An end-to-end solution for automation of biological protocols

Resumen

The inability to reproduce the results of biological research has long been the elephant in the room. Non-reproducibility of results causes billions of dollars in losses in money, time and other resources. This slows down over all scientific progress. There are some important factors which contribute to the reproducibility problem. There is ambiguity in experimental method specification, human error introduced while conducting experiment and lack of data sharing standards. Recently, some interesting approaches have been developed to alleviate the reproducibility problem. They are the use of 1) programming languages for removing ambiguity in experimental method specification, 2) use of robotic automation for conducting experiment to reduce human error and 3) use standards and ontologies for improving data sharing. The first two solutions have met with limited success. This is because use of programming languages requires the user to have prior programming knowledge. Even if the user has coding skills, different automation solutions use different programming languages. This requires a lot of time and effort on the researcher's part as they need to learn different languages to use different automation solutions. Most automated solutions today come with a graphical user interface (GUI) which circumvents the programming obstacle. However, different automated solutions again have different GUIs which need considerable effort to master. Programing languages and Automation were used as an approach to tackle the reproducibility problem. Automation also offers the additional benefits of improving efficiency and productivity in the lab. A web-based end to end automation framework was developed called the BioBlocks Automation Framework (BAF). BAF allows users to specify their experiments (Protocol Specification App) along with the machine (wet-lab hardware) on which it would it automated (Machine Specification App), followed by generation of instructions for actual execution (Protocol Execution App). It allows users to automate their experiments without the need for any prior programming knowledge. It uses a block-based graphical interface based on MIT’s Scratch and Google’s Blockly. They use a jig-saw like interface to teach beginners how to program. It makes it easier to learn programming as users do not need to learn syntaxes. Using the BAF, users can specify their protocols and machines by a simple drag and drop mechanism. The working of the three main components of BAF is explained as follows: ● A protocol specified in the Protocol Specification App using blocks is automatically translated in real-time to give multiple outputs. It is translated to English for cross-validation, a Protocol workflow for visual aid and an Output code in which the protocol is represented in JSON format for later automation. The protocol specified are platform agnostic i.e. they are independent of the platform on which they will be executed. The specified protocols are also unambiguous and can be executed manually using English Translation. ● A machine (wet-lab hardware) specified in the Machine Specification App using blocks is translated automatically in real-time to generate a machine layout to serve as visual aid and Output code (representation in JSON format). It currently has support for milifluidic devices. However, the modular architecture of the framework will allow users to specify other microfluidic devices as well. It follows a unique part-machine specification paradigm to make the specification extensible to other hardware platforms. ● Using the Protocol Execution App, the specified protocol can be automated over the specified machine. The App automatically checks if the protocol and machine are compatible, followed by routing the flow over the machine setup and finally generating machine instructions to control the actuation. BioBlocks framework is a general-purpose end to end automation framework. It is not tied to a single hardware platform and the protocols specified are independent of the platform. The protocols and machines described in the framework can be easily saved, shared and modified. It has been designed to allow it to connect with other protocol specification tools and other hardware platforms. This work aims to lower the entry barrier for non-computer scientist to automation tools by removing the programming bottleneck. It aims to enable biologists, hardware engineers and Do-it-Yourself enthusiasts by providing them a platform where experiments can be designed, executed and shared.