Team:NCTU Formosa/Integrated

Professor Lan is an expert in the field of biochemistry at NCTU. After joining up with professor Lan and introduced our project, he helpfully suggested we can focus more on the relationship between functional groups in chemicals and their contribution to mutagenicity in our dry lab work to support our biosensor. Taking his advice, we constructed a mutagenicity prediction AI based on chemical structures and mutagenicity. It is a massive step for our project! With our AI, people can now get mutagenicity with a few clicks online.

Figure 1: Meeting up with Professor Lan at NCTU.

AI and IoT Engineer Mr. Iblis Lin

After meeting up with Professor Lan, we initiated the construction of our mutagenicity prediction AI. However, we quickly met many obstacles while building it. Therefore, we visit Mr. Iblis Lin, an artificial intelligence and IoT engineer at NCTU. With his help, we started rebuilding our AI by preprocessing our data with 67 features extracted. As a result, our AI has a massive improvement in accuracy and trustworthiness. Furthermore, Mr. Lin also introduced us to IoTtalk, an easy using IoT user interface. It was Mr. Lin that we were able to integrate every part in our project on IoTtalk.

Figure 2: Working with Mr. Iblis Lin.

Toxicity Analyst Ms. Zhao, Shu-Ru

To further improve our project, we wanted to know more about the recent method of mutagen detecting. Therefore we head to the Environmental Technology Smart System Research Center to visit a particular person, Postdoctoral Researcher Zhao, Shu-Ru. During our interview, Ms. Zhao shared many of her experiences working as a toxicity analyst. After listening to our project, she credited our idea of presenting mutagenicity with an easy-understandable racecar game.

Moreover, she gave us some critical feedback. Pointing out that we should filter our training data in our mutagenicity prediction AI. Because the data we use were based on Ames test results, which turns out having a high false-positive rate. Therefore, we should better pick data that are verified with more precise methods such as mammalian cell HPRT gene mutation assay.

Besides, Ms. Zhoa also proposed we building up an all-in-one device for an analyst like her holding measurements more efficiently. Further, we discussed many of our devices in detail. As a result, we double confirmed and filtered our data, which then lead to an apparent increase in our mutagenicity prediction model. Also, we adjusted our device design, making it more user-friendly for users.

Figure 3: Photo with Ms. Zhao, Shu-Ru at Environmental Technology Smart System Research Center.

Industry Insider Mr. George Lin

To understand more about the biotechnology industry, we visited Mr. George Lin, deputy general manager of NanoSigma Biotech.Corp. Mr. Lin taught us about how mutation tests were held now in the food industry to secure food safety and suggest as to belt up standard procedures for our system. Moreover, Mr. Lin shared his experiences starting up a company, and how he introduced products to the market. With the guidance of Mr. Lin, we reconstructed our E. Phoenix system with a much robust structure and made it much acceptable for the market.

The Society of Wilderness

Besides experts, we are also desperate to know how the public and society will think about our project. Thus, we arranged a meeting with the Society of Wilderness, civil environmental society as well as frontiers in promoting sustainable development to the general public. After introducing our project to workers, volunteers, and people, we have received much feedback for our system. Especially in our biovisualize system. They recommended that we can show examples in our grading system, which could make it more understandable. Together, we enjoyed a great afternoon. We were all very exhilarated, meeting these passionate volunteers and NGO workers. What's more? They even invite us to their environmental protection activities.

Figure 4: Photo with Mr. XU, TIAN-LIN, president of The Society of Wilderness.