At Pennsylvania State University, Costas Maranas and his team are rewriting the blueprint of biology itself—teaching plants and microbes to manufacture the materials we've long relied on oil to produce. His research suggests that synthetic biology, powered by artificial intelligence and computational modeling, could fundamentally reshape how we create fuels, plastics, and chemicals worldwide, offering a genuine path toward cost-effective alternatives to petroleum-based products.
The stakes are global. As markets shift and the pressure to abandon oil intensifies, industries are desperately seeking affordable ways to produce biorenewable materials at scale. That's where Maranas, the Robert V. and Gloria H. Waltemeyer Chair and Donald B. Broughton Professor of Chemical Engineering at Penn State, enters the picture. His research specializes in metabolic modeling—a process that reads an organism's genome like a detailed instruction manual, identifying every enzyme coded in its DNA. Once scientists understand what chemical repertoire an organism possesses, they can reverse engineer it to produce new molecules, optimize growth, or convert plant sugars into biofuels like ethanol.
Consider the practical application: scientists can bioengineering organisms to overproduce lactic acid, a chemical building block of polylactic acid, which becomes a biodegradable plastic used everywhere from food packaging to medical implants. What once seemed like science fiction—using microbes and plants to manufacture plastics—is now technologically feasible. The remaining challenge is doing it affordably enough to compete with conventional petroleum-based manufacturing.
This is where data science becomes transformative. Maranas' team has developed CatPred, an online tool that uses artificial intelligence and large datasets to predict how effectively different enzymes can react with different biochemicals—a process called chemical retrosynthesis. In agriculture, multispectral imaging from drones and satellites reveals crop health by analyzing different wavelengths of light reflected from plant fields, generating enormous datasets that feed machine learning models. Each innovation generates what researchers call "big data," vast information pools that would be impossible to process without computational power.
The emergence of artificial intelligence has accelerated breakthrough after breakthrough. Specialized protein language models now operate similarly to ChatGPT, except instead of a vocabulary of words, they work with a vocabulary of amino acids—allowing researchers to identify which protein sequence will perform best for a specific biochemical task. Traditional AI tools are being repurposed to design enzymes that bind perfectly to target molecules and react in precise ways. The speed of discovery has become almost dizzying; across laboratories worldwide, new tools emerge regularly, each one edging closer to making biorenewable production economically viable.
What makes this moment genuinely hopeful is the intersection of three converging forces: breakthroughs in synthetic biology are making organism redesign possible; computational models are making that redesign predictable; and artificial intelligence is accelerating the entire process. These aren't theoretical possibilities anymore. Maranas and researchers like him are building the infrastructure—the models, the tools, the data pipelines—that will allow industries to replace oil-dependent production with biology-based alternatives. As these tools continue developing across competing labs worldwide, the question shifts from "can we do this?" to "how quickly can we scale it?"