Overview ​

Industry ​

iHP Visit – Conghua Litchi Exposition Park ​

To gain a firsthand understanding of the full production logic of the lychee industry chain and to investigate currently commercialized preservation technologies, our team visited the Conghua Litchi Exposition Park in Guangzhou. The visit was hosted in collaboration with Hualong Fruit and Vegetable Preservation Co., a leading enterprise with authoritative expertise in lychee preservation technology and full-chain production management.

During the visit, we engaged in an in-depth discussion with the Hualong team, learning about the full lychee production chain: cultivation, harvesting, pre-cooling, preservation, transportation, and postharvest processing. We inquired about technical details such as optimal harvesting conditions, innovative postharvest “tricking” methods to delay spoilage, pre-cooling strategies (including cold storage and ice-water rapid cooling), the use of chemical and biological preservation agents, and packaging solutions. The team also introduced us to frozen “hibernation” preservation technology, shelf-life challenges, variety-specific transport damage issues, and cost considerations in bio-preservation. We toured their postharvest preservation laboratory, explored high-value processed lychee products, and observed AI-powered digital agriculture systems, including drones for real-time orchard monitoring, automated irrigation, and mechanized harvesting trials.

This visit served as an immersive industrial classroom and a valuable source of scientific inspiration for Lychee Guardians:

Systematic Understanding

We gained a comprehensive view of the lychee industry chain, along with its technological strengths and current bottlenecks.

Shift in Wet Lab Focus

The visit directly reshaped our experimental direction:

1. 1. Shelf-life as the key challenge – Staff emphasized that the largest losses occur during the retail shelf-life stage, while transportation is already well protected by modern cold-chain logistics. This shifted our research priority from transportation to shelf-life preservation.


2. 2. Inspiration from “leaf covering” – Harvesters shared that freshly picked lychees are traditionally covered with a layer of leaves to “trick” the fruit and slow deterioration. This inspired us to consider whether synthetic biology could achieve a similar effect by promoting the expression of endogenous protective substances. Specifically, this advanced our exploration of the melatonin pathway as a potential biological preservation mechanism.

iHP with Farmers – Nansha Feihong Picking Garden ​

To complement the macro-level understanding gained from the Conghua Litchi Exposition Park, our team visited the Nansha Feihong Picking Garden, a production base, to directly engage with farmers and managers. We aimed to investigate preservation challenges at the production and distribution stages, especially for small-scale growers who depend heavily on direct sales and e-commerce.

During our visit, farmers shared that their primary business model relies on on-site picking, direct sales, and online e-commerce. For e-commerce distribution, they relied on cold-chain express services with ice packs. While this method effectively preserved fruit quality during transport, shipping costs were extremely high—sometimes even higher than the value of the fruit itself. The garden manager further emphasized that farmers face significant risks, since spoilage losses during long-distance sales are borne entirely by the producers without compensation. We also observed packaging and logistics practices, such as the immediate boxing of freshly harvested lychees with ice packs, which matched the insights we had learned earlier from the Exposition Park.

From these interactions, we confirmed that transportation spoilage is already well managed by cold-chain systems, but the economic burden of logistics remains a serious challenge for farmers. These findings reinforced that extending shelf-life beyond the transport stage—especially at the retail end—would reduce farmer risks and support more sustainable profits. This validated our decision to focus our wet lab design on shelf-life preservation rather than transportation.

iHP with Retailers – Local Fruit Vendors ​

Considering that small lychee vendors lack relevant knowledge about lychee preservation methods, largely leading to the spoilage of lychees and food waste, Qiqi decided to design a brochure targeting the small lychee vendors, which aims to offer practical preservation guidelines while demonstrating synthetic biology’s potential in solving lychee spoilage.

It analyzes lychee spoilage causes from three key aspects: the fruit’s inherent spoilage traits, microbial infection, and environmental factors, helping distributors grasp the problem’s essence. For practical methods, it provides four simple, low-cost techniques, including low-temperature refrigeration, ice water soaking, breathable packaging, and timely sorting suited to small businesses, with clear steps and precautions to reduce losses.

The core highlight is our dual-path biological preservation solution. Based on Lingnan’s characteristics and cutting-edge synthetic biology, it modifies Bacillus subtilis on lychee surfaces: one engineered strain synthesizes plant-derived wax to form a protective layer; another secretes a small amount of melatonin to scavenge the fruit’s reactive oxygen species, delaying senescence.

This system is safe and eco-friendly, avoiding chemical preservative residues, meeting green consumption trends, and providing a new room-temperature preservation solution. We’re committed to integrating synthetic biology with local agriculture via this technology.

Early-Stage iHP – Hualong Fruit & Vegetable Preservation Co. ​

At the start of our project, we engaged with Guangzhou Conghua Hualong Fruit & Vegetable Preservation Co., Ltd. They recognized the significance of our shelf-life preservation topic and noted our wet lab direction was on the right track. While some of their suggestions (such as possible connections to other research groups like Sun Yat-sen University) shaped our network exploration, their insights confirmed that microbial spray-based preservation was seen as innovative. For the dry lab, they emphasized the complexity of predicting lychee state with only visual features due to variety differences, yet acknowledged our multi-variety classification + prediction approach as promising. On the media side, they encouraged active outreach and awareness-building.

iHP with Consumers – Lychee Consumption Habits and Preservation ​

Summary:

1. Pain points of preservation demand:

Consumers' understanding of the preservation period is limited (85% believe it is only 1-4 days), which is far from the actual technical potential.

Waste is widespread (85.4% have experienced waste), mainly due to improper storage or excessive purchase of overnight lychees.

Acceptance of overnight lychees is highly dependent on price compensation (63% demand a price reduction, 55% expect a 50% discount).

2. Potential for technological application:

The acceptance of synthetic biology technology is relatively high (78% accept), but safety concerns need to be prioritized (17% are worried).

The stratification of payment willingness is significant: 29% have a strong willingness to pay (willing to pay a premium for preservation technology), and 47% need a reasonable price range.

Young people (aged 18-25) and residents of first-tier cities are the core groups for technology promotion (73% have a willingness to pay and 84% have an acceptance rate).

3. Educational intervention points: Cognitive enhancement: Bridge the cognitive gap regarding preservation period (only 5% know the possibility of preservation for more than 5 days).

Technical trust building: Strengthen safety empirical evidence and popular science for 17% of concerned groups (such as technical principles, transparency of toxicological data).

Differentiated strategies: Residents in production areas: Emphasize variety diversity education (35% focus on variety vs. 16% for non-production areas).

Non-production area residents: Strengthen cold chain logistics guarantee and price advantage transmission (51% focus on price).

4. Promotion strategy for preservation technology:

Stratified technology positioning: In light of the strong payment willingness of young consumers (aged 18-25) (with A combined 76% of options A and B), high-value-added preservation technology should be prioritized for promotion.

Develop low-cost solutions for the middle-aged and elderly groups. Strengthening safety education: For the 17% of the concerned group, technical trust is established through empirical data (such as the reduction rate of pesticide residues and nutrient retention) and popular science channels (84% acceptance rate in first-tier cities).

Regional differentiation application: Production areas focus on extending the timeliness of local supply chains (direct procurement channels account for 31%), while non-production areas break through seasonal restrictions (non-peak season purchase demand 36%)

Researchers Wet Lab ​

Summary of Melatonin Synthesis-Related Project Based ​

Professor Qi Feng

Professor Professor Qi Feng

Ph.D., Professor, Doctoral Supervisor. He serves as Vice Dean of the College of Life Sciences at Fujian Normal University and Deputy Director of the Engineering Research Center of Industrial Microbiology (Ministry of Education) at Fujian Normal University. He has published more than 50 academic papers in relevant fields, obtained 18 authorized invention patents, and participated in compiling 2 Chinese and English works.

Research Interests: Synthetic Biology, Microbial Metabolic Engineering, Bioenergy, Industrial Microbial Fermentation Technology

Significance of the Exchange

We performed an online meeting with Prof. Qi and exchange some technical details and asked some engineering related questions. After this exchange, the team replaced the dual-plasmid backbone and verified the issues related to BH4 and MH4 in the metabolic pathway. Through the above adjustments, the cost can be reduced while the melatonin synthesis efficiency is improved, laying a theoretical foundation for the subsequent modification of Bacillus subtilis and the application of litchi preservation in our project.

Details：Project Adjustment Details and Basis (Combined with Supplementary Data)

I. Optimization of Dual-Plasmid Expression System Backbones

Adjustment Content: Replace dual-plasmid backbones to ensure stable coexistence and efficient expression of target enzymes in the two plasmids.

Specific Basis: Dual plasmids must differ in origin of replication, resistance marker, and transcription start site (to avoid homologous exclusion). One plasmid uses pWB980-ori, and the other uses pHY300PLK-p15Aori.

II. Complementing Bacillus subtilis MH₄ Recycling Pathway Using BH₄ Recycling Enzymes

Adjustment Content: Introduce BH₄ recycling enzymes (PCD and DHPR) to complement key missing enzymes in MH₄ synthesis and recycling in Bacillus subtilis and Escherichia coli, thereby improving melatonin synthesis efficiency.

Specific Basis:

• Pathway differences and commonalities: The core difference between MH₄ and BH₄ recycling lies in reductases: MH₄ recycling relies on DHMR (1.5.1.50, missing in Bacillus subtilis 6051), while BH₄ recycling depends on DHPR (1.5.1.34); both can share PCD (4.2.1.96), as PCD shows a binding confidence of 0.99 with both MH₄-4a and 4a-OH-BH₄, with a Kcat of 18.59, suggesting universality.

• Our dry lab model shows that BH₄ recycling enzymes have higher efficiency than MH₄ recycling enzymes in MH4 recycling, primarily due to the stronger catalytic activity of DHPR on MH2.

• Supplementation strategy for Bacillus subtilis: Endogenous MH₄ recycling in Bacillus subtilis 6051 lacks DHMR (1.5.1.50), resulting in inefficient reduction of MH₂ to MH₄; however, DHPR (1.5.1.34) from BH₄ recycling can partially replace DHMR function — structural simulation shows a binding confidence of 0.50 between MH₂ and DHPR (with tag), and model-calculated Kcat = 66.07 (high catalytic efficiency), which can be verified by subsequent experiments.

• Introduce BH₄-derived PCD (sequence referenced to hPCBD1: MAGKAHRLSAEER...) and DHPR (sequence referenced to hQDPR: MAAAAAAGEARRV...) to fill the "short slab" of MH₄ recycling, achieving coenzyme self-sufficiency and avoiding the addition of expensive exogenous BH₄.

Summary of Communication with Prof. Rainer Kalscheue ​

Prof. Rainer Kalscheue

Ph.D. (Dr. rer. nat.); Professor at Heinrich Heine University Düsseldorf; Institute for Pharmaceutical Biology and Biotechnology

Research Interests: Synthetic Biology, Microbial Metabolic Engineering, Bioenergy, Industrial Microbial Fermentation Technology

Significance of the Communication This communication gave our team critical "virtual validation":

• Verified core design feasibility (e.g., low-cost wax ester detection, endogenous substrates) and found mismatches (e.g., intracellular wax ester accumulation).

• Provided solutions for real-world bottlenecks (e.g., no exogenous sodium oleate, wax ester delivery) to align with lychee preservation goals.

• Avoided resource waste (e.g., confirmed low-cost staining) and improved design rigor.

Key Details: Pre-Communication: Design & Questions

Project: Lychee preservation via introducing FarA/WS/DGAT (from Marinobacter nauticus VT8) into Bacillus subtilis (preliminary E. coli DE3 test).Questions for Prof. Kalscheuer:

1. Wax ester synthesis without exogenous sodium oleate? (Team pre-designed lactose/IPTG-induced fatty acid mechanism.)

2. Low-cost staining (Oil Red O, Sudan III/IV) for ~1% CDW wax esters?

3. Intracellular accumulation (until lysis) or extracellular release of wax esters?

Key Details: During Communication: Prep & Delivery

Team confirmed genes/hosts (E. coli DE3 for test, Bacillus subtilis for use), designed the lactose/IPTG mechanism, and selected TLC/TEM + low-cost staining. Sent email with project info and 3 questions (referencing Prof. Kalscheuer’s 2006 research).

Key Details: Post-Communication: Feedback & Iteration

• Substrate: Team’s lactose/IPTG + Prof. Kalscheuer’s thioesterase gene → boost endogenous fatty acids, no sodium oleate.

• Detection: Low-cost staining works (if lipid content rises).

• Delivery: Wax esters accumulate intracellularly (conflicts with original design); team to try controlled lysis on lychees or secretion-capable hosts.

Summary of Communication with Prof Stanley Brul ​

Prof.dr. Stanley Brul

Chair Molecular Biology and Microbial Food Safety (SILS University of Amsterdam); Research head of laboratory MBMFS, Master track coordinator Medical Biochemistry & Biotechnology (former Biochemistry & Metabolic Disease); 2002-present MSc Biomedical Sciences; Director of the College of Life Sciences & bachelor Biomedical Sciences (faculty of Science and faculty of Medicine).

Research Interests: Biochemistry; Molecular Biology; Molecular stress physiology of (micro)organisms; 'Microbiome' studies of beneficial and pathogenic microbes, in particular Bacilli.

Significance of the Communication

To further evaluate the feasibility of using a pH-responsive system to regulate gene expression and induce self-lysis in our engineered bacteria, our team consulted Prof. Stanley Brul from the University of Amsterdam, an expert in microbial physiology and pH adaptation.

Through this discussion, we realized that our initial idea—to control gene expression and programmed cell death by detecting extracellular pH—was largely impractical. Prof. Brul explained that there is no clear quantitative relationship between extracellular and intracellular pH in bacteria, meaning that extracellular pH cannot serve as a reliable indicator for intracellular changes. Even if such a linear relationship existed, the response delay between pH detection, regulatory activation, and bacterial apoptosis would make it difficult to define a stable or biologically meaningful threshold.

Additionally, Prof. Brul pointed out that pH values vary significantly among different lychee varieties and even among individual fruits, making our proposed detection system unreliable for real-world applications. This communication thus played a crucial role in redirecting our design strategy — encouraging us to explore alternative regulatory mechanisms rather than relying solely on pH-sensitive control.

What did we ask?

1. Whether there exists any quantitative relationship between extracellular pH and intracellular pH in bacteria.
2. Since it is technically easier to engineer E. coli, we asked whether transferring the plasmid from E. coli to B. subtilis would cause compatibility or expression issues.
3. Whether the Cas1 promoter would be suitable for our engineered bacteria.
4. If we could design different proteins to be activated under different pH conditions and whether it would be possible to tune the pH threshold flexibly.
5. Recommendations for pH-responsive promoters that might work effectively in B. subtilis.

What did Prof. Brul reply?

1. There is no definitive quantitative relationship between extracellular and intracellular pH. Cellular regulation mechanisms buffer intracellular pH strongly, so it cannot be linearly inferred from external pH.
2. He suggested that instead of transferring plasmids from E. coli, we could directly engineer B. subtilis, which would be easier and more reliable for expression and transformation.
3. He did not give a firm conclusion about the Cas1 promoter's suitability but mentioned it may require further experimental validation.
4. He emphasized that pH-sensing and cellular response processes are not instantaneous. The time lag between sensing, gene expression, and programmed cell death makes precise threshold design extremely difficult.
5. Prof. Brul also noted that pH variation depends greatly on lychee variety and individual fruit differences, further complicating the use of pH as a control variable.
6. As he is not a specialist in genetic engineering, he offered to help the team contact other experts in his lab who have conducted related experiments.
7. Finally, he shared two relevant papers containing examples of pH-responsive promoters that might be considered for further study.

Dry Lab & Software iHP ​

Meeting with Prof. Pan ​

Dr. Pan

Ph.D. Max Planck Institute of Biochemistry; Postdoctoral Fellow at Stanford University; PI at GBA Institute of Precision Medicine; Director of Bioinformatics at VA Palo Alto.

Areas of Focus: Precision medicine, bioinformatics, translational genomics.

Dr. Pan (Ph.D. Max Planck Institute of Biochemistry; postdoc Stanford; PI at GBA Institute of Precision Medicine; Director of Bioinformatics VA Palo Alto) kindly organized a meeting for three of our dry lab members with her and two PhD students.

In this discussion, they recommended we begin modeling with Escherichia coli, Pseudomonas, and Salmonella as initial organisms. They also noted that antibiotic resistance represents a field with abundant data but relatively limited research, suggesting that exploring output modalities related to this theme could be impactful. Additionally, they highlighted Evo2 as a popular line of research worth studying for insights and inspiration.

Communication with DeepMind AlphaGenome Team ​

During the development of BactaGenome, we reached out to the AlphaGenome team at DeepMind regarding their bioRxiv preprint. We inquired about their rationale behind using unnormalized targets in their multinomial loss function. Their team responded with detailed clarification on combining Poisson loss for counts with multinomial loss for distribution, explaining that normalizing targets would erase important information about signal strength. They also shared preprocessing steps, statistical reasoning, and ablation insights. This exchange provided valuable direction, helping us stabilize training and consider multi-modal benefits. We followed up with further technical questions, to which the team responded with clarifications about clipping thresholds, scaling interpretations, and expected loss magnitudes. This dialogue was instrumental in adapting the AlphaGenome framework to bacterial genomics.

Communication with DeepMind AlphaGenome team - our inquiry (left) and their detailed response (right). We were really surprise when receiving the first response from DeepMind AlphaGenome team regarding our inquiry.

User Survey on Bio-AI Tools ​

We conducted a survey among undergraduate researchers to understand their opinions and expectations for AI tools in biology. Respondents, who often combine experiments with data analysis, reported dissatisfaction with current tools (difficult installation, lack of integration, poor interpretability). They valued explainability, learning support, integration, and accuracy, and showed strong interest in participating in offline demonstrations.

Key recommendations from survey results:

• Optimize tools around the three most common tasks students handle.
• Emphasize result interpretability and prediction accuracy.
• Improve integration and ease of use.
• Provide offline experience sessions for feedback collection.

Insights from User Experience ​

In addition to the survey, we connected with users who expressed willingness to join offline experience sessions.

Photo of one of the offline experience sessions (participant has approved his face shown publicly).

We tracked their usage flow and collected real-time feedback, which we then analyzed using grounded theory. Through index coding, we derived the following keywords that reflect user needs and expectations:

• Multilingual support (esp. Chinese/English)
• Assistance with plasmid editing, RBS selection, enzyme site creation
• Support for sequence design, search (linked to databases), amino acid ↔ nucleotide conversion
• Critique-friendly interface (“not good” responses considered valuable)
• Integration with tools like SnapGene, NCBI BUS
• Emphasis on interactive functions: tool suggestions, combined sequence linking, environment-specific simulation, binding site prediction
• Visualization and annotation tools for images and cell counts

These insights provided a clear roadmap for enhancing SynbioMCP, with emphasis on user-friendly design, interactive AI guidance, and robust integration.