Team:SJTU-Software/Partnership

{IGEM_TopBar}}

　　 Team:SJTU-Software/Partnership - 2021.igem.org

Tongji_Software
SJTU-BioX-Shanghai

SJTU-Software & Tongji_Software

Through China iGEMer communication group, Yanping Zhao, leader of Tongji_Software, and Hu Yuanjie Hu, leader of SJTU-Software got in touch on April 20. Although the projects of two teams are not similar, because both of us act as software teams and have similar biological and computer backgrounds, we have made a lot of suggestions to each other in terms of project core and algorithms. In addition, we have helped each other in many ways.

In April-May

SJTU-Software←→Tongji_Software

During this period, the leaders of the two teams occasionally exchanged ideas via wechat about how the software teams could help the synthetic biologists and igemers remove obstacles.

It was worth mentioning that "We wanted to design a software to assist the design of Toehold RNA Switch" in the brainstorming of Tongji_Software Team at that time. Although Tongji_Software did not carry out this project, they accumulated a lot of knowledge during the investigation of this idea, including the prediction of secondary structure of RNA. This knowledge reserve facilitated the subsequent communication between the two teams.

In addition, the leaders of the two teams successfully met offline for the first time on May 1, thanks to the fact that both teams are based in Shanghai.

In June

SJTU-Software←→Tongji_Software

After the two teams roughly determined the project themes, we had a preliminary exchange with each other on respective project backgrounds, project significance, and methods involved. In the meeting, we mentioned the idea of building a DNA computing platform to assist in early cancer screening, mainly related to secondary structure prediction. Tongji_Software hopes to find a way to fight against super bacteria by comparing the sequence similarity between bacteriophages and bacteria. Tongji_Software recommended a software “NUPACK” for RNA secondary structure prediction to us, and we suggested for Tongji_Software: “The Tongji_Software should try to get in touch with other teams related to intestinal flora and phages.”

In July

SJTU-Software→Tongji_Software

We proposed to Tongji_Software that "the prediction effect of RNA secondary structure is not good". After we introduced the whole project process, Tongji_Software found that we used the "one-hot" method when encoding nucleic acid sequence information. At this time, Yanping Zhao, leader of Tongji_Software, remembered that she had read a paper "Low-N Protein Engineering with Data-efficient deep Learning". This paper points out that "the sparse encoding obtained by using one-HOT encoding to represent the word vector has a large dimension and is isolated and unrelated between words, so semantic information between words cannot be expressed". After Tongji_Software told us this point of view, we carried out an investigation and changed the encoding method from "one-hot" to the current "word-to-vector".

The advice provided by Tongji_Software paved a new way for us, which further improved the accuracy of our prediction results. Anyway, thank you very much for their valuable advice.

In early August

SJTU-Software→Tongji_Software

During the offline communication between the two teams, Tongji_Software proposed that "the method currently used is sequence alignment——Spacer in bacterial genome is first found and then compared to phage genome. But this method has high accuracy and low detection rate." After communication, we put forward valuable suggestions to Tongji_Software. We thought that "since the detection rate of comparison method is too low, why not try non-comparison method?" It was at our suggestion that Tongji_Software added alight-free sections, which supplemented the project core and made the project more complete.

Since CCIC

At the CCIC meeting, the two teams communicated offline again. In this process, Tongji_Software pointed out that our project itself was too specialized and difficult to understand. After the discussion, we made improvements in project sharing and project presentation, which made it easier for the audience to understand.

In addition, the two teams communicated with each other about education, engineering and other sectors that had doubts, and the two teams had a clearer understanding of these sectors. We are also encouraged by Tongji_Software to set up popular science web pages and make popular science mini-games.During the process of developing the project and creating the wiki, the two teams will share some resources, such as front-end knowledge of the web, art vector material websites and etc.

In September-Octorber

SJTU-Software←→Tongji_Software

During the process of developing the project and creating the wiki, the two teams will share some resources, such as front-end knowledge of the web, art vector material websites and etc.

Summary

The two teams began to contact each other from the time of topic selection and brainstorming. During the preparation period, the communication between the two teams was not only about sharing resources, but also about having a clear understanding of each other's projects and making improvements to the core of the projects under the advice of the other team.

Finally, we are very grateful to IGEM for organizing this international event, which not only enables the two teams to get to know each other, but also enables the members of the two teams to gain such a rare friendship.

https://2021.igem.org/Team:Tongji_Software/Partnership

More Metails

SJTU-Software & SJTU-BioX-Shanghai

After deciding the theme of our project, which is detecting certain microRNA biomarkers in the serum with DNA computation methods, we turned to a wide range of researchers and peer students for suggestion.

Beginning

Later, when communicating with SJTU-BioX-Shanghai, we were surprised to find that they were also conducting their project on the detection. And, more importantly, our main bodies both present as a platform.

Furthermore, they thought this DNA computation methods could not only be used in screening miRNA markers, but also it is promising to detect all other kinds of nucleic acids. This view dramatically broadened our horizons for the project itself. Therefore, we then had subsequent interviews with our collaborative laboratory and informed professors. It was really stirring that they all agreed this can be achieved with deeper efforts and innovation. Meanwhile, we also suggested that their platform could be applied to detect some more molecular besides specific toxic molecular.

Complementation

Interestingly, the detection platform provided by SJTU-BioX-Shanghai is able to test any molecular except nucleic acids, while we can do the exact opposite task. So, in the following process, we are having continuous cooperation and discussion, trying to make supplement knowledge to the project of each other.

Improvement

In a formal meeting, we exchanged our ideas and made evaluations in detail. When SJTU-BioX-Shanghai knew that we were preparing to develop a functional tool of secondary structure prediction for nucleic acids, they proposed that this could be used and, meanwhile tested in their lab experiment. Since at that time our main model was constructed on the DNA sequence, we were advised to take some RNA sequence as the training set to broaden the range of software application, as well as help them in improving the structural designing. With the lab result, part of our model’ effectiveness could also be validated.

Education

In addition, to promote the significant idea of detection platforms, we cooperated with SJTU-BioX-Shanghai to hold an educational lecture for students who were interested in this area. In the meantime, well-selected lessons related to synthetic biology and early cancer diagnosis were taught during the lecture, with students actively engaged in the activity. In later investigations, we were excited to find that students gave high praise to our lecture, from which they gained a lot precious knowledge. Moreover, they proposed their own thoughts for our projects, such as the sensitivity of the detection platform, which was very crucial feedback for us to consider and improve.

In the summer vacation, together with SJTU-BioX-Shanghai, we went to Guangxi province, southwest of China, where the education resources were very limited. As for children there, we carefully designed classes with synthetic biology knowledge, to help them learn about this beautiful field.

Verification

In the final testing part of our projects, members of SJTU-BioX-Shang used our software to make further predictions for their designed RNA sequences, which helped them to exclude those with high-impact secondary structures. Then, they made two more short-sequence DNA probes to implement TMSD design, validating our model in a deeper level. Accordingly, the introduction of TMSD design is expected to reduce the detection time from 30 minutes to less than 10 minutes, while reducing the production cost of each strip, and further adapt to practical application requirements.

https://2021.igem.org/Team:SJTU-BioX-Shanghai/Partnership

More Details

Address

NO. 800 Dongchuan Road, Minhang District, Shanghai, China

Software
Welcome to use our IntelDNA

SJTU-Software 2021

Contact Us

rosesilver@sjtu.edu.cn

Related links
Shanghai Jiao Tong University