Professor Interviews

In search of inspiration and to understand the views of professors from different fields on AI in biology, we first visited Professor Tang Aihui. TangLab is mainly engaged in the development of super-resolution imaging technology based on single-molecule imaging and spatial transcriptomics technology, and its application research in neurobiology. Their work involves a large amount of information and data, and understanding Professor Tang's work is of great significance to us in how to apply AI effectively. In our interview, Professor Tang emphasized the importance of improving analytical skills and sensitivity to data. In the era of big data, especially in the field of biology, we need not only technical operation skills but also the ability to deeply analyze data and extract meaningful information from it. In addition, Professor Tang stressed the importance of cooperation between laboratories. Each laboratory has its own area of expertise, and understanding each other's work and generating cooperation is very important.

After deep reflection, we realized the importance of AI in the field of biology, but we couldn't come to a conclusion on how AI affects and changes biology. Therefore, we wanted to find a field that has been most affected by AI in recent years and listen to the opinions of mentors.

Through our research, we found that the emergence of Alphafold2 and 3 has had a significant impact on the field of structural biology. Therefore, we found Professor Sun Linfeng. SunLab uses structural biology (mainly cryo-electron microscopy and X-ray crystallography) and biochemical methods to study the molecular mechanisms of key membrane proteins related to major diseases in cells, and the structure and function of important biomolecular machines in cells, and has published many excellent works in recent years. Professor Sun mentioned that the advent of Alphafold is actually a great promotion for those who do structure. Alphafold can provide a lot of preliminary information, accelerating the process of solving structures. In addition, Alphafold3 can make quite accurate predictions of protein-protein complexes or protein-DNA, RNA complexes, providing a lot of interaction information. In the era of big data, the mining of information, whether in the field of structural biology or in other aspects, is a great promotion.

Finally, we found Professor Duan Yi. DuanLab mainly focuses on the impact and molecular mechanisms of gut microbiota on metabolic and immune diseases, especially chronic liver disease, and microbial-related biological therapies, especially phage therapy. Professor Duan is engaged in pathogenic microbiology, and synthetic biology involves a lot of microorganisms. If these microorganisms leak or are illegally used by people, it will have immeasurable consequences. Professor Duan particularly emphasized the importance of scientific ethics review, believing that as a scientific worker, we are essentially "people" in society; and apart from science, as a "person," ethics are very important. If we violate some ethics, it is definitely wrong - science should be explored within the framework of ethics.

In addition, these teachers gave us a lot of other valuable opinions. Teachers emphasized the importance of active learning and thinking, the training of scientific research thinking, and the cultivation of experimental skills and data analysis ability. They advocate maintaining enthusiasm and curiosity for scientific research, recognizing the importance of ethics in scientific research, and encouraging interdisciplinary cooperation. At the same time, teachers remind students to be patient and persistent, because repetition and drudgery in the scientific research process are the norm, and major discoveries often require long-term accumulation. In addition, they also discussed the different pressures and degrees of freedom in academic and industrial careers after graduation, suggesting that students make choices based on their interests and suitability. Teachers also mentioned the importance of scientific research ethics and the importance of learning from failure and moving forward. These suggestions provide undergraduates with a comprehensive perspective to make wiser choices in scientific research and future career development. Teachers analyze some of the issues that students are concerned about from a professional perspective, "listening to a word from a gentleman is better than reading ten years of books." We have organized these interview materials into WeChat posts, hoping to help more undergraduates.

Literature Sharing

In an effort to familiarize more people with biology, we have created a WeChat public account as a platform to disseminate information about biology to the public. Relying on this public platform, we have organized some interesting literature we encountered in our research topics into easy-to-understand articles, allowing the public to understand biology and get closer to it through reading our literature summaries.

Our Wechat Public Account: BioByte Brigade

Exploration of Biosafety

After interviewing the above teachers, we have benefited a lot. At the same time, we are also thinking about what AI can bring to synthetic biology, or what we can use AI to bring to synthetic biology. On a very honored occasion, we invited Senior Bao Yu Han from Tsinghua University, who is the person in charge of the iGEM Human Practice project, mainly engaged in the study of emerging technology risk governance. We discussed in depth the issues of synthetic biology ethics and biosafety, and then we have been thinking about whether we can use AI to promote the governance of synthetic biology biosafety, just like the role of Alphafold in promoting structural biology.

In the end, we focused on the suicide switch. In the field of synthetic biology, researchers often use or design new biological systems. These biological systems may contain genetically engineered microorganisms, which are safe in the laboratory environment, but if accidentally released into the natural environment, they may pose unknown risks to the ecosystem. The suicide switch can be used as a safety measure to ensure that these organisms can self-destruct under certain conditions, thereby reducing the potential threat to the environment. Although we ultimately failed - due to lack of experience and guidance, we were unable to successfully apply AI to the design of the suicide switch. However, after our research and organization, we have come up with some content about the suicide switch, which we believe is very inspiring for other teams, so we have placed it on our wiki.

Ethics and Responsibility

During our participation in the 2024 CCiC, we, along with all software teams, jointly signed a white paper titled "Biology Safety with Artificial Intelligence". This white paper outlines the shared principles of ensuring safety, security, and ethical conduct in the development and deployment of synthetic biology technologies integrated with artificial intelligence. The collaboration reflects our commitment to addressing the ethical challenges that arise from the intersection of AI and biosafety, emphasizing the importance of responsible innovation.

Biosafety with AI Whitepaper Signing CCiC2024

The document addresses critical concerns, including minimizing environmental impact, ensuring public safety, and promoting transparency and public engagement in synthetic biology projects. Our team remains dedicated to following these guidelines, ensuring that all our platform's functionalities and tools align with the highest standards of biosafety and bioethics.

This white paper represents a collaborative effort among the iGEM community, with a focus on fostering a culture of responsibility in the rapidly advancing field of synthetic biology. We believe that by adhering to these principles, we can contribute to the safe and ethical development of biological technologies.

Education

Introduction to iGEM

In addition to the above activities, we are also actively involved in education work. Our team member Wang Jinbo went to the High School Affiliated to Northeast Normal University to explain the knowledge of the IGEM competition, allowing them to have a deeper understanding of the current disciplines, and lighting up the beacon of interdisciplinary knowledge in their hearts;

Synthetic Biology Promotion

Our team member Lin Xiaohui also returned to his alma mater - Zhuhai No.1 Middle School, and took the opportunity of free communication to introduce IGEM and synthetic biology to the younger students. By telling the content of the IGEM competition, the application and prospects of synthetic biology, they felt the charm and subtlety of synthetic biology and stimulated their enthusiasm for genetic engineering.

Development of PCET

In addition, we also actively fulfill social responsibilities. In our social research, we found that undergraduates in biological-related majors are very troubled by physical chemistry experiments - because there is a lot of data to be processed, and the mathematical foundation of students in the School of Life Sciences is relatively weak. Our team happens to have students from various backgrounds. We leverage the advantages of interdisciplinary, and after sufficient demand surveys, we have developed a physical chemistry experiment tool (PCET). The content covers all the data processing tools required for wet experiments by students in the School of Life Sciences. Students only need to submit the data collected in wet experiments, and PCET will automatically organize and output the results, which is very popular with students.

Communication

Participation in CCiC

On July 11th, we participated in the annual CCiC meeting, where we had the opportunity to present our Mo-BASE platform and its integrated single-cell RNA sequencing (scRNA-seq) workflow. After hearing presentations from other iGEM teams, we showcased our project through three main lenses: the development background, workflow design, and future development plans.

Our platform aims to provide a comprehensive and user-friendly scRNA-seq analysis workflow, offering tools for preprocessing, dimensionality reduction, clustering, and visualization. By leveraging open-source algorithms and integrating advanced data analysis pipelines, Mo-BASE stands out as a robust bioinformatics platform. The platform also supports real-time data visualization and batch effect correction, ensuring reliable results for research applications.

During our presentation, we introduced the modular architecture of Mo-BASE, designed for ease of use and scalability. We emphasized how our platform can seamlessly integrate with external resources and databases, offering researchers a flexible environment for conducting advanced bioinformatics analysis. Our future plans include expanding the workflow to cover additional omics data and enhancing the interactivity of the user interface.

After our introduction, we engaged in a fruitful discussion with other teams, exchanging ideas and feedback. We received insightful questions and suggestions that helped us identify areas for improvement, particularly in refining the user experience and expanding the platform’s capabilities. This exchange of ideas provided valuable input for making our project even more robust.

Collaboration with Wet Laboratory

To assist researchers in optimizing their scRNA-seq analysis pipeline, we integrated various preprocessing tools into our Mo-BASE platform. Initially, we focused on commonly used methods for filtering, normalization, and clustering, referencing established workflows in the field. However, after internal testing, we found that these standard methods, while effective in general cases, lacked flexibility and adaptability for certain datasets, especially when dealing with complex or heterogeneous samples.

Through collaborative discussions, we learned that the USTC wet lab team was developing novel single-cell analysis methods that aimed to improve batch effect correction and more accurately cluster rare cell populations. Their approach leverages advanced machine learning techniques, such as deep learning-based dimensionality reduction and unsupervised clustering algorithms, which show promising results in enhancing the robustness of scRNA-seq data analysis. After several productive meetings, both our teams decided to co-develop and integrate these methods into Mo-BASE.

After extensive testing across various single-cell datasets, we believe that the new algorithm is more adaptable and can handle a wider range of challenges, such as batch effects and the detection of rare cell types, compared to standard preprocessing tools. This makes our platform a more versatile solution for scRNA-seq analysis.

We acknowledge, however, that there are still areas for improvement in our current tools. We are actively seeking collaborations with other teams to further refine and enhance the accuracy and reliability of our scRNA-seq workflow.

Test-run

Mo-BASE Platform Usage Experience

After setting up the Mo-BASE platform, we invited students and researchers from relevant fields to test our platform and provide feedback on its usability and features.

From the users' perspective, their valuable insights encompassed both positive feedback and constructive suggestions, helping us improve the Mo-BASE platform.

Praise:

• "The Mo-BASE platform fosters a great environment for bioinformatics discussions. Its user interface is clean, and the resources provided are highly relevant to my coursework." — Xiaoming Li, Junior, Department of Bioinformatics, Fudan University
• "The scRNA-seq workflow is well-structured and easy to follow. I especially appreciate the real-time data visualization tools integrated into the platform." — Mei Zhang, Graduate Student, School of Life Sciences, Zhejiang University
• "Mo-BASE stands out as a fantastic tool for both learning and research. The community interaction aspect is well-designed, allowing me to share and access teaching resources seamlessly." — Chen Wang, Researcher, Beijing Institute of Genomics

Suggestions:

• "While the platform offers a wide array of bioinformatics tools, it could benefit from more detailed tutorials on how to use the scRNA-seq analysis workflow." — Xiaoming Li, Junior, Department of Bioinformatics, Fudan University
• "The image segmentation tool works well, but I would suggest adding a feature that allows for side-by-side comparison of pre- and post-segmentation images for better accuracy." — Mei Zhang, Graduate Student, School of Life Sciences, Zhejiang University
• "The course sharing feature is excellent, but it would be helpful if there were a way to categorize resources by difficulty level or topic for easier navigation." — Chen Wang, Researcher, Beijing Institute of Genomics

Summary and Outlook:

We are thrilled to see that Mo-BASE has successfully aligned with our vision of creating an interactive community for bioinformatics learning and collaboration. However, through feedback from real users, we have identified areas for improvement, particularly in enhancing user guidance for advanced workflows and refining tool usability.

In the future, we plan to expand the platform's features and streamline the user experience. We are committed to addressing the feedback provided and will continue to evolve Mo-BASE into an even more effective resource for students and researchers alike.

We encourage other research teams and institutions to explore Mo-BASE, and we eagerly await your feedback and contributions to our growing bioinformatics community!

Contact us: USTC_Software2024@163.com