Experimental Biological Protocols with Formal Semantics

Alessandro Abate, Luca Cardelli, Marta Kwiatkowska, Luca Laurenti, Boyan Yordanov

Bridging the Gap Between Models and the Wet-Lab

In modern biological research, a significant disconnect exists between computational models and physical laboratory execution. On one hand, researchers use precise mathematical frameworks, such as Chemical Reaction Networks (CRNs), to describe biological processes. On the other, experimental protocols—the step-by-step instructions for humans or liquid-handling robots—are often treated as "black boxes" lacking formal definition.

When experimental results deviate from model predictions, it is often impossible to determine the cause: was the underlying biological model flawed, or did a minor pipetting error cascade into a significant discrepancy? This paper introduces a formal programming language designed to bridge this gap by providing mathematical semantics for laboratory protocols, allowing for the simulation, verification, and optimization of experiments under real-world conditions.

A Core Language for Biological Samples

The proposed language centers on a set of core operations that manipulate biological samples while adhering to physical constraints. Key operations include:

  • Mix: Combining two distinct samples into one.
  • Dispense: Dividing a sample into specific fractions (e.g., extracting 40% of a solution).
  • Equilibrate: Allowing a sample to react for a defined duration.
  • Dispose: Safely discarding a sample.

To ensure physical realism, the language enforces a linearity restriction. Samples cannot be arbitrarily duplicated or deleted; every volume must be accounted for through explicit mixing, dispensing, or disposal. This mirrors the conservation of mass in a physical laboratory.

Formal Semantics: Deterministic and Stochastic Models

The true power of this framework lies in its dual semantic approach, which maps protocol steps to rigorous mathematical models:

  1. Deterministic Semantics: Protocols are mapped to systems of Ordinary Differential Equations (ODEs). This allows researchers to track how concentrations, volumes, and temperatures evolve over time as the protocol executes.
  2. Stochastic Semantics: Recognizing that laboratory environments are inherently noisy, the authors extend the model into Piecewise Deterministic Markov Processes (PDMPs). This accounts for uncertainties such as equipment tolerances (e.g., ISO 8655 standards for pipettes) and timing variations in equilibration steps.

Case Studies: From Titration to DNA Computing

The authors evaluated their framework using two primary case studies. The first is a classic Acid-Base Titration, used to demonstrate the language's ability to handle basic mixing and reaction dynamics.

Figure 4: Protocol flow for an acid-base titration, illustrating the use of Dispense (p), Mix (blue diamond), and Equilibrate (t) operations.

The second, more complex case study involves a DNA Strand Displacement (DSD) AND Circuit. This molecular programming example demonstrates how the framework can integrate a complex biological model with a multi-step protocol.

Figure 2: (A) The protocol flow for a DSD AND gate circuit. (B) The underlying chemical reaction network (CRN) that the protocol is designed to execute.

Key Insights and Results

By placing the protocol and the biological model within a Unified Context, the framework enables automated synthesis. Instead of manually estimating volumes, researchers can use the model to calculate optimal parameters to reach a specific experimental goal. Furthermore, by incorporating sensor noise through an Observe command, the model acts as a "digital twin" of the actual laboratory hardware.

In the DSD case study, the authors used Statistical Model Checking (Monte-Carlo sampling) to simulate 1,500 executions. They successfully identified "optimal bands" for experimental parameters—specific ranges of dispense volumes that maximized the probability of the DNA circuit functioning correctly despite inherent noise in reaction rates and pipetting accuracy.

Limitations and Future Directions

While this core language provides a robust foundation, several challenges remain for real-world adoption. Current semantics rely on ODEs, which may not be suitable for systems with very low molecule counts (e.g., single-cell gene expression), where Continuous-Time Markov Chains (CTMCs) would be more appropriate. Additionally, the language must be expanded to include more complex laboratory operations like centrifugation, thermal cycling, and plate shaking. Future work will also focus on improving the scalability of formal verification techniques to handle massive, high-throughput protocols.

Summary updated

Abstract

Both experimental and computational biology is becoming increasingly automated. Laboratory experiments are now performed automatically on high-throughput machinery, while computational models are synthesized or inferred automatically from data. However, integration between automated tasks in the process of biological discovery is still lacking, largely due to incompatible or missing formal representations. While theories are expressed formally as computational models, existing languages for encoding and automating experimental protocols often lack formal semantics. This makes it challenging to extract novel understanding by identifying when theory and experimental evidence disagree due to errors in the models or the protocols used to validate them. To address this, we formalize the syntax of a core protocol language, which provides a unified description for the models of biochemical systems being experimented on, together with the discrete events representing the liquid-handling steps of biological protocols. We present both a deterministic and a stochastic semantics to this language, both defined in terms of hybrid processes. In particular, the stochastic semantics captures uncertainties in equipment tolerances, making it a suitable tool for both experimental and computational biologists. We illustrate how the proposed protocol language can be used for automated verification and synthesis of laboratory experiments on case studies from the fields of chemistry and molecular programming.
BibTeX
@inproceedings{abate2017experimental,
  author = {Alessandro Abate and Luca Cardelli and Marta Kwiatkowska and Luca Laurenti and Boyan Yordanov},
  title = {{Experimental Biological Protocols with Formal Semantics}},
  booktitle = {CMSB},
  year = {2017},
  doi = {10.1007/978-3-319-99429-1\_10},
}
Figure