Team:Imperial College/Awards

Awards

for Judges

Medal criteria

bronze medal

Attribution

The creative, wet lab, and dry lab work was almost fully done by the members of the team. Several academic lab group leaders and staff gave us advice on the project, on assembly protocol details, and helped us source reagents. Imperial College Ellis lab postdoc Will Shaw created a dCas9 yeast strain for us and helped us source, prepare and extract BioBricks parts. Imperial College PhD student Liam Hallett helped us further with the BioBricks parts preparations, which turned out to be more challenging due to being remainders from 2016 and 2018. We additionally received MoClo parts from Alberto Scarampi, a Cambridge PhD student.

Project Description

We have created a comprehensive software pipeline, taking users from part design to automated assembly and transformation. We are empowering our users through easy access to combinatorial design for BioBricks, MoClo, and BASIC, all in a standardised format, enabling future scalability. We made our software tool accessible through an intuitive, web-based GUI and thorough documentation, communicated with potential end-users about functionality and validated it through our wet-lab project and partnership. The feedback we received was translated into modifications and features, such as improving SBOL Designer’s visuals for added user friendliness, adding customisability to labware, and providing detailed information to users. In development of our software we identified issues and added new features to tools including pysbol2, the python library for SBOL. We can confidently say that we fulfilled all of the requirements we set out for our software pipeline.

Contribution

Our contribution to future iGEM teams encompasses enhancing the accessibility of SBOL as a programming standard through the creation of an SBOL parser that uses the open-source Python labware library Plateo. Although developing with pySBOL compared to the established Java API libSBOLj cost us weeks of debugging and implementing features, we are better able to share our understanding of SBOL as a programming interface and elucidate its functionality through the Python API, as the language itself is much more novice-friendly, modern and interoperable. Other teams wanting to appropriate the standard into their own software tools will be able to look at our documentation, use case, and benefit from the pull requests we got accepted to the official version of pySBOL, not only serving to enrich the type of data that iGEM teams may benefit from, but also encouraging further community development of the SBOL standard.

Throughout the course of our project we aimed to contribute to and collaborate with the opens source community. SBOL was chosen as a data standard for its open source nature and that it could be amenable to development by a dedicated community. Members of our team have directly contributed to pysbol2, the python library for SBOL, and through the use of the SBOL standard have helped to push for improvements and fixes to existing functionalities. Our additions in pysbol2 include combinatorial derivation and enumeration. We were fortunate to speak with one of the creators of SBOL, Professor Christ Myers, and directly receive his feedback and guidance.

We would like to acknowledge the Imperial College tool DNABot, and the DAMP Lab’s MoClo tool. These were an excellent basis for our BASIC and MoClo protocols respectively. We updated both DNABot and DAMP Lab’s MoClo tool to Opentrons version 2, added many customisable features, and created additional information for the user. Our supervisor, Professor Baldwin, was one of the authors of DNABot, and provided invaluable inspiration and feedback. The DAMP Lab presentation at the International Workshop on Biodesign Automation helped inform many of the additions we made to their tool and others.

silver medal

Collaboration

We mentored 5 iGEM teams over the course of their projects in mathematical modelling to simulate their genetic circuits, which culminated in an iteratively rigorous yet simple, feedback-defined modelling introduction and tutorials package. In addition, we collaborated with LiU iGEM Linköpings team who guided us through the beta test of our tool, analysing the relevance of each feature we were implementing thanks to their feedback testing assessment form.

Go to our Collaboration page here.

Human Practices

Our project has three core values that have been fundamental thoughout its completion: communication, accessibility and validation. We have been fully committed to them and we looked for feedback in multiple ways from other iGEM teams, both high school and undergraduate level and from the general public with webinars and the Science Slam.

Go to our Human Practices page here.

Engineering Success

We achieved engineering success through our Tryptophan project following the engineering design cycle. We initially came up with ideas for producing various important products. After discussions with academics, we found that Tryptophan optimisation in yeast would be interesting due to its limiting the production of many such products. We then came up with the idea of simultaneously using gene overexpression and CRISPRi. We chose several targets which we modelled with parsimonious Flux Balance Analysis, narrowing down a large design space. Unfortunately, lab access was restricted though the project would have proceeded using automated DNA assembly with SOAP-Lab to carry out a combinatorial build sampling the design space. We would then test each design using LC-MS. LC-MS data would then be used to build models predicting the optimal design with Design of Experiments. Finally, we would build and characterise several of the designs predicted to give high tryptophan yields.

Go to our Engineering page here.

Proposed Implementation

We started our iGEM project by noticing that there was a discrepancy between biologists specialised in wetlab work and the tools that allow lab automation, designed to help them. The biologists often lack awareness and the skills and might prefer to use previously established practices so we created a tool that would be easier to utilise than current workflows for DNA Assembly. Every step is incorporated in the flow so from the production of Genetic designs, access to libraries of genetic parts such as SynBioHub as well as being able to produce your own, all in a data format (SBOL) that allows for the sharing of genetic designs with large amounts of information, particularly useful for collaborations. The final part of the pipeline the generation of an opentrons script was validated with iGEM Hamburg in just an afternoon.

To bring SOAP Lab to a wider audience we plan to integrate it to a larger pipeline such as Galaxy SynBioCAD, which incorporates the entire design-build-test-learn cycle, and are discussion with the creators at the moment about incorporating part of our software tool into their open-source platform.

Go to our Implementation page here.

gold medal

Proof of Concept

We designed several genetic constructs for validating the specific assembly methods supported by our automatic liquid handler program generation to debug our software and check that the protocols were defined correctly in the code. After getting access to our lab at Opencell, we managed to test all of our assembly methods and iteratively update our code upon lengthy observations of the robots in action. We put our hardened software into action with user testing through our partnership with iGEM Hamburg, who helped us test our software from beginning to end. Giving us feedback throughout, they were able to construct both the constructs for their own project and the BioBricks parts for the later tests that they carried out for us. Despite identifying many points for our own improvement, we were very happy to see that Hamburg were able to set up the robot and run the scripts in an afternoon, showing the usability not only of SBOL in our product, but also proving the assembly protocols resulting from these designs ran as expected.

Go to our proof of concept page here.

Integrated HP

Integrating the community-maintained SBOL data standard into our software forced us to climb the steep learning curve to understanding the technicalities and justifications behind its inner workings. We made the standard more accessible by creating an in-browser version of SBOL Designer Java application, enhancing and adding features in the SBOL Python API, which is currently less developed compared to the SBOL Java API, and integrating the script generation into an established synthetic biology software workflow platform, GALAXY SynBio-CAD. In the process of interfacing with the developers of the standard, we honed in the interoperability between our software and others by adopting standardised intermediates such as the Plateo Python labware library.

Go to our Integrated Human Practices page here.

Science Communication

A key aspect of our project was inviting the general public into the conversation. This is why we participated in three events, two of our own making. We participated in the Science Slam to bridge the gap between scientific research and the general public, giving them the opportunity to be aware of the development in science. We decided to focus on the topic of ‘Ethics of automation’. To fully understand the ethical implications automation has, we invited experts to talk about their opinion in a webinar format. We held an additional webinar to understand how lab scientists and software engineers have been collaborating both in academia and tech companies. Our aim with this webinar was to promote interoperability and communication between bench scientists and developers, and to learn for ourselves how the communication we want to promote is happening in the real world.
Thanks to one of our team member's previous iGEM experiences on a Highschool team, we realised how many teams had limited modelling experience, especially after several enthusiastically responded upon our offer to help them. We mentored 5 iGEM teams over the course of their projects in mathematical modelling, primarily of genetic circuits. Through these meetings, we noticed a trend in the guidance they required and found that there were no modelling resources accessible to high school and undergraduate teams. Out mentorship program therefore culminated in an iteratively rigorous yet simple, feedback-defined introduction to mathematical modelling package, consisting of a handbook and a video tutorial series. As well as the theory, we identified programming as a vital, yet challenging skill for modelling, which we integrated into the educational package. Our resource provides a broad, solid foundation which enables readers to access more complex literature and develop their own models.

Go to our Education page here.