According to the article An Automated Scientist to Design and Optimize Microbial Strains for the Industrial Production of Small Molecules (pdf here):

Although industrial fermentation of engineered microbes has proven impactful in that it can deliver high purity biomolecules in sustainable fashion at stable cost, the R&D process of bringing a single biomolecule from idea to small scale production (milligrams) to applications testing (kilograms) to market (tons) is lengthy (years to a decade or more) and expensive ($100-$200M), a barrier that prevents many new biomolecules from going to market.

These costs have come down over the decades due to such factors as the technological cost curves in DNA Synthesis and Sequencing, improved machine learning, new genome engineering tools such as CRISPR-Cas9, and rapid tool advances.

On June 3-4, 2024, Metaculus and the Astera Institute held a Scale Is All You Need workshop in Berkeley, California, to identify a set of technical innovations or scientific insights that, if developed, could help resolve the bottlenecks that currently make it difficult for synthetic biology companies to innovate. Operating under the Chatham House Rule, subject matter experts identified the continued high costs of engineering, prototyping, and deployment of engineered organisms as an enormous bottleneck in the synthetic biology industry.

The scientists identified the following potential interventions, if they happened before 2035, as having potential to unlock this bottleneck:

• Greatly-reduced costs of fermentation tests are available before 2035, assuming a sufficiently high Positive Predictive Value is maintained.

• Universal molecular biology tools for use across organisms are available before 2040.

• A model is developed that can predict protein function (catalytic activity, binding affinity, etc.) from any genotype with 90% accuracy before 2035.

• In general, overcoming the significant long-term challenge in synthetic biology of being able to accurately predict phenotype from genotype would be enormous, especially if it could be done with a high degree of accuracy. While the group did not find this to be very feasible, as one participant put it, "Solving genotype<>phenotype would guarantee the 100x drop."

• Before January 1, 2035, at least 10 scalable non-model organisms that grow on waste are developed and available.

• Greatly-increased amounts of publicly-available resources and datasets. One example would be the creation of gene modules, organisms, assay and molecular biology technologies -- and other physical materials -- that are open-source, free to use and commercialize, and easy to access.

See Also

Genetic Engineering & Biotechnology News: Fermentation Margins and Cost of Goods

AlChE: SBE Supplement: Commercializing Industrial Biotechnology - Use Cost Models to Guide R&D

Nielsen J, Keasling JD. Engineering Cellular Metabolism. Cell. 2016 Mar 10;164(6):1185-1197. doi: 10.1016/j.cell.2016.02.004. PMID: 26967285.