Construction of a Calibration Curve for Lycopene on a Liquid-Handling Platform—Wider Lessons for the Development of Automated Dilution Protocols

TL;DR

This paper describes how to automate liquid-handling for lycopene in DMSO, offering lessons for broader automation in synthetic biology.

Contribution

A practical framework for identifying and monitoring transfer errors in automated liquid-handling protocols is developed and tested.

Findings

A calibration curve for lycopene in DMSO was successfully constructed using automated liquid-handling.

The study revealed issues in automated protocols when handling complex mixtures like lycopene/DMSO.

A regression-based framework was effective in controlling and monitoring transfer errors.

Abstract

Liquid-handling is
a fundamental operation in synthetic biology—all
protocols involve one or more liquid-handling operations. It is, therefore,
crucial that this step be carefully automated in order to unlock the
benefits of automation (e.g., higher throughput, higher replicability).
In the paper, we present a study, conducted at the London Biofoundry
at SynbiCITE, that approaches liquid-handling and its reliable automation
from the standpoint of the construction of the calibration curve for
lycopene in dimethyl sulfoxide (DMSO). The study has important practical
industrial applications (e.g., lycopene is a carotenoid of industrial
interest, DMSO is a popular extractant). The study was also an effective
testbed for the automation of liquid-handling. It necessitated the
development of flexible liquid-handling methods, which can be generalizable
to other automated applications. In addition, because lycopene/DMSO
is a difficult mix, it was capable of revealing issues with automated
liquid-handling protocols and stress-testing them. An important component
of the study is the constraint that, due to the omnipresence of liquid-handling
steps, errors should be controlled to a high standard. It is important
to avoid such errors propagating to other parts of the protocol. To
achieve this, a practical framework based on regression was developed
and utilized throughout the study to identify, assess, and monitor
transfer errors. The paper concludes with recommendations regarding
automation of liquid-handling, which are applicable to a large set
of applications (not just to complex liquids such as lycopene in DMSO
or calibration curves).

Figures