Cultivated Meat Needed a $500M Factory. Molecular Farming Needs a Field.
Moolec Science just expressed bovine myoglobin in pea seeds, proving its molecular farming platform works across multiple crop species. A first-of-its-kind capex comparison reveals why growing animal proteins inside plants could undercut every other alternative protein approach by orders of magnitude.
Zero. That is how many new factories molecular farming requires to produce animal proteins at commercial scale.
On April 7, Moolec Science announced it had successfully expressed bovine myoglobin in pea seeds (Pisum sativum), confirming stable expression across multiple generations after 28 months of sponsored research with an unnamed U.S. academic institution. This is the second crop species in Moolec's platform, after soybeans, and it changes the competitive calculus for the entire alternative protein industry.
Here is why. Cultivated meat companies spent $3 billion trying to build bioreactor-based production at scale. Most of them are now dead. Precision fermentation works but still needs purpose-built steel tanks costing $100-200 million per facility. Molecular farming skips all of that by engineering crops to produce animal proteins directly in their seeds, then processing those seeds through the same crushing and extraction infrastructure that already handles 69.3 million metric tons of soybeans per year in the United States alone.
Three Paths, Three Cost Structures
Nobody has published a direct capex comparison across all three alternative protein production methods. So I ran one, using publicly available data from company filings, USDA records, and the American Soybean Association.
| Metric | Cultivated Meat | Precision Fermentation | Molecular Farming |
|---|---|---|---|
| Facility cost | $250-500M | $100-200M | ~$0 incremental |
| Annual capacity (tons) | 5,000-10,000 | 5,000-20,000 | 693,000+* |
| Capex per ton/yr capacity | $25,000-50,000 | $5,000-40,000 | Approaches $0 at scale |
| 2025 investment | $74M (-48% YoY) | $357M (-43%) | Part of $450M plant-based (+39%) |
| Production status | Multiple bankruptcies | Shipping (mozzarella, egg whites) | USDA approved, expanding platforms |
*If 1% of existing U.S. soy crush capacity were allocated to molecular farming varieties. Source: American Soybean Association, Scott Gerlt (Chief Economist). Current capacity: 2.55 billion bushels/year across 68 solvent crushing plants, growing to 2.78 billion bushels by 2030.
Read that right column again. Molecular farming's incremental facility cost is essentially zero because the crushing plants already exist, the fields already exist, and the combines already exist. Moolec operates its own 10,000-ton/year facility in Argentina and has a supply contract with Bioceres for 15,000 tonnes of soybeans. It did not need to raise $500 million for a purpose-built factory. It needed to develop the seed genetics, a one-time R&D cost of $30-100 million that amortizes toward zero at scale.
Showing the Math
Soybeans are roughly 40% protein by weight. Total soluble protein represents about 25-30% of seed weight. Moolec's Piggy Sooy line expresses porcine myoglobin at 26.6% of total soluble proteins, four times its initial projections. Running the numbers: recombinant animal protein = 26.6% × 27.5% (midpoint of soluble protein range) ≈ 7.3% of seed weight.
Per metric ton of molecular-farmed soybeans: about 73 kg of animal protein. At 2025 average soybean prices of ~$370/metric ton, the raw feedstock cost for that animal protein is $370/73 kg ≈ $5.07/kg.
For context, conventional pork retails at $4-8/kg, but that price includes an entire animal, with all its feed conversion inefficiency, water consumption, land use, and slaughter infrastructure. Molecular farming delivers the target protein directly, with soybeans as the only input. No growth media at $50-400 per liter. No sterile environments. No bioreactors.
Caveat: This $5.07/kg figure represents raw feedstock cost only. Downstream extraction, purification, and food-grade processing add significant cost. Nobody has published fully loaded production costs for molecular-farmed proteins at commercial scale, because nobody is operating at that scale yet. Treat this as a floor, not a final price.
Why Peas Matter
Moolec's April 7 announcement extends beyond soybeans for two reasons. First, peas are a different crop family, proving the underlying genetic engineering platform is transferable rather than a one-crop trick. Second, peas carry different regulatory and consumer profiles. Pea protein is already a $2.5 billion market growing at 12% annually, and peas are not a top-8 allergen in the U.S. or EU. For food manufacturers worried about soy allergen labeling requirements, a pea-based molecular farming ingredient offers a cleaner formulation path.
Moolec is not alone in this space. Alpine Bio (formerly Nobell Foods) engineers casein production in soybeans for cheese applications. PoLoPo in Israel is targeting egg white proteins (ovalbumin) in potatoes. Pigmentum uses Romaine lettuce. Each company bets on a different host crop, but all share the same structural thesis: use photosynthesis instead of fermentation as the production engine.
The GMO Wall
Before molecular farming bulls start counting profits, they should talk to the Golden Rice people. Golden Rice, engineered to produce beta-carotene to combat vitamin A deficiency that the WHO estimates affects 190 million preschool children globally, took 20 years to reach regulatory approval and still faces active opposition in the Philippines and Bangladesh, countries where children are going blind from the deficiency it was designed to prevent.
Molecular farming soybeans contain animal DNA. Let that phrase sit for a moment. It triggers both anti-GMO and animal-rights constituencies simultaneously. Non-GMO Project has already labeled the approach "synbio on steroids" and flagged contamination risks to conventional and organic supply chains. USDA APHIS cleared Moolec's Piggy Sooy as "unlikely to pose increased pest risk" in April 2024, meaning it can be planted and transported without permits, like any other soybean. But APHIS evaluates plant pest risk, not food safety. FDA still needs to evaluate the food product, and that review has no published timeline.
Cross-pollination containment presents another challenge. Soybeans are largely self-pollinating (outcrossing rates below 1%), which limits gene flow risk. But "limited" is not "zero," and coexistence protocols for GM and non-GM soy are already a point of friction in grain markets. Adding animal genes to the mix raises the stakes.
What We Don't Know
There are gaps in this analysis that deserve honest accounting. Moolec's 26.6% expression level comes from controlled research conditions; field-scale data across different climates, soil types, and growing seasons has not been published. Expression levels may degrade under environmental stress, and yield penalties from producing non-native proteins remain unquantified at commercial scale. Moolec is pre-revenue or generating minimal revenue from its legacy safflower-derived products (chymosin, gamma-linolenic acid). No taste panel data for molecular-farmed proteins is publicly available. The $3.5 billion market forecast comes from a single analyst firm (QY Research). Our capex comparison uses ranges and estimates where exact figures are proprietary, and the cultivated meat facility cost range is derived from Believer Meats' $150M production center and UPSIDE Foods' scale-up projections.
What You Can Do
If you work in food manufacturing: Watch for Moolec's FDA food safety submission, likely 2027. If it clears, pea-derived bovine myoglobin could be the cheapest heme protein source on the market. Start evaluating reformulation feasibility now.
If you invest in alternative proteins: Rebalance your mental model of the sector. Investment data from the Good Food Institute shows cultivated meat collapsed to $74 million in 2025, down 48% year-over-year. Meatable, Believer Meats, and Aqua Cultured Foods all folded. Plant-based (which includes molecular farming) rose 39% to $450 million. Follow the capex math, not the hype cycle.
If you are a consumer: Products made with molecular-farmed ingredients are at least 2-3 years from U.S. grocery shelves, pending FDA review. When they arrive, they will likely appear first as ingredients in processed foods (burger patties, sausages, cheese) rather than as standalone items. Look for them in the plant-based section, not the meat aisle.
The Bottom Line
Alternative protein's first two attempts followed a familiar playbook: build expensive, purpose-built infrastructure and hope scale drives costs down fast enough to survive. Cultivated meat failed that bet. Precision fermentation is grinding through it. Molecular farming asks a different question entirely: what if the infrastructure already existed, and all you needed was a better seed? Moolec's pea announcement suggests the platform is real and expanding. Whether consumers and regulators agree is the $3.5 billion question. But on raw economics, growing animal proteins in a field beats growing them in a tank. It is not even close.