BioGPU

Biology GPU (BioGPU), developed by AnimaBiotech, is the first experimental runtime for AI in drug discovery, enabling AI models and scientists to experimentally interact with living cells through a paradigm called Visual Biology Computing. Designed for pharmaceutical researchers and drug discovery teams, BioGPU bridges the gap between computational hypothesis generation and real biological validation by executing AI-derived hypotheses directly inside disease-relevant cellular systems at scale.

BioGPU addresses a fundamental challenge in drug discovery: under perturbation, cells activate many pathways simultaneously, making it difficult to distinguish causal drivers from associated passengers. By making the full biological consequences of perturbation visible early in the discovery process, BioGPU enables teams to make better-informed decisions across the entire discovery lifecycle — from pathway and target biology resolution through to molecule characterisation.

Core Platform Capabilities

• Large-scale experimental infrastructure: Executes AI and omics-derived hypotheses directly in cells, across multiple disease-relevant cellular systems, enabling parallel rather than sequential target investigation.
• Visual computing as a new biological modality: Observes pathway activity and target effects across millions of cellular images, interpreted by a model trained on over 2 billion pathway images spanning approximately 140 cell types and 25 diseases.
• Parallel experimental execution: Resolves disease and target biology at scale simultaneously, rather than investigating one target at a time.
• BioGPU Agent: Enables scientists or large language models (LLMs) to design experiments, run visual compute, and iteratively test hypotheses against observed pathway activity in an autonomous experimental reasoning loop.

Biology Resolution Across the Discovery Lifecycle

• Pathway Biology Resolution: Experimentally resolves the truly active pathways under a given perturbation or disease condition.
• Target Biology Resolution: Separates causal driver targets from associated passenger targets, improving confidence in target selection.
• Molecule Biology Resolution: Identifies how active molecules work mechanistically and reveals where they act off-target.
• Disease Biology Resolution Models: Experimentally generated foundational models that serve as the core biological layer of the BioGPU platform, covering multiple cell types across diseases, conditions, and perturbations.

Visual Biology Trained Models

• Large Visual Computing Model: Generated from 25 disease models trained in real projects with pharmaceutical partners, encompassing over 2 billion process visualisations in total.
• Visual mRNA Biology Model: A domain-specific model that visualises mRNA biology regulatory mechanisms, reflecting the team's deep expertise in mRNA biology.

BioGPU Agent and Experimental Reasoning Loop

• Enables LLMs such as ChatGPT to function as autonomous drug discovery engines by iterating within an experimental loop with the Biology GPU.
• Demonstrated capability to resolve causal biology in disease areas such as ALS through iterative hypothesis testing and visual compute cycles.
• Supports both human scientists and AI agents in designing experiments, executing visual biology compute, and refining hypotheses based on observed cellular pathway activity.

Discovery Workflow

1. Hypothesis generation: Hundreds of disease hypotheses are generated computationally from genomics, transcriptomics, proteomics, pathology, and cellular morphology data.
2. BioGPU execution: The platform sees and computes on biological processes and pathways inside cells, applying visual computing to interpret pathway activity across millions of images.
3. Causal target identification: Targets are ranked by cellular activity, with causal drivers clearly separated from passenger associations to prioritise the most promising candidates.

Team and Expertise

• The AnimaBiotech team combines deep expertise across mRNA biology, artificial intelligence, software development, computational biology, medicinal chemistry, imaging, and operations.
• Leadership includes co-founders with advanced degrees in mRNA biology and scientific research, supported by experienced executives in software, business development, and finance.
• The scientific advisory board includes recognised experts in mRNA biology, and the platform has been validated through real-world projects with pharmaceutical industry partners.

BioGPU integrates with AI systems including large language models and supports omics data modalities such as genomics, transcriptomics, and proteomics. The platform has been developed and validated in collaboration with pharma partners across 25 disease models, positioning it as a foundational experimental layer for AI-driven drug discovery programmes.