Integrated
human practices

SJTU-Software participated in the online exchange event titled “Global Green Movement: Discussion on Sustainable Development among Youth Leaders from the P5 Countries in International Organizations” organized by IOING CHINA. According to reports from institutions such as the United Nations Environment Programme and the United Nations Commission on Sustainable Development, the Earth's ecological environment faces multiple threats and challenges, including global climate change and the energy crisis. In the discussions, Mr. Wang Zhijia emphasized that the youth can play a vital role in raising public environmental awareness and forming a powerful network to collectively work towards sustainable development.

Specifically, we concern about reducing the negative environmental impact caused by chemical pesticides use in agriculture, which exerts significant pressure on ecosystems. Nowadays, we are faced with serious pest infestations and pesticide residues at the same time. Thus it is critical to seek out a high-efficient pesticide to put into use.

Spinosad is a new type of microbial insecticide with contact and ingestion toxicity. It can effectively control a variety of pests, with very little dosage and long duration of effect, when the concentration of 1mg/L can protect the grain for at least 4 months, it is considered a promising "green" insecticide and protective agent. In 2005, the United States Environmental Protection Agency approved spinosad as a protective agent for stored grain. In 2008, the European Union approved the use of spinosad on organic crops, and spinosad has become the most promising biological insecticide in the world.

Spinosad has been applied to more than 200 different crops. It has been used to control caterpillars in cotton, leafminers in various crops, leaf-rollers in apples, etc. However, due to factors such as low unit yield, many by-products, complex extraction steps, and low product yield, the production cost is high, which restricts the popularization and application of spinosad. The spinosad products and APIs sold on the market are basically the United States products of Dow Yinong Company, and the production of domestic spinosad APIs cannot compete with them.

Therefore, we came up with the improvement of the Spinosad production, which is a biological insecticide produced through the fermentation of Streptomyces spinosa. At the same time, we are dedicated to increasing public awareness of environmental protection, encouraging more people to take action in preserving the environment.

Professor Ting Wei

Enzymes play a central role in biochemical processes as biocatalysts. They function in aqueous solutions at mild temperature and pH conditions and are highly substrate-specific. Enzymes bind to substrates through their active sites and catalyze chemical reactions that are fundamental to energy conversion and substance metabolism in living organisms.

When predicting the type of reaction between enzymes and substrates, we can apply machine learning models to process and learn from these huge data sets. Through iterative optimization, we are able to gradually improve the prediction accuracy of the model until the model performance reaches the optimal level or completes the predetermined number of iterations.

According to the Prof.Ting Wei, we know that we can have access to a large amount of high-quality data at our disposal when it comes to predicting the type of reaction between enzyme and substrate, covering a wide range of enzyme-substrate reactions. Our datasets include information about the structure and sequence of enzymes, as well as information about the structure of small molecules. The advantage of such data is that they are not only large in quantity and high quality, but they are also well suited for validation and iterative optimization through wet experiments, and at the same time, we can experiment with a variety of modeling methods. After experimental verification and evaluation, we can further optimize the model in a diversified manner.

Professor Yan Feng

Prof. Yan Feng mentioned that in molecular evolution, convergent evolution is a fascinating phenomenon in which proteins derived from the same ancestor may have two distinct protein functions that may be high in sequence homology and similarity, but develop completely different structures and functions. This phenomenon not only sheds light on the richness of biodiversity, but also provides us with an important perspective on how biomolecules adapt and evolve.

From a propositional point of view, Prof.Yan Feng suggested that we can start from existing molecules and develop higher quality or more goal-oriented molecules through the process of evolution. For example, Spinosad is a natural product produced by soil microorganisms, and it is an insect control product with a unique structure. Through computational tools, particularly artificial intelligence (AI)-based quantitative structure-activity relationships (QSAR) and computer-aided modeling and design (CAMD), scientists have been able to explore and exploit Spinosad and have discovered a range of synthetic Spinosad simulants. These synthetic Spinosad simulants have shown potential as insecticides against lepidopteran insect pests, and their effectiveness has been shown to be comparable to that of Spinosad and Spinetoram in field studies.

Now, the challenge is to design a better system to facilitate this type of molecular evolution. This may involve more advanced computational models to predict how molecules evolve through mutations and natural selection processes, or the development of new experimental techniques to simulate and accelerate this evolutionary process in the laboratory. In this way, we can better understand and harness the power of molecular evolution to create novel molecules with specific functions that may have a wide range of applications in medicine, agriculture, and industry.

When designing such a system, we need to take into account the diversity of protein functions and the expansion of protein sequence space. Using deep learning techniques such as generative adversarial networks (GANs), we can explore and generate protein sequences with novel functions. This approach not only helps us understand how proteins acquire new functions through sequence changes, but also guides us in designing novel proteins with specific properties.

In conclusion, by combining computational modeling, laboratory evolutionary experiments, and deep learning techniques, we are expected to develop new systems to more effectively explore and harness the potential of molecular evolution. This will provide us with new tools and strategies to address the many challenges facing the world today, including disease treatment, crop protection, and environmental management.

The teachers suggested that we should be in line with social issues. Prof. Xiaoli Xue also provided suggestions on the selection of our project objectives- it is best to choose the one that has data support from the college or has more data from major databases, and the actinomycete research of our college has a certain accumulation, which is of great help to our project determination.

Professor Linquan Bai

Prof. Linquan Bai told us that the purity of spinosad is not high, and the purity and unit yield need to be improved.

At the same time, he mentioned that improving the production efficiency of spinosad is of far-reaching significance for agricultural development. First, it reduces the cost of biopesticide production, making this environmentally friendly pest control method affordable to more farmers, thereby reducing reliance on chemical pesticides. Secondly, with the improvement of production efficiency, the availability of spinosad will increase, which will help to promote the use of biopesticides more widely, which will not only help protect the agricultural ecological environment and reduce chemical residues, but also improve the quality and safety of agricultural products and meet consumer demand for healthy food.

In addition, increased productivity means faster response to pest outbreaks, reduced crop losses and food security. In the long run, this will promote the sustainable development of agriculture and contribute to the stability of agricultural ecosystems and the conservation of biodiversity. Ultimately, this will create a virtuous circle that will improve the overall efficiency and sustainability of agricultural production, bringing economic benefits to farmers and environmental benefits to society.

After comprehensive consideration, as spinosad is a biological insecticide produced through the fermentation of Streptomyces, but remains expensive and challenging to promote widely, our team are committed to make some changes for this situation.

Professor Yue Zhang

Now we are devoted in optimizing spinosad’s yeild. But we haven’t worked out a best way to manage this project. Through consultation before, we are considering build up an AI model to simulate the whole producing process. Thus, we turned to Prof. Yue Zhang for further advice. Prof. Yue Zhang suggestion for our project is that the metabolic pathway models of many organisms have not been studied in detail, and most of them are limited to specific model organisms, whether it is the establishment of a metabolic model of a specific organism or a model with wide applicability, it is very innovative work, so we will focus on the production of spinosad bacteria and hope that our model can be adapted to a variety of organisms through adjustment.

At the same time, he suggested that the needs of target users should be considered in the AI design process, and the interpretability factor should be considered in AI design, so as to help us understand the deeper biological mechanism under the AI model.

Prof.Yan Feng suggested that when we select specific enzymes for directed evolution, we can think beyond the single compound synthesis pathway, and judge the enzyme system that has an important impact on the production of spinosad from the metabolic network of the whole organism.

Finally, we again sought guidance from Prof. Linquan Bai as he is the professor in actinomycetes, and spinosad is the metabolite of actinomycetes. Prof. Linquan Bai pointed out that the properties of biological enzymes are not limited to Km and kcat related to chemical catalysis, and he suggested that we can explore the correlation between the optimal temperature of enzymes or their suitable salt ion environment, pH value and other physical and chemical environments and enzyme sequences and structures.

Therefore, our project aims to enhance the production efficiency of spinosad by optimizing the metabolic pathways within Streptomyces spinosa. One significant factor affecting its yield is the discrepancy between industrial fermentation temperatures and the optimal temperatures for the enzymes involved in metabolism. Our project intends to develop a deep learning model to predict the optimal temperatures for enzymes within Streptomyces spinosa and employs ecGEM to identify key enzymes that impact yield, thereby pinpointing those restricted by temperature factors.

For these key enzymes, the project will use the model to propose optimization strategies and validate these through experimental verification.

Qilu Pharmaceutical Group is a large-scale comprehensive modern pharmaceutical enterprise in China. Qilu Pharmaceutical (Inner Mongolia) Co., Ltd. Hulunbuir Branch Arongqi Base Biopharmaceutical Industry Project has a total investment of 10 billion yuan to build high-end biological pesticides, biological veterinary drugs, high-end human drugs and other products and their research and development bases. It will be built in three phases. The first phase of the project invested 3 billion yuan, mainly to build 6 agricultural and veterinary drug production lines, and started construction in March 2019. Among them, the annual production capacity of spinomycin is 600 tons. They pointed the problems found in actual production that the reasons for the low purity were high optimal temperature of enzymes and low catalytic efficiency.

After the representatives of the company's R&D department Mr. Dang learned about our idea, they affirmed our model to a certain extent, and proposed to us the key parameters including pH value, dissolved oxygen, carbon and nitrogen source, trace elements and inorganic salts, etc., and they suggested that we can infer the cause from effect in the research process, and explore the mechanism after discovering a certain factor to increase the yield. At the same time, they also suggested that the SJTU-Software team should apply AI technology in the project, focusing their research on the optimization of optimal temperature and fermentation conditions in the early stage

This year's project by the SJTU-Software team focused on spinosadin, a low-toxicity, high-efficiency, broad-spectrum insecticide, a fermentation product of spinosad. However, due to the difference between the optimal temperature of the key enzymes required in the production process of spinosad and the actual temperature of the fermentation process, and the lack of optimal temperature information of the key enzymes, the problems of low catalytic activity of key enzymes, the generation of by-products, and the low yield and purity of spinosad per unit of biosad have restricted the further promotion and application of spinosad insecticides. In response to this problem, the R&D department of Chenghua Biology put forward three solutions and suggestions: first, with the continuous addition of enzymes as the reaction progresses, the total enzyme activity is maintained; Second, according to the different optimal temperatures of key enzymes in different reaction processes, adjust the reaction temperature in different time periods; Third, mutations were made to mutate the key enzyme genes in Polyspora spinosa so that the optimal temperature was equal to the actual reaction temperature, so as to improve the catalytic efficiency. The SJTU-Software team members were inspired and hoped to further optimize the team's project based on Chenghua Biotech's suggestions for the project

To seek for solutions, we came to Chenghua Biology, who has done continuous research in the synthetic biology filed. In response to our problem, the R&D department of Chenghua Biology put forward three solutions and suggestions: first, with the continuous addition of enzymes as the reaction progresses, the total enzyme activity is maintained; Second, according to the different optimal temperatures of key enzymes in different reaction processes, adjust the reaction temperature in different time periods; Third, mutations were made to mutate the key enzyme genes in Polyspora spinosa so that the optimal temperature was equal to the actual reaction temperature, so as to improve the catalytic efficiency. The SJTU-Software team members were inspired and hoped to further optimize the team's project based on Chenghua Biotech's suggestions for the project.

The organism contains a large number of metabolic reactions and enzymes, which together support the regular metabolic activities of the organism. These reactions are interconnected, and altering any one of them may potentially affect the entire metabolism of the organism. Traditional wet lab methods rely on random mutations, which means that most mutations do not produce beneficial traits and may even be harmful. Therefore, it usually takes researchers long time to get an improved strain.

If we can model the metabolic reactions within the organism and predict the key enzymes involved in the synthesis of spinosad, its results can guide wet lab mutations, which can significantly enhance efficiency of optimization.

So we put forward a scientific solution that we aimed to further optimize the prediction model by collecting and analyzing relevant data. Specifically, our goal was to predict the optimal operating temperature for all enzymes involved in the synthesis of clavulanic acid by Streptomyces clavuligerus. Spinosad is an important β-lactamase inhibitor that is widely used in the synergistic treatment of antibiotics to enhance the efficacy of antibiotics.

When we talked with Prof. Yue Zhang for this idea, he recognized our thoughts and proposed that we would better first collect data on the various enzymes involved in the synthesis of spinosad in the process of polyspora spinosa, including their biochemical properties, active temperature range, etc. Subsequently, this data was used to optimize existing prediction models to more accurately predict the optimal temperature for each enzyme in the synthesis of spinosadin. This step may involve the application of machine learning algorithms, statistical analysis methods, or bioinformatics tools.

Next, we can compare the predicted results with actual industrial production conditions, focusing in particular on those reaction temperatures used in the actual production process. This comparison allows for the identification of enzymes that are optimally different from the reaction temperature in the actual application. These differences can affect the activity of the enzyme and the efficiency of the entire synthesis process.

Finally, Prof. Yue Zhang suggested that these differences should be analyzed in depth to determine whether the production conditions need to be adjusted, or whether these enzymes can be modified by genetic engineering to show higher activity and stability at the temperature of practical application. Such research would not only improve the efficiency of spinosad production, but could also provide valuable experience and insights into other biosynthetic processes.

Professor Hongzhong Lu

Genome-scale metabolic models (GEMs, GSMM) is a mathematical model constructed based on the genomic information, transcriptomics, proteomics, and other data of a specific organism. It integrates all known metabolic reactions and pathways, enabling the simulation and prediction of the complex metabolic networks within an organism.

However, after literature investigation, we found that there is currently no available GEMs model for Saccharopolyspora spinosa, which produces spinosad. Considering that building a GEMs model from scratch would be very time-consuming, we decided to seek solutions from our professor.

Fortunately, when we discussed with Prof. Hongzhong Lu, the expert in digital modeling and characterization of cell metabolic networks, we learned that he is also conducting research on spinosad. After communication, he offered us an initial GEMs model.

Prof. Hongzhong Lu presented a research topic on the latest genome-scale metabolic model (GEM) of Hypoglycosaperos. GEM is a powerful computational tool that simulates and analyzes the metabolic network of organisms by integrating genomic data, metabolic pathways, and biochemical reaction information. This model not only helps us understand the metabolic function of organisms and the complexity of metabolic networks, but can also be used to predict gene knockout effects, metabolic engineering, drug development, and other applications.

In the study of Hypoglycosaperos, the latest GEM model may contain a detailed reconstruction of this biological metabolic pathway, which may involve an in-depth analysis of its metabolic activity under different environmental conditions. Such a model could reveal key metabolic nodes that may be critical for the growth and survival of Hypoglycosaperos. With this model, researchers can identify potential drug targets or devise genetic engineering strategies that can manipulate their metabolic processes.

In addition, GEM models can also be used to study the role of Hypoglycosaperos in specific ecological niches and how it interacts with other organisms. For example, if Hypoglycosaperos is a microorganism associated with human health, its GEM model may help to understand its role in disease development or how certain diseases can be treated by regulating its metabolic activity.

In conclusion, Hongzhong Lu's research theme highlights the potential of GEM models in understanding microbial metabolism and developing new therapeutic strategies. By constructing and analyzing the GEM of Hypoglycosaperos, we can better understand the biology of this organism and explore its applications in medicine and industry.

Traditional GEMs only consider stoichiometric and reaction constraints, which cannot accurately simulate cellular metabolic responses under different conditions. By incorporating enzyme kinetics information into GEMs, we can further constrain reaction boundaries. Therefore, we can construct an ecGEMs for Saccharopolyspora spinosa, which includes enzyme-constrained reaction flux matrices that account for enzyme abundance and activity.

Prof. Yue Zhang's point highlights the importance of introducing enzyme concentration limits when constructing enzyme-constrained genome-scale metabolic models (ecGEMs). This model can more accurately predict the metabolic behavior of cells under different conditions, including growth rate and carbon source utilization. By integrating enzyme kinetic information and thermodynamic information, ecGEM can simulate the total amount of enzymes within the cell, allowing for more realistic predictions of metabolic flux distribution.

When constructing ecGEM, researchers often consider the spatial congestion of the enzyme in the cell, the catalytic activity, and the thermodynamic properties of the metabolic reaction. For example, with the FBAwMC method, the concentration of the enzyme can be limited based on the principle of macromolecular crowding, allowing for a more accurate simulation of the molar volume of the enzyme in the cytoplasm as a function of the molar number. In addition, the GECKO method takes into account the average saturation of enzymes, which provides a new way to construct enzyme confinement models.

The applications of ecGEM are not limited to theoretical research, they can also guide practical metabolic engineering modifications. For example, with the enzyme concentration distribution predicted by ecGEM, researchers can identify key rate-limiting enzymes and design experimental strategies accordingly, such as optimizing metabolic pathways through overexpression or enzyme modification, to improve the yield of the target product. In addition, ecGEM can predict the impact of environmental disturbances, such as temperature changes, on cellular metabolic behavior to guide process control and strain optimization.

Recent research progress shows that ecGEM has significant advantages in simulating the growth of microorganisms under a single carbon source, studying metabolic overflow phenomena, and guiding metabolic engineering transformation. For example, the ecGEM of Saccharomyces cerevisiae constructed by the GECKO method was able to successfully predict its maximum growth rate under different carbon sources, and it was in good agreement with the experimental data. These studies show that ecGEM has important potential for both basic biology research and industrial applications.

In living organisms, even slight temperature changes can have fatal effects on microorganisms. However, in GEMs and ecGEMs, the models can only simulate conditions at a single temperature. Since temperature is crucial for Saccharopolyspora spinosa, we plan to incorporate thermodynamic constraints into the model, constructing etcGEMs to simulate the metabolic conditions of the organism at different temperatures.

Prof. Yue Zhang proposed an innovative concept to construct enzyme and temperature-constrained genome-scale metabolic models (etcGEMs) based on genome-scale metabolic models (GEMs) and enzyme-constrained genome-scale metabolic models (ecGEMs). Such models can simulate the metabolic state of microorganisms at different temperatures, and for temperature-sensitive microorganisms such as Saccharopolyspora spinosa, etcGEMs can provide more accurate predictions of metabolic behavior.

In existing GEMs and ecGEMs, models generally assume that all metabolic reactions take place at a constant temperature, but in practice, temperature has a significant effect on the metabolic activity of microorganisms. Changes in temperature can affect the activity of enzymes, the stability of proteins, and the behavior of other biomolecules within the cell. Therefore, incorporating temperature as a constraint into the model can more realistically reflect the physiological state of microorganisms at different ambient temperatures.

To construct etcGEMs, he suggested that we collect data on the activity of various enzymes in microorganisms at different temperatures, including parameters such as the melting temperature (Tm), optimal temperature (Topt), and heat capacity change (ΔCp) of the enzymes. These parameters can be obtained through experimental measurements or literature studies. Then, by integrating these parameters into the metabolic network model, it is possible to simulate which metabolic pathways will be activated at different temperatures, which enzymes will be inhibited, and how these changes affect metabolic flow throughout the cell.

In addition, he told us that etcGEMs can also help us understand how microbes adjust their metabolic strategies at different temperatures to maintain growth and survival. For example, by simulating metabolic flux at high or low temperatures, researchers can identify key metabolic pathways and enzymes that may be affected at extreme temperatures, thus providing a rationale for microbial strain improvement in industrial applications.

Enzyme optimal temperature, a fundamental property, has long been a focus of enzyme research. However, even yeast, as a chassis cell, lacks optimal temperature data for more than a quarter of its metabolic network. This is because experimental measurement of optimal temperatures remains complex, and existing databases provide limited information on key enzymes.

After we discussed the problem with Prof. Chaochun Wei, we thought that AI might help to solve the problem, since the data that existed can be used to train our model.

We held the view that developing a deep learning model to predict the optimal temperature of the enzyme in the body of Polyspora spinosa may work. Therefore, we tended to Prof. Chaochun Wei, expert in bioinformatics. He suggested that we collect data and optimize the model to predict the optimal temperature of all enzymes in the reaction pathway for the synthesis of spinosad in Polyspora spinosa, and compare it with the actual reaction temperature to find which enzymes have a large difference between the optimal temperature and the actual reaction temperature.

Our project aims to improve the efficiency of the production of spinosad from Polyspora spinosa, which is an important ingredient in biopesticides and has significant effects and environmental friendliness in pest control, thereby reducing production costs and promoting the widespread use of this highly effective, safe, and easily degradable bioinsecticide.

In this project, by collecting and analyzing the relevant data of Polyspora spinosa, we developed a deep learning model to predict the optimal temperature of each enzyme in the process of spinosad synthesis in vivo, and compared it with the actual industrial fermentation temperature to find the enzyme with a large difference between the optimal temperature and the actual temperature. It is the fact that the activity of these enzymes is not fully realized during industrial fermentation that results in a decrease in yield.

In order to further optimize, we added enzyme constraints to the existing GEM model and constructed the ecGEM model to identify the key enzymes that were both limited by temperature and affected by yield by analyzing the rate-limiting enzymes in the metabolic pathway.

For these key enzymes, we will combine the prediction results of the deep learning model to propose an optimization strategy and conduct directed evolution experiments to make the optimal temperature close to the actual fermentation temperature, so as to improve their catalytic efficiency.

Prof. Bai Linquan gave affirmation of our ideas.

Our team went to Mengniu Dairy for more practical suggestions. We consulted about the problems we may met during the directly opimization of metabolic pathways through wet lab. Mengniu Diary shared real producing progress and insisted that this straight optimization is too difficult and troublesome to realize, which confirmed us that we need to use drt lab, the etcGEM, to find out key enzymes in the first step and mutate them through experiment later.

In the last meeting, the representative of Qilu Pharma proposed that the key parameters include pH, dissolved oxygen, carbon and nitrogen source, trace elements and inorganic salts in addition to temperature, and Mr. Dang suggested that our team should apply AI technology in the project to focus on the optimization of the optimal temperature and fermentation condition in the early stage.So our team put their ideas into practice on the model. After repeated attempts and modification, we got our AI model greatly optimized and achieved phased success.

Professor Qun Jiang

The competition project aims to improve the yield of polysaccharides produced by Ganoderma lucidum through synthetic biology techniques. By using the ecGEM model to identify enzymes that restrict the synthesis of multiple fungicides within the bacterial body, the enzymes with significant differences were modified to increase the production of active compounds, resulting in strain modification. The ethical issues involved in this are mainly biosafety.

We know that the problems caused by synthetic biology in the fields of agriculture and environment, in terms of biosafety, are mainly reflected in the possibility of synthetic biology creating completely new and unprecedented organisms and ecosystems. Due to the absence of such new species in existing ecological history, it may result in their inability to evolve or deviate from human expectations to evolve into other unknown species, making it difficult to estimate the risks of their escape or intentional or unintentional release. Furthermore, newly synthesized species are strong and harmful, and may replicate or evolve uncontrollably, causing a "crowding out effect" on other organisms. Some synthetic toxins used to resist specific pests may affect the survival of other invertebrates.

In 2010, after Venter announced the creation of the first synthetic life form, the US government released a research report titled "New Directions: Ethics of Synthetic Biology and Emerging Technologies," emphasizing the need for risk assessment before and after project implementation. This report also proposes five ethical principles for evaluating biotechnology: public good, responsible management, academic freedom and responsibility, democratic evaluation, fairness and impartiality.

Synthetic biology is also constrained by some international regulatory documents. The "Laboratory Biosafety Manual" published by the World Health Organization stipulates the protection of microbial resources from theft, loss, or transfer, and the avoidance of inappropriate use of microbial resources that threaten public health safety.

For this project, on the one hand, it is necessary to assess biosafety risks before implementation, and on the other hand, attention should be paid to protecting microbial resources during laboratory operations to prevent arbitrary transfer and unintentional release. Attention must be paid to the proper handling of any synthetic biomaterials.

Researchers should take the concept of "responsible innovation" as an important guiding principle for synthetic biology research and enhance their sense of social responsibility. Specifically, before conducting research, it is necessary to consider the potential social and environmental impacts of the project, and select the most urgent and beneficial topics to meet human needs. After the research is completed, it is necessary to choose whether to protect the invention. The iGEM competition requires not only the creativity and academic value of the project, but also the sense of responsibility of the participants. Participants are required to demonstrate the rationality of the project from the perspectives of law, safety, and ethics, and pay attention to the social benefits of the technology.

Synthetic biology utilizes engineering principles to manufacture complex life systems by designing and developing standardized biological components, methods, and tools. However, there are many safety issues with the development of standardized biological components:

At present, there is no safety level classification in standardized biological components, and toxic components cannot be identified, which poses a safety hazard. Even if some components themselves do not pose biosafety risks, they may have negative impacts during the assembly process. For example, biological switches that cannot predict whether the assembled organism contains toxins. In a novel gene circuit synthesized from standardized components, the lack of sufficient functional separation units can lead to unexpected interactions between components. The combination of components that have been proven to be safe may also lead to unsafe overall performance of the genetic circuit, and so far no method has been invented for conducting safety checks in the genetic circuit.

The urgent problem that needs to be solved now is whether it is necessary to classify components, parts, and life systems into different security levels. How to implement biosafety measures into the design process, so that the system can automatically alert designers of safety issues in their plans.

Our team discussed the social acceptance and safety of synthetic biology products with two PhDs from Mengniu. Dr. Huang first said that the products of synthetic biology have been around for a long time, and there are also a large number of experiences from developed countries that can be used as references in the formulation of relevant domestic regulations, and the approval process is complete and reliable. In addition, synthetic biology-based R&D products will also be less expensive, which can help increase their popularity in society. Dr. Zhao added that synthetic biology products can exclude allergens such as antibiotics and are safer than traditional dairy products for special populations such as lactose intolerance. Therefore, synthetic biology has great application potential in the fields of food, medicine, and chemical products.

In this roundtable discussion, we have delved into the widespread application of artificial intelligence (AI) in the field of biological sciences and the ethical challenges it brings. The conference has been divided into six main topics, each aimed at promoting a comprehensive understanding of the potential and risks of AI technology in biological sciences.

Firstly, we have discussed the application of AI in gene editing technology, especially CRISPR technology. This issue has analyzed how AI can accelerate the gene editing process, while exploring the ethical issues that arise from it, including genetic discrimination, the concept of designing babies, and potential threats to biodiversity.

Subsequently, the conference has shifted towards the application of AI in biomedical research, particularly in drug development and personalized medicine. We have discussed how to effectively utilize AI for research while protecting patient privacy and data security.

The third topic has already focused on the security of AI designed biological systems. We have evaluated the application of AI in synthetic biology, including the design of new biological systems and microorganisms, and discussed the safety and potential risks of these designs.

In the fourth topic, we have discussed the role of AI in public health monitoring and epidemic warning. The meeting has discussed the issue of transparency in data collection and how to maintain public health while protecting personal privacy.

The fifth topic has already focused on the application of AI technology in agricultural biosafety. We have analyzed the role of AI in crop genetic improvement and pest monitoring, as well as the potential impacts of these technologies on ecosystems and biodiversity.

Finally, the conference has discussed the ethical and legal issues of AI generated biological data. We have discussed the application of AI generated data, such as virtual experiment data, in scientific research, as well as the authenticity and intellectual property issues of these data.

We have thoroughly analyzed the ethical issues of AI assisted diagnosis and treatment, including its application in medical diagnosis and treatment, including image recognition and personalized treatment plans, and explored potential ethical issues such as algorithm bias and decision transparency.

This conference has gathered experts from different fields to discuss the future development of AI technology in biological sciences through interdisciplinary dialogue, and how we can ensure that the application of these technologies meets ethical standards while maximizing the benefits for human society.

We had an in-depth discussion about the background of our respective projects, including research goals, methodologies and expected outcomes. We exchanged valuable experiences and insights, which deepened our understanding of each other's projects.

We focused on discussing the Human Practice work, including how to effectively use social media for outreach, creative ways to promote science education and strategies for external collaboration and communication.

We discussed about the design of the Software, covering specific functionalities, the choice of system architecture and ways to enhance software stability and operational efficiency. We also exchanged experiences about the technical challenges faced during development and the solutions implemented.

We shared and presented our iGEM project with teams from various domestic universities, engaging in comprehensive discussions on project design, experimental methods, and practical applications. This exchange not only provided us with new ideas and insights but also fostered mutual learning and collaboration among the participating teams.