Partnership
Formation of the Partnership
Since the beginning of our project, our team knew that generating sequences only to satisfy the team's needs was not enough. In order to create a useful genetic tool used by different users around the globe, we need to work closely with at least one new customer, learn their needs and customize our algorithm accordingly.
After creating our first algorithm pipeline, we sent a message to the IGEM community, explaining our project and goals. Wishing to help with our new tools to other teams.
The massage we sent on the iGEM slack.
Several groups sent us messages, but no fit was found after a zoom session with each one. As time passed, we encountered Aboa messages about Toehold sequences. After a zoom session with them, we knew it was the right fit for our tool and goal. Even though our main focus was primarily not on bacteria as our experimental model, our advising committee encouraged us to add prokaryotic toeholds to the scope of our designing tool. hearing about Aboa’s project’s requirements, we understood that a demand also exists for bacterial toehold optimization tools, despite the system being better developed than in eukaryotes. We therefore made the necessary adaptations in our algorithms to allow optimization for the E. coli toeholds, which ended up in our software having an entire model adapted for prokaryotic toeholds’ design.
We clicked right after the first zoom meeting and since then, our journey began:
Our first meeting with Aboa team (Rei).
Win-Win Situation:
Our team goal was to create a generic tool for a variety of groups and needs around the globe. Moreover, we wished to test as many sequences as possible in a wet lab. Of course, we are limited in lab resources and researchers. Hence reaching out to other IGEM teams for help was the best solution in the spirit of the IGEM organization. We reached the Aboa team to learn about their new needs and to test our algorithms in their wet lab.
Aboa team has created sequences using NUPACK to achieve their goals. Our algorithm is NUPACK based, with highly sophisticated advances. Factors that NUPACK doesn't take into account. Therefore our sequences might be better than the regular use of NUPACK. With new optimized sequences, Aboa team might reach better results.
Additional zoom meeting with Aboa team (Rei and Efi).
Learning Along the Way:
By this point, we started working together on two weeks-based communication, having meetings and through emails and WhatsApp. At first, we sent Aboa team the sequences they requested, but shortly after, they mentioned we had a problem with our second-generation sequences. Aboa team method was to rank all the sequences to determine which one might work best. We were unfamiliar with that rating method at this point in the project. We learned about it only because of our partnership and meeting session with Aboa. Since then, we have kept using these methods and integrated them into our model. We worked hard on our second generation. A few weeks after, we sent Aboa team our new sequences which reached higher metric results on the B-series sequences, Aboa achieved higher metric results on the A-series sequences. By reaching better results on the B-series sequences, Aboa learned new ways to improve their algorithm.
An important input from working with Aboa’s team came up when they shared with us the metric system they used in their project, which helped us to improve our own rating system. Moreover, we exchanged feedback and experiences with them concerning the challenges we must overcome while dealing with lab and modeling works .After a back and forth exchange, two sequences were eventually chosen to be tested in their system. They used their assessment metric to compare our toeholds’ score compared to theirs, with one of our toeholds being ranked highest of all sequences examined. Sadly, wet lab validation could not be realized by the time of the wiki freeze (see Aboa team’s Results section).
Figure taken from Aboa team’s Results section, Two of our sequences are mentioned, one of them ranked highest by Aboa’s assessment metric.
To conclude, both teams stepped into this partnership with one goal in mind. We wished to expand our ability to test our program, and Aboa wished for better sequences. But along the way, each of us learned more than expected. even if our joint effort did not lead to our sequences being tested experimentally, our metric system improved by working with aboa and our work with prokaryotic toeholds inspired by them both proved detrimental for shaping our project, and we thank them for the opportunity to work together.