# VCPI Launches First Open-Source Transcriptomics Dataset: Profiling 2,200+ Compounds in THP-1 Cells to Power AI Drug Discovery
When we launched the Virtual Cell Pharmacology Initiative last November, we made a promise: community-driven, high-quality pharmacology data, released on a rolling basis. Today, we're delivering on that promise with VCPI's first data release.* Available at thevirtualcell.com, under CC BY 4.0. *(requires Superuser early access, see below for details).
What's in the First Release
~2,280 small molecules were profiled across 6 doses [0.03, 0.1, 0.3, 1, 3, 10 µM] with a 24-hour treatment window. By looking at the structural classes — which span complex carboxamides and sulfonamides to diverse heterocycles — this dataset provides models with a rich and diverse chemical space to learn from. This diversity is key to training models that understand the fundamental rules of cellular response.
VCPI uses arrayed DRUG-seq — the most scalable, high-quality arrayed transcriptomics assay. Each compound is treated in its own barcoded well, which means every transcriptomic profile is a clean, compound-specific measurement. No computational deconvolution, no guessing which molecule did what. Transcriptional and killing controls are present in all plates, which allow us to track quality across batches. Median gene recovery is 12,000 genes/well sequenced at 2M reads/well, indicating high quality and robustness of the assay.
We originally announced V-Ref293 — an engineered HEK293 derivative — as the standardized backbone for the initiative. We made the call to switch to THP-1 cells. The reasoning is straightforward. THP-1 is a widely established human monocytic cell line that serves as a cornerstone for immunology, oncology, and inflammatory disease research, and allows us to study the intersection of immune biology and drug action — exactly where virtual cell models need to perform if they're going to be useful for real-world drug discovery.
Join the Movement: Become a VCPI Superuser
The first dataset is now available to Superusers at thevirtualcell.com. There are two simple ways to unlock Superuser status:
- Submit compounds for us to profile in future runs.
- OR Invite 3 friends or colleagues to join the platform.
This gives 6 months early access to this and all future data releases. We invite dry and wet-lab scientists and engineers across the globe to dive in and participate. If you have molecules you want profiled, send them to us and we will run them through DRUG-seq at no charge. Options to embargo and protect IP-sensitive compounds are available.
Head over to thevirtualcell.com today to sign up, become a Superuser, and download the first batch. Programmatic access via a Python client is also available using your same credentials.
Help us build the Virtual Cell.