# 🔬Biology Is Turning Into Software — Matt McPartland & Neil Patel, Chai Discovery

## Executive summary

The intersection of biology and software is transforming drug discovery from a slow 'waterfall' process into an agile, iterative loop. Chai Discovery leverages advanced AI models (Chai-2, Chai-3) that function as sophisticated design suites—more akin to SolidWorks or Figma than ChatGPT. These platforms enable the co-design of protein sequences and structures, allowing researchers to move beyond simple structure prediction toward generating novel therapeutic candidates with high precision for complex modalities like ADCs and bispecifics.

## Key takeaways

- Platform Design vs. Chatbot Interface: The product is designed as a visual, highly functional design suite (like Autodesk or SolidWorks), allowing users to 'paint' epitopes and generate binders, rather than operating through conversational prompts.
- AI Model Progression for Drug Design: Chai models progressed from Chai-1 (structure prediction) to Chai-2 (all-atom diffusion model capable of design), crossing the threshold into generating candidate molecules that bind to a target structure, which is critical for drug development.
- Shifting Drug Discovery Paradigms: The process is moving from a costly, multi-year 'waterfall' model (target discovery $ ightarrow$ hit discovery $ ightarrow$ optimization) toward an agile, iterative loop where AI models provide rapid, promising candidates for continuous refinement.

## Technical details

- AI Model Architecture and Capabilities: The core technology involves advanced generative models: Chai-1 uses a combination of tokenizers (handling multimodality), transformers, and image diffusion models to predict structure from sequence. Chai-2 is an all-atom diffusion model that enables co-designing both the amino acid sequence and the 3D structure simultaneously, moving beyond simple prediction.
- Computational Engineering & Infrastructure: The platform requires robust engineering primitives, including sharding computation across GPU fleets and implementing durable execution logic (using tools like Temporal) to manage long-running, complex data pipelines that are prone to infrastructure flakiness.
- Protein Design Modalities: The platform supports advanced therapeutic design beyond simple binders, including: 1) Anti-Antibody Drug Conjugates (ADCs); 2) Bispecifics; and 3) Targeting specific epitopes or binding sites (GPCR agonists). The ability to control selectivity and cross-reactivity is a key differentiator.
- Model Validation & Metrics: Validation relies on comparing the model's output against independent structure prediction methods for self-consistency. Confidence scores are crucial, as models can provide calibrated confidence predictions regarding specific parts of a predicted structure.

## Practical implications

- The platform enables the design of highly sophisticated drug modalities (e.g., GPCR agonists) that are difficult or impossible to achieve through traditional immunization methods.
- By providing a neutral software factory model, Chai can scale its services across multiple pharma partners and diverse target areas without needing to develop individual drugs itself.
- The focus on engineering primitives allows for the development of complex workflows—such as those involving cross-reactivity and selectivity constraints—that accelerate pre-clinical research.

## Topics

AI/ML, Protein Engineering, Drug Discovery, Computational Biology, Software Architecture, Chai Discovery, Eli Lilly, Pfizer, Novartis, argenx

Source: https://www.youtube.com/watch?v=Qp5xklyJySI