Week 3 HW: Lab Automation

Workflow for the testing of SARS-CoV-2 using Opentrons OT-2 robots

Villanueva-Cañas et al. (PLOS ONE, 2021) describe the setup of a full SARS-CoV-2 RT-qPCR testing workflow using Opentrons OT-2 robots. This allowed them to automate what would otherwise be manual pipetting. Instead, they automated the critical steps and connected them like a lab “production line”: multiple OT-2 units to prepare plates and reagents, a system to extract RNA, and then a thermocycler to run the PCR.

What’s valuable (and novel) is that beyond simply “putting a robot in charge of pipetting,” they developed open-source Python software to make the process robust and traceable (including smart liquid handling, pipetting height control, volume control, and logging), and they show it can process 96-sample plates in a few hours with results comparable to commercial platforms. In short: they demonstrate that accessible, open automation can support a real, high-impact biological application—clinical diagnostics at scale.

Reference: Villanueva-Cañas et al. (2021) Implementation of an open-source robotic platform for SARS-CoV-2 testing by real-time RT-PCR. PLoS ONE 16(7):e0252509

An Automated Agnostic Platform for the Detection of Synthetic DNA Signatures via Genomic Foundational Models and E-CRISPR Biosensors

My proposal is focused on the design of a biosecurity platform that integrates Artificial Intelligence and CRISPR biosensors to detect human-engineered DNA sequences in the environment. Its goal is to identify synthetic biological threats based on engineering patterns, even if the pathogen is entirely new or unknown.

Below are two methodological steps of the workflow that could be implemented using Lab Automation:

1. Evasion Stress-Test (via Ginkgo Bioworks)

This involves utilizing Ginkgo’s DNA “foundries” to create synthetic sequences specifically designed to attempt to deceive the AI. By manufacturing these genomic “disguises,” we can train the model to be significantly more resilient against potential adversarial bypass attempts.

2. Diagnostic Robotization (via Opentrons)

This stage focuses on programming a robot to prepare wastewater samples and perform biosensor assays automatically (this is the component where, if synthetic sequences are detected, a field diagnostic system is deployed). This eliminates human error and ensures the system can detect minute DNA concentrations (less than 100 copies) with consistent precision. Environmental Sample Workflow for Synthetic Signature Detection The following describes a potential automated workflow for processing environmental samples to identify synthetic signatures or sequences: Reaction Preparation (Echo 525 & Bravo): The Echo 525 transfers ultra-low volumes (nanoliters) of the 10 specific gRNAs and activators designed in Phase 2 into the destination plate. Then, the Bravo performs the “stamping” of the Master Mix containing the Cas13a enzyme and RPA pre-amplification components into all wells of a 384-well plate. Detection Initiation (Multiflo & PlateLoc): The Multiflo rapidly dispenses the pre-processed environmental sample extract into all wells to initiate the molecular recognition reaction. Immediately, the PlateLoc thermally seals the plate to prevent micro-droplet evaporation and cross-contamination from DNA aerosols. Incubation and Activation (Inheco): The plate is transferred to the Inheco module for a controlled incubation at 37°C. In this stage, the Cas13a—upon identifying an “engineering signature”—is activated and begins cleaving the molecular reporters. Readout and Confirmation (XPeel & PHERAstar): Following incubation, the XPeel automatically removes the seal. Finally, the PHERAstar FSX performs a fluorescence readout (or electrochemical signal if the sensor is integrated) to quantify the presence of synthetic sequences. Data is sent directly to the foundational model to confirm if the signal corresponds to a real threat.