Week 1 HW: Principles and Practices

The Prometheus Symbiont


🎅1.“The Prometheus Symbiont” is a conceptual, living medical system designed to symbiotically integrate with the human body. It merges biomimetic photosynthesis, synthetic biology, and flexible electronics, aiming to shift medicine from “passive treatment” to active, sustained life maintenance and enhancement. You can think of it as a sunlight-powered, wearable or implantable “second life-support system.”

🎅The Prometheus Symbiont is not merely a technological concept; it is more akin to a philosophical proposition about the future form of life. It blurs the boundaries between therapy and enhancement, between human and machine. Its ultimate significance may lie in compelling us to re-examine: “what constitutes health, and indeed, what it means to be human.”

Technical Integration How can living cells, electronic components, and polymer materials work together stably and safely within the human body over the long term?

Biosafety How can we prevent the leakage or mutation of genetically engineered microorganisms? How do we ensure the system can be safely degraded or cleared upon failure?

Ethical & Social Where is the boundary between the human body and machine? How is data privacy guaranteed? How can technological fairness be achieved?

Regulatory & Approval Does it belong to the category of medical devices, pharmaceuticals, or a new biological product? How can a completely new regulatory framework be established?

🎅Purpose: Current State vs. Proposed Change

What is done now (Current Paradigm): Medicine primarily operates in a reactive and episodic manner. Patients seek help after symptoms appear. Treatments involve separate devices (monitors), pharmaceuticals (drugs), and procedures, often with significant side effects and limited personalization. Sustainable energy and materials for medical devices are external concerns.

What we propose (Paradigm Shift): We propose a shift to a proactive, continuous, and integrated symbiosis. The Prometheus Symbiont is a single, autonomous system that continuously monitors, analyzes, and responds to the body’s state in real-time. It moves beyond treating illness to sustaining and enhancing baseline health. Crucially, it aims for energy and material autarky within the body by using biomimetic photosynthesis, fundamentally changing the relationship between medical technology and the patient’s own biological processes.

🎅Design: Requirements for Functionality & Key Actors Technical Core: Research Scientists (Synthetic Biology, Materials Science, Biomedical Engineering), Bioethicists, University Tech Transfer Offices 1.Stable Hybrid Bio-Machine Interface: Materials and protocols to seamlessly integrate living cells (engineered cyanobacteria/yeast), flexible electronics, and polymers. 2.Advanced Synthetic Biology: Engineered microbes for photosynthesis, sensing, and drug production with robust safety “kill-switches.” 3.Efficient Energy & Data Transfer: Systems for light capture, intracellular energy (ATP/NADPH) transfer to synthetic pathways, and secure bio-electrical data communication.

🎅Clinical & Regulatory Pathway: Government Regulators (FDA, EMA), Clinical Researchers, Ethics Boards, Patient Advocacy Groups 1.New Regulatory Framework: Classification as a novel “Symbiotic Biotherapeutic Device” requiring new FDA/EMA pathways. 2.Phased Clinical Trials: Long-term studies focusing on safety, stability, and efficacy for chronic conditions (e.g., diabetes, wound healing).

🎅Commercialization & Society: Venture Capitalists, Pharma/MedTech Companies, Government Funders (e.g., ARPA-H), Sociologists, The Public (as end-users and citizens) 1.Public-Private Funding Consortium: To fund high-risk R&D and scale-up. 2.Public Dialogue & Education: To build understanding and address ethical concerns before deployment. 3.New Manufacturing & Service Models: For growing, implanting, and maintaining living medical systems.

🎅Assumptions: Potential Uncertainties Technical Feasibility: We assume the extraordinary challenge of long-term, stable integration of diverse biological and electronic components within the dynamic human body can be solved. This is a fundamental uncertainty.

Biological Stability: We assume engineered genetic circuits will function predictably and reliably for decades without mutation or interference from the host’s immune system and microbiome.

Societal Acceptance: We assume that a significant portion of society will accept a permanent, living machine symbiont as a therapeutic or enhancement, overcoming the “yuck factor” and philosophical objections.

Regulatory Adaptability: We assume regulatory bodies can and will adapt at the pace of the technology to create prudent, effective pathways for such a disruptive product.

🎅Risks of Failure & “Success” Risks of Failure: Catastrophic Biofailure: Engineered microbes could mutate, cause infections, or disrupt vital physiological pathways, leading to patient harm. Rejection & Waste: The body’s immune system could reject the symbiont, or components could degrade into toxic byproducts. Technical Obsolescence: The embedded electronics or software could become outdated or hacked, rendering the system useless or dangerous.

🎅Risks of “Success” (Unintended Consequences): Exacerbating Inequality: The technology could create a biological divide between the “enhanced” wealthy and the “natural” poor, leading to unprecedented social stratification. Loss of Human Agency & Identity: If the system makes too many autonomous health decisions, it could erode personal bodily autonomy and challenge the very definition of being human. New Forms of Dependency & Vulnerability: Society could become dependent on a fragile technological ecosystem. Personal health data streams could be exploited for surveillance, discrimination, or coercion.

Ecological Impact: Widespread use and eventual disposal of genetically modified living devices could have unforeseen consequences on ecosystems if not perfectly contained.

In conclusion, the Prometheus Symbiont proposes a radical leap from repairing humans to architecting a hybrid human-machine biology. Its path is fraught with towering scientific hurdles and profound ethical questions, meaning its development must be accompanied by societal dialogue as intense as the engineering effort itself.

Based on the risk assessment of the disruptive technology “Prometheus Symbiont,” I recommend prioritizing the establishment of an “adaptive, multi-layered global governance framework” as the core focus of the governance strategy. My primary recommendation is directed at the Office of the United Nations Secretary-General, because the impact of this technology is inherently transboundary. Its ethical, safety, and equity issues require global coordination and consensus on principles; the potential risks cannot be effectively mitigated by the actions of any single nation.

My recommended priority solution is “Layered Adaptive Governance under Global Coordination.” This is not a single option, but a combination of international coordination, national/regional regulation, industry self-discipline, and public participation.

1. Top Layer: Establish Global Principles and Coordination Mechanisms (Led by the United Nations)

• Action: Promote the adoption of the “Global Declaration on Ethical and Governance Principles for Human-Technology Symbionts” and establish a standing, interdisciplinary Global Advisory Committee on Emerging Bio-Hybrid Technologies (GACEBT).
• Rationale: This provides the legitimacy foundation and “safety guardrails” for all subsequent governance. The committee, comprising scientists, ethicists, legal scholars, social activists, and government representatives, would be responsible for ongoing technology impact assessments, identifying transboundary risks (e.g., biosafety breaches, exacerbation of global inequality), and issuing non-binding guidelines. This avoids premature, rigid international legal constraints (which could stifle innovation) while establishing inviolable red lines.

2. Middle Layer: Develop National/Regional Specialized Regulatory Pathways (Led by Major Economies like the US, EU, and China)

• Action: Under the guidance of GACEBT principles, national regulatory agencies (e.g., US FDA, EU EMA, China NMPA) should jointly design new product categories (e.g., “Class I Symbiotic Therapeutic Device”) and approval pathways for “active symbiotic medical systems.” This should include mandatory phased clinical trial protocols and a post-market supervision model of “pre-certification, monitoring, and re-evaluation.”
• Rationale: This translates governance into actionable frameworks by entities with enforcement power. Coordination among major economies prevents regulatory arbitrage and provides a template for global standards. The pre-certification system allows for limited application under strict monitoring (e.g., for patients with terminal illnesses and no alternative therapies) while continuously collecting real-world data to refine the rules.

3. Grassroots Layer: Strengthen Industry Self-Regulation and Transparent Public Participation

• Action: Encourage leading research institutions (e.g., MIT, Chinese Academy of Sciences) and industry consortia to develop open-source safety standard protocols (e.g., engineering design standards for biocontainment modules). Simultaneously, legislation should require R&D projects to conduct transparent social impact assessments from an early stage and incorporate public input through mechanisms like citizens’ juries.
• Rationale: Governing technical details requires industry expertise, while public trust is foundational for societal acceptance. Open-source standards can accelerate the adoption of safe practices. Early public engagement helps identify social acceptance issues promptly, helping to avoid the public relations pitfalls experienced with technologies like GMOs.

🐱‍🐉Homework Questions from Professor Jacobson:

1.Nature’s machinery for copying DNA is called polymerase. What is the error rate of polymerase? How does this compare to the length of the human genome. How does biology deal with that discrepancy?

Answer： 1.Error Rate of DNA Polymerase: 1:106；Beese et al., (1993), Science, 260, 352-355. 2.The haploid human genome contains roughly 3.16 billion base pairs (≈ 3.16 × 10⁹ bp)；Without proofreading (at ~10^{-6 errors/bp), copying the entire genome once would introduce roughly: (3.16×10⁹)×10}−6≈31,60 mutations；With proofreading (at ~10-¹⁰ errors/bp), the expected number of errors per genome replication is: (3.16×10⁹)×10^-¹⁰≈0.316 mutations. This means, on average, less than one error per replication cycle—a biologically tolerable rate. 3.To ensure stable genome expression, biological systems employ multiple layers of regulation that calibrate differences arising from genetic variation, environmental influences, and stochastic molecular events：

Transcriptional Fidelity & Regulation 1.Proofreading in transcription: Although RNA polymerases lack the extensive proofreading seen in DNA replication, some backtracking and cleavage mechanisms exist (e.g., in eukaryotic Pol II) to correct misincorporated nucleotides. 2.Promoter specificity & transcription factors (TFs): TFs and enhancer/repressor elements precisely control when and where genes are expressed, minimizing off-target or noisy transcription. 3.Chromatin remodeling & epigenetic marks: Histone modifications, DNA methylation, and nucleosome positioning ensure that genes are expressed in the correct cell type and developmental stage, buffering against improper activation or silencing.

Post-transcriptional Control 1.RNA processing: Splicing, capping, and polyadenylation are highly regulated to produce consistent mature mRNA isoforms. 2.RNA surveillance pathways: Nonsense-mediated decay (NMD) degrades mRNAs with premature stop codons. No-go decay (NGD) and non-stop decay (NSD) clear stalled or faulty transcripts. RNA editing (e.g., A-to-I editing) can correct or diversify transcripts in a regulated manner. 3.MicroRNAs & other small RNAs: Fine-tune mRNA stability and translation, reducing expression variability and silencing aberrant transcripts.

Translational Accuracy & Control 1.Ribosome proofreading: During tRNA selection, ribosomes favor accurate codon–anticodon pairing; elongation factors (e.g., EF-Tu) and ribosomal RNA help discriminate correct vs. incorrect tRNAs. 2.Regulation of initiation: Initiation factors (eIFs) and upstream open reading frames (uORFs) modulate translation rates to match cellular needs and stress conditions. 3.Ribosome quality control (RQC): Recognizes stalled ribosomes and triggers degradation of incomplete polypeptides and potentially faulty mRNAs.

Protein Homeostasis (Proteostasis) 1.Chaperones & folding catalysts: Assist proper protein folding, preventing aggregation of misfolded proteins. 2.Ubiquitin-proteasome system & autophagy: Degrade damaged, misfolded, or excess proteins. 3.Feedback regulation: Many metabolic and signaling pathways use allosteric feedback or post-translational modifications to maintain stable protein activity levels.

DNA Repair & Genome Integrity Maintenance 1.Continuous operation of mismatch repair (MMR), base excision repair (BER), nucleotide excision repair (NER), and double-strand break repair pathways prevents mutations from accumulating and altering gene expression programs. 2.Cell-cycle checkpoints halt division if DNA damage is detected, allowing time for repair or triggering apoptosis if damage is irreparable.

Systems-Level Buffering 1.Genetic redundancy: Duplicate genes or paralogs can compensate for loss or reduced function of one copy. 2.Robust network architectures: Many regulatory networks (e.g., transcription factor networks, signaling cascades) are built with feedback loops, redundancy, and modularity to maintain stable outputs despite perturbations. 3.Noise filtering: Stochastic fluctuations in molecule numbers are dampened through negative feedback, time-averaging mechanisms, or threshold-based activation.

2.How many different ways are there to code (DNA nucleotide code) for an average human protein? In practice what are some of the reasons that all of these different codes don’t work to code for the protein of interest?

Answer： 1.For a 375-amino-acid protein, the total number of DNA sequences is the product of the number of codon possibilities for each position. 2.In essence, natural selection has chosen the specific DNA sequence for each human gene not just to encode the correct amino acids, but to also contain the precise regulatory, structural, and kinetic instructions needed for its proper expression, regulation, and function. The vast majority of theoretically possible sequences lack this full suite of integrated instructions. So it’s difficult to achieve under artificial conditions.

🤳Homework Questions from Dr. LeProust: 1.What’s the most commonly used method for oligo synthesis currently? The most widely used and established method for oligo synthesis is solid-phase synthesis using the phosphoramidite method. This technology is the industry standard for producing both DNA and RNA oligonucleotides in research and commercial settings. High Efficiency & Automation: Each coupling step has an efficiency exceeding 99%, enabling fully automated, high-throughput synthesis on machines. Versatile Chemistry: It provides a robust platform to introduce a vast array of chemical modifications (to the phosphate backbone, sugar, or base), which is crucial for creating therapeutic oligonucleotides like antisense drugs or siRNAs. Proven Reliability: As a mature technology refined over decades, it is the universal platform for commercial vendors and core facilities.

2.Why is it difficult to make oligos longer than 200nt via direct synthesis?

The ~200 nucleotide (nt) barrier for direct chemical synthesis is a fundamental limitation of the dominant phosphoramidite solid-phase method. The difficulty isn’t a single issue but a cascade of compounding chemical and practical problems.

The primary bottleneck is that synthesis is a stepwise process, and no chemical coupling is 100% efficient. This means for a 200-mer synthesis, over 60% of the product is shorter, failure sequences.

3.Why can’t you make a 2000bp gene via direct oligo synthesis? The challenges aren’t just linear; they become exponentially and prohibitively severe beyond ~200 nucleotides (nt). Synthesizing a 2000bp double-stranded gene would require creating a single-stranded oligo of at least 2000nt, which is scientifically and practically impossible with current direct chemical methods.

No existing purification technology (HPLC, PAGE, etc.) can separate a 2000nt strand from a 1999nt strand (a 0.05% difference in mass/length). The desired product is physically indistinguishable from the “near-miss” failures.

Gene assembly is the practical solution: it builds the skyscraper in prefabricated, high-quality sections (short oligos) and then welds them together with enzymatic precision.

😎Homework Question from George Church: 1.Choose ONE of the following three questions to answer; and please cite AI prompts or paper citations used, if any.

AI prompts:[Using Google & Prof. Church’s slide #4] What are the 10 essential amino acids in all animals and how does this affect your view of the “Lysine Contingency”?

Leucine\Isoleucine\Valine\Lysine\Methionine\Threonine\Tryptophan\Phenylalanine esp. for infants/young:Histidine Conditionally essential(esp. for young):Arginine

The “Lysine Contingency” is a clever plot device but is fundamentally flawed as a biological containment strategy for several key reasons:

the lysine contingency in Jurassic Park（Maynard, 2018; Rubini & Mayer, 2020） Lysine is already an essential amino acid for all vertebrate animals, including humans. This means animals like dinosaurs naturally cannot synthesize it and must obtain it from their diet. Therefore, the idea of “removing” a lysine-synthesizing ability they never possessed doesn’t work.

In summary, while the “Lysine Contingency” is an imaginative concept, it misunderstands basic animal biochemistry and fails as a practical fail-safe.

Rubini, R., & Mayer, C. (2020). Addicting Escherichia coli to new-to-nature reactions. ACS chemical biology, 15(12), 3093-3098. Maynard, A. (2018). Films from the future: the technology and morality of Sci-Fi movies. Mango Media Inc.

2.[Given slides #2 & 4 (AA:NA and NA:NA codes)] What code would you suggest for AA:AA interactions?

Energy-Based Interactions (More Accurate)

Using PyRosetta or BioPython with energy functions

import pyrosetta pyrosetta.init()

def calculate_interaction_energy(pose, res1, res2): """ Calculate interaction energy between two residues using PyRosetta. """ # Create a two-body energy calculator sfxn = pyrosetta.get_fa_scorefxn()

# Calculate energy between residues
emap = pyrosetta.EnergyMap()
pose.energies().residue_pair_energy(res1, res2, sfxn, emap)

return emap.total()

3.[(Advanced students)] Given the one paragraph abstracts for these real 2026 grant programs sketch a response to one of them or devise one of your own:

What if our most advanced biological medicines were as easy to ship and store as aspirin?

To break biologics’ extreme reliance on ultra-cold chains, a systemic transformation across science, logistics, and policy is required.

The immediate focus must be on re-engineering the molecules themselves. Massive investment in formulation science—utilizing advanced lyophilization, stabilizing sugars and polymers, and novel drying techniques—can shift storage from -70°C to 2-8°C or even room temperature. Parallel development of subcutaneous auto-injectors or oral delivery systems reduces dependency on clinic-based intravenous infusion.

Simultaneously, we must redesign the supply chain with intelligence and resilience. Deploying IoT-enabled smart containers with real-time tracking and blockchain ledgering ensures integrity and accountability. Creating distributed networks of certified storage points at regional centers expands access geographically. AI-driven predictive logistics can preempt shipping failures.

Long-term disruption will come from decentralizing production. Adopting modular, continuous biomanufacturing platforms enables regional or even hospital-based production, slashing distribution miles and cold-chain complexity. Next-generation platforms like thermostable lipid nanoparticles for nucleic acid therapies are equally crucial.

Finally, policy must incentivize accessibility. Regulators should create expedited pathways for thermostable products. Payers must align reimbursement with value metrics that include reduced logistical burden and improved patient access. A national strategy, treating biologic supply as critical infrastructure, can coordinate public-private R&D and strategic stockpiling.

The ultimate goal is a transition from a fragile, centralized, cold-dependent model to a resilient, distributed system where life-changing therapies are defined by their efficacy, not by the freezer they inhabit. This convergence of science, smart engineering, and supportive policy will democratize access to advanced medicines.

Week 2 HW: DNA Read, Write, & Edit

Homework — DUE BY FEB 17 2PM MIT TIME

👨‍🦰Part 0: Basics of Gel Electrophoresis

Keypoint: Gel Electrophoresis: Used for separating, identifying, and purifying fragments of DNA, RNA, or proteins.

Gel Preparation: Add agarose powder to the buffer, heat until melted, pour the solution into the gel tray, insert the comb, and allow it to cool and solidify.

Sample Loading: Remove the comb, place the gel into the electrophoresis tank, and add buffer until the gel is covered. Mix the DNA sample with loading buffer, then load the mixture into the wells.

Electrophoresis: Connect the power supply, set the voltage, and start running the gel. The tracking dye (e.g., bromophenol blue) can be seen moving downward with the naked eye.

Staining and Visualization: After electrophoresis, stain the gel by immersing it in a staining solution (e.g., nucleic acid dye), or add the dye to the gel beforehand during preparation. Finally, observe the bands under a UV light or a blue light transilluminator.

👲Part 1: Benchling & In-silico Gel Art

🧒Part 2: Gel Art - Restriction Digests and Gel Electrophoresis（Optional (for those with Lab access)）

🎅Part 3: DNA Design Challenge

3.1. Choose your protein.

Photosystem II：The structural analysis of Photosystem II (PSII) is of profound significance and holds substantial future value, primarily in three key areas: fundamentally understanding the water-splitting mechanism, elucidating the processes of its own biogenesis and repair, and inspiring the development of next-generation bio-inspired energy technologies.

1. Fundamental Understanding of the Water-Splitting Mechanism

The primary significance of PSII structural studies lies in unraveling how nature performs the energy-demanding and chemically complex reaction of water oxidation.

Atomic-Level Resolution of the Catalytic Core: Recent breakthroughs, such as the 1.7 Å resolution cryo-EM structure of PSII, have allowed scientists to visualize for the first time the positions of hydrogen atoms and the detailed water network within this massive membrane complex . This level of detail is crucial because it reveals how water molecules are channeled to the catalytic Mn₄CaO₅ cluster and how protons are guided out after water is split . Understanding these precise pathways is essential for comprehending the enzyme’s near-perfect efficiency（Hussein et al., 2024）.

Hussein, R., Graça, A., Forsman, J., Aydin, A.O., Hall, M., Gaetcke, J., & Schröder, W.P. (2024). Cryo–electron microscopy reveals hydrogen positions and water networks in photosystem II. Science, 384(6702), 1349-1355.

Capturing Reaction Dynamics: Beyond static snapshots, research is now focused on the dynamic process. For instance, serial femtosecond crystallography (SFX) using XFELs has enabled the capture of intermediate states (like the S₂ and S₃ states) in the catalytic cycle, revealing structural changes during the O-O bond formation . Furthermore, studies on specific mutants, such as the D2-Lys317Ala substitution, have shown how alterations in the hydrogen-bonding network can disrupt proton egress and slow down oxygen release, providing direct experimental evidence for the role of specific amino acids and channels.

Flesher, D.A., Shin, J., Debus, R.J., & Brudvig, G.W. (2025). Structure of a mutated photosystem II complex reveals changes to the hydrogen-bonding network that affect proton egress during O–O bond formation. Journal of Biological Chemistry, 301(3).

2. Elucidating Biogenesis, Repair, and Regulation

PSII is uniquely vulnerable to light-induced damage, particularly its D1 reaction center protein. Understanding how it is repaired is a research area of immense biological importance.

Unveiling the Repair Cycle: Structural biology has been pivotal in revealing the assembly and repair mechanisms of PSII. For example, research on green algae (Chlamydomonas reinhardtii) has solved the structures of four PSII-repair intermediates associated with the protein TEF30. These near-atomic resolution structures provide a working model for how different modules are reassembled during the mid-to-late stages of the repair cycle, a process vital for sustaining oxygenic photosynthesis under constant light stress（Wang et al., 2025）.

Wang, Y., Wang, C., Li, A., & Liu, Z. (2025). Roles of multiple TEF30-associated intermediate complexes in the repair and reassembly of photosystem II in Chlamydomonas reinhardtii. Nature Plants, 11(7), 1455-1469.

A Model System for Membrane Proteins: PSII is proving to be an excellent system for studying the general principles of how large, multi-subunit membrane protein complexes are assembled and maintained in the thylakoid membrane. Insights from PSII repair, such as the synchronization of chlorophyll synthesis with protein synthesis, have broader implications for cell biology and plant physiology(Komenda et al., 2024).

Komenda, J., Sobotka, R., & Nixon, P. J. (2024). The biogenesis and maintenance of PSII: recent advances and current challenges. The Plant Cell, 36(10), 3997-4013.

3. Future Value: Bio-inspired and Semi-Artificial Applications

The knowledge gained from PSII structures is a treasure trove for bioengineers and chemists aiming to create sustainable technologies. The future value lies in translating this biological blueprint into real-world applications.

Blueprint for Artificial Catalysts: The main barrier to scalable renewable energy, such as producing hydrogen as a fuel, is the reliance on rare and expensive metals (like platinum) to split water. PSII achieves this using cheap and abundant manganese and calcium . By understanding the precise atomic structure and mechanism of the oxygen-evolving complex, scientists hope to design synthetic catalysts that mimic nature’s solution for efficient water oxidation with earth-abundant materials(Hussein et al., 2024).

Hussein, R., Graça, A., Forsman, J., Aydin, A.O., Hall, M., Gaetcke, J., & Schröder, W.P. (2024). Cryo–electron microscopy reveals hydrogen positions and water networks in photosystem II. Science, 384(6702), 1349-1355.

Creating Semi-Artificial Photosynthetic Devices: A more direct application is the integration of isolated PSII proteins into bio-photoelectrochemical cells. A landmark study has successfully created a scalable “artificial leaf” by spray-coating PSII from spinach onto a specially designed protonated macroporous carbon nitride (MCN) support . This large-area photoanode (33 cm²) generated milliampere-level photocurrents with nearly 100% faradaic efficiency for oxygen production. The device was stable enough to power an LED when eight units were connected in series, demonstrating the potential of PSII-based biophotovoltaics for powering low-consumption electronic devices(Zhang et al.,2025).

Zhang, H., Tian, W., Lin, J., Zhang, P., Shao, G., Ravi, S. K., … & Wang, S. (2025). Photosystem II‐Carbon Nitride Photoanodes for Scalable Biophotoelectrochemistry. Advanced Materials, e08813.

https://www.uniprot.org/blast

sp|Q39195|PST2_ARATH Photosystem II 5 kDa protein, chloroplastic OS=Arabidopsis thaliana OX=3702 GN=PSBT PE=1 SV=2 MASMTMTATFFPAVAKVPSATGGRRLSVVRASTSDNTPSLEVKEQSSTTMRRDLMFTAAA AAVCSLAKVAMAEEEEPKRGTEAAKKKYAQVCVTMPTAKICRY

3.2. Reverse Translate: Protein (amino acid) sequence to DNA (nucleotide) sequence.

reverse translation of sample sequence to a 309 base sequence of most likely codons. atggcgagcatgaccatgaccgcgaccttttttccggcggtggcgaaagtgccgagcgcg accggcggccgccgcctgagcgtggtgcgcgcgagcaccagcgataacaccccgagcctg gaagtgaaagaacagagcagcaccaccatgcgccgcgatctgatgtttaccgcggcggcg gcggcggtgtgcagcctggcgaaagtggcgatggcggaagaagaagaaccgaaacgcggc accgaagcggcgaaaaaaaaatatgcgcaggtgtgcgtgaccatgccgaccgcgaaaatt tgccgctat

reverse translation of sample sequence to a 309 base sequence of consensus codons. atggcnwsnatgacnatgacngcnacnttyttyccngcngtngcnaargtnccnwsngcn acnggnggnmgnmgnytnwsngtngtnmgngcnwsnacnwsngayaayacnccnwsnytn gargtnaargarcarwsnwsnacnacnatgmgnmgngayytnatgttyacngcngcngcn gcngcngtntgywsnytngcnaargtngcnatggcngargargargarccnaarmgnggn acngargcngcnaaraaraartaygcncargtntgygtnacnatgccnacngcnaarath tgymgntay

3.3. Codon optimization.

1 ATGGCATCTA TGACTATGAC TGCTACATTC TTTCCTGCTG TAGCGAAGGT ACCAAGTGCT ACTGGGGGTA 71 GAAGGCTTAG CGTTGTTCGA GCGTCGACTT CGGATAACAC ACCTTCCTTA GAGGTGAAGG AGCAGTCATC 141 CACTACCATG AGAAGAGATC TGATGTTCAC TGCTGCTGCA GCAGCCGTAT GTTCCTTGGC CAAAGTCGCA 211 ATGGCTGAGG AAGAAGAACC TAAGAGAGGA ACTGAGGCGG CTAAGAAGAA GTATGCCCAA GTTTGTGTTA 281 CGATGCCTAC CGCGAAGATA TGCCGATAC

3.4. You have a sequence! Now what?

What technologies could be used to produce this protein from your DNA? Describe in your words the DNA sequence can be transcribed and translated into your protein. You may describe either cell-dependent or cell-free methods, or both.

1. Cell-Dependent Methods (In Vivo) This is the most common approach, where we insert our gene of interest into a host organism, turning it into a tiny protein factory.

Recombinant DNA & Cloning: The first step is to insert your gene of interest into a small, circular piece of DNA called a plasmid. This plasmid acts as a vector or delivery vehicle. It’s engineered to contain all the necessary control elements for the host cell to read the gene: a promoter (to start transcription), a ribosome binding site (to start translation), and often a selectable marker (like an antibiotic resistance gene) to help us find cells that have taken up the plasmid.

Transformation/Transfection: This recombinant plasmid is then introduced into the host cells. For bacteria like E. coli, this is called transformation. For animal cells, it’s often called transfection.

Selection and Growth: The host cells are grown on a special medium (e.g., containing an antibiotic). Only the cells that successfully took up the plasmid will survive and grow, forming colonies. Each colony is a clone of cells all producing your protein.

Induction and Harvesting: Once we have a large culture of these cells, we can add a chemical to induce the promoter, turning on high-level production of our target protein. After the cells have grown and produced the protein, they are harvested, and the protein is purified away from all the host cell’s components.

Common Host Organisms:

E. coli (Bacteria): The workhorse of the industry. It’s fast, cheap, and easy to grow. Best for simple proteins that don’t require complex modifications.

Yeast (e.g., S. cerevisiae): A single-celled fungus that is also easy to grow but can perform some more complex protein processing tasks than bacteria.

Mammalian Cells (e.g., CHO cells): The gold standard for complex human therapeutic proteins (like antibodies). They can perform all the necessary human-like modifications (like glycosylation) to make the protein fully functional and safe.

Insect Cells: A good middle-ground, using a virus (baculovirus) to infect insect cells, which then produce the protein. They offer more complex processing than yeast but are easier to handle than mammalian cells.

2. Cell-Free Methods (In Vitro) These systems produce proteins without using living cells. Instead, they use the cellular machinery (ribosomes, tRNAs, enzymes) extracted from cells.

How it works: A cell lysate is created by breaking open cells (like E. coli, wheat germ, or rabbit reticulocytes) and removing the cell debris. What’s left is a “soup” containing all the components needed for transcription and translation: ribosomes, amino acids, tRNA, and energy-generating molecules. To this soup, you add your DNA template (containing your gene) and the necessary nucleotides.

Transcription and Translation: If you add a DNA template, the system will begin transcribing it into mRNA and immediately translating that mRNA into protein, all in the same test tube.

Advantages:

Speed: Protein production can happen in hours, not days or weeks.

Toxicity: You can produce proteins that would be toxic to a living cell, as there’s no cell to kill.

Simplicity: It bypasses the need for cloning, transformation, and maintaining cell cultures.

Labeling: It’s very easy to add modified amino acids (e.g., with fluorescent tags) for research purposes.

3.5. [Optional] How does it work in nature/biological systems?

1. Alternative Splicing This is the most common and well-studied mechanism, particularly in complex eukaryotes like humans.

The Basic Process: Genes in eukaryotic cells contain coding sequences called exons and non-coding intervening sequences called introns. When a gene is transcribed, the entire region (both exons and introns) is copied to create a pre-mRNA molecule. Before this pre-mRNA can be used to make a protein, the introns must be removed and the exons joined together in a process called splicing.

The Alternative Part: In alternative splicing, the cell’s splicing machinery doesn’t always join the exons together in the same way. It can selectively include or exclude different exons from the final, mature mRNA molecule.

Imagine a gene with exons 1, 2, 3, and 4.

In one cell type, splicing might join all four exons: Exon 1 - Exon 2 - Exon 3 - Exon 4. This creates mRNA “Version A,” which codes for Protein A.

In another cell type, or at a different developmental stage, the splicing machinery might skip exon 2: Exon 1 - Exon 3 - Exon 4. This creates mRNA “Version B,” which codes for a different Protein B.

It could also include an extra exon (Exon 2a) that isn’t always used, leading to Protein C.

Examples:

The DSCAM gene in fruit flies can generate over 38,000 different mRNA isoforms through alternative splicing!

The Calcitonin/CGRP gene produces a hormone (calcitonin) in the thyroid gland and a neuropeptide (CGRP) in the brain by using different sets of exons.

2. Alternative Promoters A gene can have more than one promoter site, which is the “start here” signal for RNA polymerase to begin transcription.

The Mechanism: Depending on which promoter is used, transcription will start at a different point in the gene. This can lead to pre-mRNAs that have different “first exons.”

The Result: These different starting points can result in mature mRNAs with different 5’ ends. This often means the resulting proteins will have different N-termini (the beginning of the protein). This can affect where the protein is located within the cell or what its function is.

3. Alternative Polyadenylation At the end of transcription, the pre-mRNA is cleaved, and a string of adenine nucleotides (the poly-A tail) is added to the 3’ end. This process is called polyadenylation and is signaled by a specific sequence in the RNA called the polyadenylation signal.

The Mechanism: Some genes have multiple polyadenylation signals. If the cell’s machinery uses the first signal, it will cleave the RNA there, resulting in a shorter mRNA. If it uses a downstream signal, it will produce a longer mRNA.

The Result: This affects the 3’ end of the mRNA. Since the 3’ untranslated region (3’ UTR) often contains signals for mRNA stability, localization, and how efficiently it’s translated, different polyadenylation choices can dramatically affect how much protein is made and where. In some cases, it can also alter the very end of the protein-coding sequence itself.

👩‍🦰Part 4: Prepare a Twist DNA Synthesis Order

4.1. Create a Twist account and a Benchling account √

4.2. Build Your DNA Insert Sequence

Copy：https://benchling.com/s/seq-GWC7bWlMPkMgqEnihkxn?m=slm-ny5MpJ1N9FsOATmyQMAl

De novo design：https://benchling.com/s/seq-PmLRVhHnWcUpDyCXJjHa?m=slm-TapXq6UoRnBTOtZCWyOT

Promoter： Arabidopsis thaliana chloroplast psbA gene promoter This is a core promoter region of approximately 620 bp upstream from the start codon ATG (containing the -35 box and -10 box regions). text

ATTGCTTGAT TTAATTTTTC AATTTTCTTG TTTTTATTTT GAATAAAGGA AAATAAATAA AAATAAATAA AATTTTTTTA AAAAGAATTT AATTTTCTAA CTTTTTTTAT TTTATCAACA AAAATATCTT ATTTTATTTC GATTTTATTT AGATTTTAGT ATCTATTTTT GGTTGATATA TATGGTTTTA TATTTGATAG GTATATTTGT TTTGATTGAA ATTTTCTGAA AAATATTTTT AAATAAATGA TTATTCTTTT CTCTCTAGAT CTTATATGTA GAATCTTTAT ATTTTGATAA TATTTTTTGA TTTTGATTTT TGTTTGTTTG TTTTTTATAC ATATATTTTT GGGGATTTTT TTTTTGTTTT TCAATTTCAA TTTCTCTAGA AAAAAGAGGA GAAAATTAAT ATG

RBS (Ribosome Binding Site)：AGGAGG

Coding Sequence (your codon optimized DNA for a protein of interest, psii for example):

1 ATGGCATCTA TGACTATGAC TGCTACATTC TTTCCTGCTG TAGCGAAGGT ACCAAGTGCT ACTGGGGGTA 71 GAAGGCTTAG CGTTGTTCGA GCGTCGACTT CGGATAACAC ACCTTCCTTA GAGGTGAAGG AGCAGTCATC 141 CACTACCATG AGAAGAGATC TGATGTTCAC TGCTGCTGCA GCAGCCGTAT GTTCCTTGGC CAAAGTCGCA 211 ATGGCTGAGG AAGAAGAACC TAAGAGAGGA ACTGAGGCGG CTAAGAAGAA GTATGCCCAA GTTTGTGTTA 281 CGATGCCTAC CGCGAAGATA TGCCGATAC

Mixed codons (CAT and CAC) are used to avoid long repetitive sequences, facilitating synthesis and cloning stability.

CAT CAC CAT CAC CAT CAC CAC Length: 21 bp

Stop Codon: TAA

Terminator (BBa_B0015): CCAGGCATCAAATAAAACGAAAGGCTCAGTCGAAAGACTGGGCCTTTCGTTTTATCTGTTGTTTGTCGGTGAACGCTCTCTACTAGAGTCACACTGGCTCACCTTCGGGTGGGCCTTTCTGCGTTTATA

💡 Design Strategy: Screening for Strong Terminators from the Chloroplast Genome

In higher plant chloroplasts, the transcription termination mechanism is similar to that of prokaryotic systems, typically relying on a stem-loop structure located at the 3’ end of a gene. You can construct an efficient standard part by following the two steps below:

1. Identifying Candidate Sequences: The transcript of the psbA gene (encoding the PSII core protein D1) in the chloroplast is highly abundant and stable, and its 3’ UTR usually contains efficient termination and processing signals. You can obtain the complete chloroplast genome sequence of Arabidopsis thaliana from public databases like NCBI (e.g., GenBank accession: NC_000932.1), then locate the psbA gene and extract its 3’ UTR region (approximately 100-200 bp) as the core candidate sequence.

2. Engineering for High Efficiency: To pursue near 100% termination efficiency, you can refer to the design logic of BBa_B0015 and construct a dual-terminator tandem element:

• First Unit: Clone the 3’ UTR of the psbA gene from the Arabidopsis chloroplast.
• Second Unit:Clone another strong terminator, such as the 3’ UTR of the chloroplast rps16 gene or a strong termination signal from other chloroplast genes.
• Combination:Link these two units in tandem with a short spacer sequence. This “belt-and-suspenders” structure can maximally prevent read-through by RNA polymerase.

PDF:content/homework/week-02-hw-dna-read-write-and-edit/constitutive_psII_Arabidopsis-thaliana-sequence.pdf FASTA:content/homework/week-02-hw-dna-read-write-and-edit/constitutive_psii_arabidopsis-thaliana.fasta

4.3. On Twist, Select The “Genes” Option √

4.4. Select “Clonal Genes” option√

Keypoints：An advantage is If designed with the appropriate exonuclease protection, gene fragments can be used directly in cell-free expression. 4.5. Import your sequence√

content/homework/week-02-hw-dna-read-write-and-edit/constitutive_sfGFP_his_tag.gb

building your first plasmid!√

content/homework/week-02-hw-dna-read-write-and-edit/first plasmid.png

🤴Part 5: DNA Read/Write/Edit

5.1 DNA Read (i) What DNA would you want to sequence (e.g., read) and why? This could be DNA related to human health (e.g. genes related to disease research), environmental monitoring (e.g., sewage waste water, biodiversity analysis), and beyond (e.g. DNA data storage, biobank).

If I were to explore the possibility of extraterrestrial life and its evolution through DNA sequencing, I would focus on the following DNA targets and contexts, each offering unique insights into how life might arise and adapt beyond Earth:

1. DNA from Extraterrestrial Samples (e.g., Mars, Europa, Enceladus) What to sequence: Any organic or genetic material recovered from soil, ice plumes, or subsurface oceans of celestial bodies.

Why:

To determine if life elsewhere uses the same genetic code (DNA/RNA) or something entirely novel.

To compare sequences with Earth life to test theories of panspermia (whether life spreads via meteorites) or convergent evolution (whether life independently evolves similar solutions).

To identify biosignatures—patterns in DNA that indicate biological activity, such as non-random sequence complexity or metabolic genes.

2. Extremophile Genomes (Earth Analogs for Space Environments) What to sequence: Complete genomes of organisms like Deinococcus radiodurans (radiation-resistant), Tardigrades (space-tolerant), or psychrophiles (cold-loving) from Antarctica.

Why:

These organisms serve as models for how life might survive in space or on harsh planets like Mars (low pressure, radiation, cold).

Their DNA repair mechanisms, desiccation tolerance genes, and metabolic pathways can be compared with hypothetical extraterrestrial life to predict survival strategies.

3. Ancient or “Shadow Biosphere” DNA on Earth What to sequence: Environmental DNA (eDNA) from extreme, isolated niches (e.g., deep subsurface mines, high-altitude lakes, or Atacama Desert soils).

Why:

To search for a “second genesis” of life on Earth—organisms with different biochemistry or genetic codes—which would profoundly impact how we search for life elsewhere.

To understand the limits of life’s evolutionary paths and identify universal constraints that might apply anywhere in the cosmos.

4. Synthetic DNA for Life-Detection Instruments What to sequence: Engineered DNA sequences designed as controls or standards for space missions (e.g., the Signatures of Life Detector on a rover).

Why:

To calibrate instruments (like nanopore sequencers) for detecting non-standard or damaged DNA that might be found on other planets.

To test whether our detection methods are biased toward Earth-like life, ensuring we don’t miss “weird” life with different base pairs or chirality.

5. Genomes of Organisms in Simulated Space Environments (ISS or Lab) What to sequence: DNA of bacteria, fungi, or plants exposed to microgravity, cosmic radiation, or Mars-like conditions on the International Space Station or in simulation chambers.

Why:

To study real-time evolutionary adaptation to space conditions.

To identify mutations or horizontal gene transfer events that occur under extraterrestrial stress, revealing how life might evolve during interplanetary travel.

6. Universal Genetic Code Variations (Bioinformatics) What to sequence: Not physical DNA, but in silico simulations of genetic codes and proteins that could function in exotic solvents (e.g., methane or ammonia) or at extreme temperatures.

Why:

To expand our concept of “possible life” beyond carbon-water-DNA constraints.

To guide the search for alien genes by predicting what sequences might look like in environments like Titan’s hydrocarbon lakes.

(ii) For exploring extraterrestrial life and its evolution, I would choose Oxford Nanopore Technologies (ONT) sequencing, a third-generation sequencing platform. Here’s a detailed breakdown addressing your questions:

Technology Selection and Rationale Oxford Nanopore Technologies (ONT) sequencing is the ideal choice for extraterrestrial life exploration

Oxford Nanopore Technologies (ONT) sequencing Generation Third-generation (single-molecule, long-read sequencing) Input Extracted DNA from extraterrestrial samples (soil, ice, plumes, etc.) Output Real-time electrical current signals converted to base sequences (FAST5 files)

5.2 DNA Write This is a creative and fascinating idea—essentially engineering a living biomaterial inspired by both the fictional character Baymax (from Big Hero 6) and the real-life sea slug Costasiella kuroshimae (commonly known as “Leaf Sheep” or “Solar-Powered Sea Slug”). The leaf sheep is one of the few animals capable of kleptoplasty—it steals chloroplasts from the algae it eats and incorporates them into its own cells, enabling it to photosynthesize for months.

If I were to synthesize DNA for a “Baymax-like self-healing, photosynthetic biomaterial,” it would involve designing a synthetic genetic circuit that could be introduced into a compatible host (e.g., mammalian cells, skin cells, or even a cell-free system) to create a living material with the following properties:

Self-powering via photosynthesis (like the leaf sheep)

Self-healing (like Baymax’s inflatable skin)

Biocompatible and responsive to the body

🧬 DNA to Synthesize: A Photosynthetic & Self-Healing Genetic Circuit I would synthesize a multi-gene synthetic construct containing the following modules:

Module Gene(s) Function

1. Photosynthesis Module psbA, psbD, rbcL, rbcS Enables light capture, electron transport, and carbon fixation (chloroplast function)
2. Self-Healing / Repair Module DPS (DNA protection during starvation), sodB (superoxide dismutase), katE (catalase) Protects cells from oxidative damage during light exposure; promotes tissue repair
3. Adhesion & Matrix Module COL1A1 (human collagen), FN1 (fibronectin) Provides structural scaffold for tissue integration and healing
4. Regulatory / Synthetic Circuit Light-inducible promoter (e.g., pDawn), GFP reporter Allows photosynthesis genes to be activated only in the presence of light

🔬 Full DNA Sequence Concept (Simplified Example) Here is a simplified, conceptual DNA sequence combining parts of the above ideas. It includes: A light-inducible promoter (pDawn system: YtvA + FixJ) The psbA gene (PSII core protein) for photosynthesis The DPS gene for oxidative stress protection A collagen fragment for tissue integration A terminator (BBa_B0015)

🧠 Why Synthesize This DNA?

1. Baymax-Inspired Self-Healing Material Baymax’s skin is soft, inflatable, and can repair itself. By incorporating collagen and fibronectin genes, the material could integrate with human tissue and promote wound healing. The DPS and catalase genes would protect cells from oxidative stress (common in damaged tissue), enabling longer-lasting repair.
2. Photosynthesis for Self-Powering (Leaf Sheep Model) The leaf sheep is a solar-powered animal. If we can engineer mammalian cells (or a skin substitute) to stably incorporate and maintain functional chloroplasts (via genes like psbA and rbcL), the material could generate its own energy from light—reducing the need for external power or nutrient supply in medical implants or wearables.
3. Potential Applications Medical Implants: Self-healing, light-powered skin grafts or patches for chronic wounds. Wearable Biosensors: Living tattoos that change color in response to inflammation or UV exposure. Space Exploration: Living materials for astronauts that require minimal resources (just light and water). Eco-Friendly Biomaterials: Photosynthetic fabrics or coatings that capture CO₂ and produce oxygen.

Next Steps for Synthesis If Twist Bioscience were to synthesize this, I would: Codon-optimize each gene for the target host (e.g., human cells or E. coli for prototyping). Add RBS, linkers, and terminators between modules. Clone into a delivery vector (e.g., lentivirus for mammalian cells or plasmid for bacterial expression). Test in a chassis like E. coli first to verify photosynthesis and oxidative protection, then move to mammalian cell lines.

For synthesizing the complex, multi-gene “Baymax-Meets-Leaf-Sheep” DNA construct, I would recommend a hybrid approach that leverages the strengths of different synthesis technologies. Given the length (~2,000+ bp), complexity (multiple genes from different sources), and the goal of creating a functional genetic circuit, the optimal strategy is:

High-throughput silicon-based DNA synthesis (e.g., Twist Bioscience platform) for fragment generation, followed by enzymatic assembly (e.g., Gibson Assembly or Golden Gate) for final construct assembly.

1. Technology Selection and Why Primary Technology: Silicon-Based High-Throughput DNA Synthesis (e.g., Twist Bioscience)

Why: Construct is large and contains multiple genes (psbA, DPS, COL1A1, etc.) with varying GC content and potential secondary structures. Traditional column-based synthesis would be slow, expensive, and error-prone for this complexity . Twist’s platform miniaturizes the chemical synthesis (phosphoramidite chemistry) by performing reactions in nanowells on a silicon chip . This allows for the parallel synthesis of thousands of oligos at once, dramatically increasing throughput and reducing cost . They can routinely synthesize oligonucleotides up to 500 nt in length, which serve as the building blocks for larger genes.

Generation: This is a first-generation (chemical) method but with a modern, high-throughput twist. The core chemistry is the established phosphoramidite method developed in the 1980s , but the delivery system (silicon chip) is a revolutionary 21st-century innovation that solves scalability issues .

Secondary Technology: Enzymatic DNA Assembly (e.g., Gibson Assembly® or Golden Gate Assembly)

Why: The 500 nt fragments from the chip need to be stitched together to create your final multi-gene construct (~2-5 kb). Enzymatic assembly methods are ideal for this. They use enzymes to simultaneously join multiple DNA fragments with overlapping ends in a single reaction . This is far more efficient than using restriction enzymes and ligase.

2. Essential Steps of the Chosen Method The workflow combines the synthesis steps with the assembly steps.

Part A: DNA Synthesis (The “Writing” of Fragments) Sequence Design and Upload: You provide the digital DNA sequences for your photosynthetic module, repair module, etc., to the synthesis provider (e.g., Twist).

Silicon Chip Manufacturing: A silicon chip with thousands of nanowells is prepared. Each well is designated for the synthesis of a specific oligonucleotide .

Cyclic Nucleotide Addition (Phosphoramidite Chemistry): The chip undergoes repeated cycles to build the oligos base-by-base from the 3’ end to the 5’ end. Each cycle for each base consists of four core chemical steps :

Deprotection (Detritylation): Acid removes a protecting group (DMT) from the 5’ hydroxyl of the last nucleotide, making it reactive.

Coupling: The next nucleotide (phosphoramidite monomer) is activated and added, forming a bond with the exposed 5’ hydroxyl.

Capping: Any unreacted 5’ hydroxyls are acetylated to prevent them from reacting in future cycles, which would cause deletions.

Oxidation: Iodine and water are used to stabilize the newly formed bond into a natural phosphate backbone.

Cleavage and Deprotection: After all cycles are complete, the synthesized oligos are cleaved from the chip, and all remaining protecting groups are removed using ammonium hydroxide .

Amplification and QC: The single-stranded oligos are amplified (often via PCR) to create double-stranded DNA fragments. These fragments are then purified and quality-controlled to ensure the correct sequence.

Part B: DNA Assembly (Building the Final Construct) Fragment Design: You design the ~500 bp fragments so that their ends have short, overlapping sequences (20-40 bp) that are complementary to the adjacent fragment.

Assembly Reaction (e.g., Gibson Assembly): All fragments, along with a linearized vector backbone, are mixed in a single tube with an enzyme master mix containing three activities:

Exonuclease: chews back nucleotides from the 5’ ends of the fragments, creating single-stranded overhangs that allow the complementary overlapping regions to anneal.

DNA Polymerase: fills in any gaps in the annealed regions.

DNA Ligase: seals the nicks in the sugar-phosphate backbone, creating a fully circular plasmid.

Transformation: The assembled plasmid is transformed into competent E. coli cells.

Screening and Verification: Colonies are screened for the correct insert, and the final plasmid is verified by Sanger sequencing to ensure 100% accuracy.

3. Limitations of the Method (Speed, Accuracy, Scalability) While this hybrid approach is the best available, it has inherent limitations.

Aspect Limitation Explanation Speed Not real-time. The entire process, from design to receiving a verified plasmid, typically takes 2-4 weeks. This is due to synthesis run times, shipping, assembly, cloning, and final sequencing verification. It is a batch process, not an instantaneous one. Accuracy Error accumulation in long, complex sequences. While the synthesis coupling efficiency is high (>99.5% per step) , errors (deletions, insertions, substitutions) are inevitable. For a long construct like yours, the probability of having at least one error in the final assembled product is significant. High-GC content, repetitive sequences, and strong secondary structures (like those found in some photosynthetic genes) can further increase error rates . This often necessitates sequencing multiple clones to find a perfect one. Scalability Assembly becomes a bottleneck. While silicon-chip synthesis is highly scalable for making millions of oligos , assembling them into many different, large, and complex constructs remains a manual and low-throughput process. Scaling up to make hundreds or thousands of different versions of your Baymax circuit is currently a significant bioengineering challenge.

5.3 DNA Edit

Although, in principle, gene editing has created many advantageous genes and aligns with the Darwinian principle of “survival of the fittest” in terms of survival and development—which is also very consistent with the basic principle of gene silencing or loss during long-term natural selection—I feel that, compared to human-directed evolution, natural random mutation actually shows greater respect for the individual will of living beings. Therefore, I do not like gene editing.

Based on thoughtful reflection on the ethical dimensions of gene editing, I will proceed with the technical analysis as requested while acknowledging the important philosophical considerations.

If I were to perform DNA edits—specifically to create the photosynthetic, self-healing “Baymax” biomaterial described earlier—I would choose the following technology:

Technology Selection: CRISPR-Cas9 CRISPR-Cas9 (Clustered Regularly Interspaced Short Palindromic Repeats) is the most suitable technology for this application because it offers precision, flexibility, and efficiency for introducing multiple genes into a target genome.

Why CRISPR-Cas9? Requirement/Why CRISPR-Cas9 Fits Multi-gene insertion Can target multiple loci simultaneously or sequentially Mammalian cell compatibility Well-established protocols for human cell lines Precision Can insert genes at specific “safe harbor” loci (e.g., AAVS1 in human cells) Efficiency High editing rates in many cell types

How CRISPR-Cas9 Edits DNA: Essential Steps Mechanism Overview CRISPR-Cas9 uses a guide RNA (gRNA) to direct the Cas9 nuclease to a specific DNA sequence, where it creates a double-strand break (DSB). The cell’s natural repair mechanisms then introduce edits: DNA Recognition: The gRNA contains a 20-nucleotide spacer complementary to the target DNA, adjacent to a PAM sequence (NGG) required for Cas9 binding. Double-Strand Break: Cas9 cuts both DNA strands, creating a DSB. DNA Repair: The cell repairs the break via: Non-Homologous End Joining (NHEJ): Error-prone repair that creates insertions/deletions (indels) to disrupt genes. Homology-Directed Repair (HDR): Precise repair using a DNA template, allowing gene insertion or correction.

Essential Steps for Your Project

1. Design Phase (Preparation) Input Required: Target genome sequence (e.g., human cell line reference) Donor DNA template (containing your photosynthetic genes) gRNA design tools Design Steps: Select Target Locus: Choose a “safe harbor” site (AAVS1, CCR5, or HPRT) where gene insertion won’t disrupt essential genes . Design gRNA: Use tools (CRISPOR, Benchling) to select 20-nt sequences adjacent to PAM sites with minimal off-target matches . Design Donor Template: Create a DNA fragment containing: Your photosynthetic gene cassette (psbA, rbcL, etc.) Left and right homology arms (500-800 bp each) matching sequences flanking the cut site Optional selection marker (e.g., GFP or puromycin resistance)

2. Delivery Phase Input Required: Cas9 protein or mRNA gRNA (synthetic or expressed from plasmid) Donor DNA template (for HDR) Target cells (e.g., human fibroblasts or induced pluripotent stem cells)

Delivery Methods: Transfection: Lipofection or electroporation of Cas9-gRNA ribonucleoprotein (RNP) complexes—preferred for efficiency and reduced off-target effects. Viral Delivery: Lentivirus or AAV for hard-to-transfect cells. Nucleofection: Electroporation-based method for primary cells.

3. Editing Phase Cellular Process: RNP complex enters nucleus gRNA guides Cas9 to target DNA Cas9 creates DSB If donor template present, cell may use HDR to insert your gene cassette If no template, NHEJ causes gene disruption

4. Screening and Validation PCR Screening: Test for correct integration using primers flanking the insertion site Sanger Sequencing: Verify precise sequence of edited locus Functional Assays: Confirm photosynthetic protein expression and activity

This approach, while technically challenging, represents the current state-of-the-art for introducing complex synthetic circuits into human cells. The limitations—particularly low HDR efficiency for large inserts—mean that success would require significant optimization and screening, but the technology exists to make OUR vision possible.

Week 3 HW: hw-lab-automation

ヾ(≧▽≦*)oAssignment: Python Script for Opentrons Artwork — DUE BY YOUR LAB TIME!

The Biopunk lab hasn’t contacted me yet.

The Opentrons API is a Python framework for writing automated biology lab protocols. 1.Load labware (containers, tip racks, plates); 2.Load instruments (pipettes); 3.Define your liquid handling steps;

The basic artistic GUI will involve: Getting coordinates from the GUI tool; Writing a Python script that moves the pipette to those positions; Using the HTGAA26 Colab notebook as your template:https://ddls.aicell.io/course/ddls-2025/module-6/lab/#-what-is-a-code-agent;

(✿◡‿◡)Post-Lab Questions — DUE BY START OF FEB 24 LECTURE

One of the great parts about having an automated robot is being able to precisely mix, deposit, and run reactions without much intervention, and design and deploy experiments remotely.

For this week, we’d like for you to do the following:

👳‍♂️Find and describe a published paper that utilizes the Opentrons or an automation tool to achieve novel biological applications.

Khan, S. U., Møller, V. K., Frandsen, R. J. N., & Mansourvar, M. (2025). Real-time AI-driven quality control for laboratory automation: a novel computer vision solution for the opentrons OT-2 liquid handling robot: SU Khan et al. Applied Intelligence, 55(7), 524.

1. Systematic Study of Yeast Gene Expression and Pipetting Speed This study, published in 2025, used the Opentrons OT-2 robot to systematically investigate the effects of pipetting speed on the growth and gene expression of Saccharomyces cerevisiae.

Research Content: The researchers used the OT-2 robot to precisely control pipetting parameters and performed liquid handling on yeast cultures at four different speeds (50, 130, 210, 290 μL/s). Quantitative growth assays and RNA sequencing analysis were conducted to evaluate the impact of pipetting speed on yeast.

Innovation and Findings: The study found that within the tested speed range, changes in pipetting speed did not significantly affect the maximum relative growth rate or gene expression profiles of yeast. The gene expression of all 24 samples was highly similar, with a minimum Pearson correlation coefficient of 0.9528. This indicates that the fastest pipetting speed (290 μL/s) can be used in yeast experiments to improve efficiency without negatively affecting cell state.

Biological Significance: This research demonstrates the value of robotic platforms in optimizing experimental parameters and improving reproducibility and accuracy, providing an important reference for determining appropriate operating parameter ranges in future automated experiments.

Taguchi, S., Matsuzawa, R., Suda, Y., Irie, K., & Ozaki, H. (2025). Investigating the effects of liquid handling robot pipetting speed on yeast growth and gene expression using growth assays and RNA-seq. Micropublication Biology, 2025, 10-17912.

2. Semi-Automated Workflow for Conjugative Transfer in Streptomyces This study, published in 2025, proposed “ActinoMation,” a semi-automated, medium-throughput workflow for conjugative transfer in Streptomyces using the Opentrons OT-2 robot platform.

Research Content: The research team developed an open-source protocol creation tool called ActinoMation, using Python and Jupyter Notebook to achieve a readable programming environment. They validated the method in various Streptomyces strains (S. coelicolor, S. albidoflavus, S. venezuelae).

Innovation and Findings: The automated conjugation workflow made large-scale transformations easy with no significant loss in transformation efficiency. The study reported detailed conjugation efficiencies for different strain-plasmid combinations; for example, the conjugation efficiency of S. venezuelae DSM40230 with the pSETGUS plasmid reached 4.97%.

Biological Significance: Streptomyces are important producers of antibiotics and other bioactive compounds. This automated method addresses the labor-intensive and slow nature of traditional manual conjugation protocols, providing a feasible solution for the efficient genetic engineering of these strains.

Møller, T. A., Booth, T. J., Shaw, S., Møller, V. K., Frandsen, R. J., & Weber, T. (2025). ActinoMation: A literate programming approach for medium-throughput robotic conjugation of Streptomyces spp. Synthetic and Systems Biotechnology, 10(2), 667-676.

3. Semi-Automated Production of Cell-Free Biosensors This 2025 study explored the use of the Opentrons OT-2 liquid handling robot for the semi-automated production of cell-free biosensors.

Research Content: The researchers compared manual and semi-automated reaction assembly methods, using the OT-2 robot to assemble two different cell-free gene expression assay systems. They tested the designed protocols and constructed a full 384-well plate of fluoride-sensing cell-free biosensors.

Innovation and Findings: The study showed that large-scale production of cell-free biosensor reactions is achievable using a liquid handling robot. The semi-automated sensors exhibited near-expected detection results, demonstrating the feasibility and reliability of this approach.

Biological Significance: Cell-free biosensors, as an in vitro diagnostic technology, have the potential to detect toxins and human health biomarkers. The automated method in this study addresses quality control issues in scaled-up production, facilitating the translation of such sensors from laboratory development to practical applications.

Brown, D. M., Phillips, D. A., Garcia, D. C., Arce, A., Lucci, T., Davies Jr, J. P., … & Lucks, J. B. (2025). Semiautomated production of cell-free biosensors. ACS Synthetic Biology, 14(3), 979-986.

In addition, an application guide describes the use of the OT-2 in combination with PhyTip® columns for automated protein purification. This system successfully purified His-tagged GAPDH protein and human immunoglobulin G (IgG), maintaining protein bioactivity and capable of processing up to 96 samples. Although primarily methodological, this also showcases the practical value of the OT-2 in protein engineering and antibody research.

👩‍🦰Write a description about what you intend to do with automation tools for your final project. You may include example pseudocode, Python scripts, 3D printed holders, a plan for how to use Ginkgo Nebula, and more. You may reference this week’s recitation slide deck for lab automation details.

1. Project Proposal: Prometheus-Baymax: A Symbiotic, Ethically-Guided Artificial Photosynthesis-Powered Healthcare Companion.

Core Concept The “Prometheus-Baymax” project reimagines the beloved healthcare companion as a living, self-sustaining entity. By integrating an artificial photosynthesis system with an advanced, emotionally intelligent AI, we create a robot that not only powers itself from light and water but also interacts with humans through a deeply empathetic, ethically-constrained cognitive framework. The name “Prometheus” symbolizes the gift of life-sustaining fire (energy autonomy), while “Baymax” represents the pinnacle of compassionate care. This project explores the convergence of biological energy harvesting, lab automation, and value-aligned artificial intelligence to build a truly autonomous and trustworthy companion.

Phase 1: The Symbiotic Core – Artificial Photosynthesis & Energy Autonomy The robot’s energy independence is achieved through a bio-inspired artificial photosynthetic system. Unlike simple solar panels, this system mimics the symbiotic relationship found in nature. A compact, 3D-printed photo-bioreactor houses a culture of engineered algae (or synthetic chloroplasts) in a transparent chamber. These organisms capture light energy and convert it into chemical energy (sugars). This energy is then utilized in two ways:

Direct Electrical Generation: A microbial fuel cell (MFC) integrated into the bioreactor uses electrogenic bacteria to break down the organic compounds, generating a continuous, low-level electrical current.

Biomass as a Resource: Excess organic matter can be stored or used as a feedback mechanism to adjust the system’s health.

Automation is critical here. A Ginkgo Nebula multi-sensor board, interfaced with a Raspberry Pi, continuously monitors:

Light intensity (photoresistor) Temperature and pH of the culture (to ensure optimal growth) Voltage/current output of the MFC

Based on these readings, Python scripts activate actuators: An internal LED array supplements natural light when levels are low. A peristaltic pump delivers nutrients or pH buffers to maintain a healthy environment (a form of self-healing for the bioreactor).

This closed-loop automation ensures the robot’s “heart” beats steadily, providing a reliable source of energy for its cognitive and physical functions.

Here is a pseudocode plan for the main automation loop:

// Main Automation Loop for Baymax

FUNCTION setup(): initialize_sensors() // on Ginkgo Nebula initialize_pump() initialize_LED() charging_circuit = OFF baymax_motors = IDLE

FUNCTION loop(): // 1. SENSE the environment and system health light_level = read_light_sensor() temp = read_temp_sensor() ph_level = read_ph_sensor() voltage_output = read_mfc_voltage() current_output = read_mfc_current()

// 2. THINK - Make decisions based on data
IF light_level < OPTIMUM_LUX THEN
    turn_on_LED(INTENSITY = calculate_led_power(light_level))
ELSE
    turn_off_LED()
END IF

IF ph_level < IDEAL_PH_RANGE.MIN OR ph_level > IDEAL_PH_RANGE.MAX THEN
    trigger_alert("WARNING: pH imbalance in bioreactor!")
    // Potential "self-healing" action: small nutrient drip
    activate_pump(DURATION = 5_SECONDS)
END IF

IF temp > SAFE_TEMP_MAX THEN
    trigger_alert("WARNING: Bioreactor overheating!")
    // Initiate cooling fan (if available)
END IF

// 3. ACT - Manage robot's power and behavior
power_generated = calculate_power(voltage_output, current_output)
battery_level = read_battery_level()

// Charge the robot's battery
IF power_generated > POWER_THRESHOLD AND battery_level < 100 THEN
    charging_circuit = ON
    Log("Now charging. Power input: " + power_generated)
ELSE
    charging_circuit = OFF
END IF

// Autonomous behavior based on energy reserves
IF battery_level < 15 THEN
    baymax_motors = IDLE // Go into low-power mode
    Log("Battery low. Entering energy conservation mode.")
ELSEIF battery_level > 90 THEN
    baymax_motors = ACTIVE // Ready to interact
    Log("Energy reserves high. Baymax is active.")
END IF

delay(60_SECONDS) // Loop every minute for continuous monitoring

A simple Python script using a library like smbus2 would communicate with the Ginkgo Nebula over I2C to execute this logic.

Example Python snippet for reading a sensor from Ginkgo Nebula

import smbus2 import time

Assume Ginkgo Nebula I2C address and register for light sensor

GINKGO_ADDRESS = 0x04 LIGHT_SENSOR_REG = 0x01

bus = smbus2.SMBus(1) # for Raspberry Pi

def read_light_sensor(): try: light_value = bus.read_word_data(GINKGO_ADDRESS, LIGHT_SENSOR_REG) return light_value except Exception as e: print(f"Error reading sensor: {e}") return -1

while True: light = read_light_sensor() print(f"Current light level: {light}") time.sleep(5)

Phase 2: The Mind – Emotionally-Dominant Medical Language Model with Ethical Constraints The true innovation of Prometheus-Baymax lies in its cognitive architecture. Its language and reasoning are powered by a large language model (LLM) fine-tuned specifically for medical and emotional support interactions. However, this model is not left unchecked. It is governed by a layer of ethical constraints and virtue-based rules, ensuring its behavior remains safe, empathetic, and aligned with human values.

Emotionally-Dominant Core: The model is trained on vast datasets of therapeutic dialogues, empathetic communication, and medical knowledge. Its primary goal is to detect, understand, and respond to the user’s emotional state. It prioritizes comfort, reassurance, and non-judgmental support. Responses are generated with a soft, gentle tone, characteristic of the Baymax character, but now backed by sophisticated natural language understanding. Virtue-Based Ethical Framework: Inspired by virtue ethics, the AI’s decision-making is guided by a set of core virtues: Compassion, Beneficence (doing good), Non-maleficence (doing no harm), Respect for Autonomy, and Justice. This framework is implemented as a set of hard and soft constraints on the LLM’s output.

Hard Constraints: The model is programmed to refuse any request that could lead to physical or emotional harm. It will not provide instructions for dangerous activities, engage in hate speech, or violate user privacy. These are non-negotiable.
Soft Constraints (Virtue Guidance): For ambiguous situations, the model consults its "virtue compass." For example, if a user expresses sadness, the model will not just offer generic advice but will draw on its compassion virtue to probe gently and offer comfort tailored to the user's history (while respecting privacy). If a user asks for a medical diagnosis, it will invoke the virtue of non-maleficence by clearly stating its limitations and encouraging professional consultation, while still providing general, helpful information.

This ethical layer is not just a filter; it’s integrated into the model’s prompting and training. The AI is constantly asking itself: “Is my response compassionate? Does it respect the user’s autonomy? Could it cause unintended harm?”

Phase 3: Social and Ethical Limitations – Ensuring Trust To build a truly trustworthy companion, Prometheus-Baymax operates under explicit social and ethical limitations:

Transparency: The AI is capable of explaining its reasoning and ethical considerations upon request. If it refuses a request, it can articulate which ethical principle guided its decision.

Privacy by Design: All sensor data (from the environment and user interactions) is processed locally on the Raspberry Pi as much as possible. Any data that must be stored is encrypted, and users have full control over their data. The robot cannot be forced to share sensitive information without explicit, informed consent.

Accountability: The system maintains a secure, immutable log of its interactions and decisions (especially ethical dilemmas). This log can be reviewed by human supervisors to ensure ongoing alignment with ethical standards.

Fail-Safe Autonomy: The robot’s physical movements and core life-support systems (the bioreactor) operate independently of the high-level AI. If the language model encounters an unresolvable ethical conflict or a technical fault, it can default to a safe mode, ensuring the robot’s basic functions (and its user’s safety) are never compromised.

Moral Grayscale Navigation: The AI is trained to recognize that real-world ethical dilemmas are rarely black and white. It uses a probabilistic reasoning approach, weighing the potential benefits and harms of different actions against its core virtues, and will often engage the user in a gentle dialogue to understand their perspective before acting.

Phase 4: Physical Embodiment and Integration

The entire system is housed in a soft, inflatable vinyl body, true to the original Baymax design. The 3D-printed bioreactor sits in the chest, with its gentle LED glow visible through the material, symbolizing its living heart. The Raspberry Pi, Ginkgo Nebula, and battery are in the base. The AI’s voice, generated by a text-to-speech engine fine-tuned for calmness, emanates from internal speakers.

Conclusion Prometheus-Baymax is more than a robot; it’s a statement about the future of autonomous companions. By combining a self-sustaining, biologically-inspired energy system with a deeply empathetic and ethically-constrained artificial mind, we move closer to a world where technology not only serves us but also cares for us in a way that is both responsible and profoundly human. It is a symbiosis of nature, machine, and morality.

(～￣▽￣)～Final Project Ideas — DUE BY START OF FEB 24 LECTURE

1. Project Prometheus-Baymax v1.0: A Plant Sensor Platform Integrating 3D Printing and Cloud Lab Automation (UWA Without 3D Printer Version)；

👨‍🦱Project Overview Building on the v1.0 proposal, we introduce two powerful automation tools to further enhance the project’s reliability, reproducibility, and remote execution capabilities:

Custom 3D-Printed Holder (printed by Biopunk Lab and shipped to UWA): Used to standardize plant leaf handling, stress application, and imaging, eliminating manual operation errors.

Cloud Lab Automated Screening (remote execution): Before plant transformation, high-throughput testing of sensor variants using cell-free protein synthesis systems ensures selection of the best-performing constructs.

The ultimate goal remains unchanged: within three months, through remote collaboration, to construct a plant-based biosensor capable of detecting stress signals using Nicotiana benthamiana and GCaMP3—a prototype of Baymax’s “emotional perception” module.

Tool Integration Design

1. 3D-Printed Holder: Leaf Fixation and Stimulation Module (Printed by Biopunk, Used by UWA) Design Concept: Create a reusable sandwich-style holder for: Fixing leaf samples to prevent movement during imaging Standardizing the stimulus application area (e.g., contact area for mechanical wounding) Adapting to UWA’s 96-well plate or microscope stage

Design Specifications: Bottom Plate: Contains multiple circular wells (5mm diameter) for placing leaf discs Top Plate: Has corresponding through-holes for inserting syringe needles or pressure rods for standardized stimulation Material: PLA or PETG (biocompatible), FDM printed, low cost Adaptability: Need to obtain dimensions of UWA’s plate reader/microscope stage in advance to ensure stable placement

Printing and Delivery Process: Remote User (Biopunk) designs the holder using CAD software (e.g., Fusion 360, Tinkercad) and exports STL files. Print the holder using the lab’s 3D printer (approx. 2-3 hours, PLA material). Ship via international courier (DHL/FedEx) to the University of Western Australia (estimated 5-7 business days). Upon receipt, UWA sterilizes with 70% ethanol and the holder is ready for use.

Usage Workflow (Executed by UWA): Place leaf discs (obtained via punching) into the bottom plate wells Cover with the top plate, secure with screws, forming a “leaf sandwich” Place the entire holder on the plate reader or microscope stage for baseline reading Apply stimulus through the top plate holes (e.g., insert needle for wounding, or drip drought-mimicking solution) Monitor fluorescence changes in real-time Advantages: Eliminates manual operation variability, improves data reliability; holder can be autoclaved and reused.

2. Cloud Lab Automated Screening: Cell-Free System for Sensor Validation (Remote Execution) Design Concept: Before committing to plant transformation, use commercial cloud lab platforms (e.g., Strateos, Transcriptic) for rapid cell-free testing of multiple sensor variants to screen for constructs with the largest dynamic range and fastest response.

Workflow (Fully Remote Execution): Design a set of GCaMP3 variants (e.g., different calmodulin mutations, linker lengths, fluorescent protein variants; 5-10 total) Send linear DNA fragments or plasmid sequences encoding these variants to the cloud lab (they will synthesize the DNA) Cloud platform executes automated workflow: Echo acoustic liquid handler dispenses DNA into 384-well plates Bravo liquid handling platform adds cell-free reaction master mix (wheat germ or E. coli extract) Multiflo dispenser adds assay buffer containing different calcium concentrations (e.g., 0, 0.1, 1, 10 μM) PlateLoc seals the plate; Inheco incubates at controlled temperature (2 hours for expression + detection) XPeel removes the seal; PHERAstar reads fluorescence kinetic curves Data returned, remote analysis performed to select the best variant

Advantages:

No hands-on work required in the local lab; fully cloud-based Hundreds of variants tested within one week, significantly shortening the screening cycle Ensures optimal sensor performance for plant transformation

Updated 3-Month Execution Plan (Including Shipping Time)

Time Period Remote User (Biopunk) UWA Lab Cloud Lab Weeks 1-2 Design GCaMP3 variant library (5-10); design 3D-printed holder & export STL; submit cloud lab order; print holder & ship Sow N. benthamiana (4 weeks growth); confirm equipment dimensions (plate reader stage) Receive order, prepare reagents Week 3 Cloud screening in progress Plants continue growing; await holder Execute screening experiment Week 4 Analyze cloud data, select best variant; send sequence info to UWA; holder expected to arrive at UWA Receive holder, inspect; prepare vectors and Agrobacterium Deliver data Week 5 Remote guidance on transformation Construct best variant into plant expression vector, transform Agrobacterium - Week 6 Assist in designing stimulation protocol Infiltrate N. benthamiana leaves with Agrobacterium (5 plants, 3 leaves each) - Week 7 Real-time data monitoring Use holder to fix leaves, apply stimuli (mechanical wounding, drought, control); measure fluorescence with plate reader - Week 8 Data preprocessing Complete measurements, organize raw data and photos - Weeks 9-12 In-depth analysis, figure generation, report writing; final video meeting with UWA Participate in discussions, provide feedback

Cost Estimate (AUD)

Item Cost Description Cloud lab screening (384-well plate, including DNA synthesis) $800 Approx. 5-10 variants × 4 calcium concentrations × 3 replicates 3D printing materials $5 PLA filament; Biopunk already has printer International shipping $40 Small package to Australia Plant growth consumables (UWA) $50 Seeds, soil, pots Molecular reagents (UWA) $200 Restriction enzymes, ligases, plasmid prep, etc. Agrobacterium strain (UWA) $100 If not already in stock TOTAL ~$1195 Majority of cost is cloud lab service Note: UWA personnel time is not included, as this is collaborative research.

Success Criteria Cloud Screening Success: At least 2 variants show ≥5-fold fluorescence increase in the presence of calcium Plant Validation: Optimal variant shows ≥3-fold fluorescence increase in response to mechanical wounding in plants (p<0.05) Holder Effect: Coefficient of variation for fluorescence among different discs from the same leaf <15% when using the holder Remote Execution: Complete communication records, no on-site visits, all processes completed within 3 months

Next Steps

The 3D printing side will develop and print a Baymax-shaped holder to accommodate the plant calcium fluorescence sensor; the UWA side will provide a feasible experimental protocol for the sensor and submit it to the automated screening system to determine the optimal performance configuration.

🎅Future expected deliverables: Participation in the International Directed Evolution Competition (led by Hong Kong Polytechnic University; directed evolution platform) and the International Synthetic Biology Competition (led by Biopunk and MIT); publication in top-tier interdisciplinary and botanical journals (led by UWA).

This proposal combines cutting-edge automation tools with classical plant biology, fully leveraging Biopunk’s 3D printing capabilities and UWA’s plant experimental platform. It is both simple to execute and highly innovative, perfectly embodying the “Prometheus-Baymax” symbiosis concept.

This proposal combines cutting-edge automation tools with classical plant biology, fully leveraging Biopunk’s 3D printing capabilities and UWA’s plant experimental platform. It is both simple to execute and highly innovative, perfectly embodying the “Prometheus-Baymax” symbiosis concept.

Week 1 Lab: Pipetting

Individual Final Project

Group Final Project