An alternative protein data lake could contain anonymized data from processing runs across many manufacturers, informing processing improvements and aiding process failure troubleshooting.
A number of cellular processes occurring after slaughter are known to affect the quality and sensory properties of conventional meat. Cultivated meat will offer unprecedented control over these parameters and therefore over the quality of the final product, but it is critical to understand exactly how post-harvest processes for cultivated meat can or should differ from post-slaughter processes in conventional meat. This research can enable subsequent innovations in bioprocess design, media formulation, cell line development, or harvesting techniques to confer consistently high levels of meat quality from cultivated meat processes.
Development of humanely-sourced and thoroughly documented and characterized cell lines from a variety of common food species—together with a mechanism for licensing and distributing these lines to researchers and companies—will remove a key barrier to entry into the field of cultivated meat. In addition, development of open-access, standardized protocols for performing cell isolation from a variety of source tissues and establishing robust cell lines will streamline the processes for those who do end up needing to perform their own isolation and cell line establishment.
Plant-based food manufacturers often struggle with batch-to-batch ingredient inconsistency and variability between suppliers. Better analytical tools for predicting plant-based ingredient performance could improve manufacturing efficiency and create more transparent ingredient markets. Tools are needed to predict how ingredients will perform after various processing methods and in end-product applications like plant-based meat and dairy.
Processing crops into flours, isolates, and concentrates often relies on chemical and mechanical methods. Biological processing techniques may impart the desired composition and molecular structure for optimal functionality with increased precision, lower cost, and greater suitability for small-scale processing. Biological processing techniques include using enzymes to fine-tune functional properties like solubility, gelling capacity, and fat- and water-binding capacity or using microbial fermentation to convert plant protein feedstocks into more functional forms.
Animal cell metabolism within cultivators can produce useful co-product side streams that provide monetary value to the manufacturer while creating a novel source of inputs for other industries. Potential side streams should be identified and analyzed for their utility and economic viability, in addition to developing methods for efficient side stream capture.
Techno-economic models are critical for process design and cost of goods projections. Open-access models based on generalized or exemplar processes with standardized unit operations and designs can form the foundation for individual companies’ work, reducing duplicative effort. Furthermore, techno-economic models can identify key cost drivers and opportunities for process improvements to guide future research efforts. The independent research consultancy CE Delft recently published a cultivated meat techno-economic analysis. However, similar efforts are needed for fermentation-derived and plant-based meat production.
After identifying specific target molecules or desired functionalities in animal-derived foods, scientists can work backward, mining microbial sequences for candidate molecules in the microbial realm that might provide similar functionality. This process can also elucidate the pathways that produce these molecules and inform strategies for designing microbial strains that produce these molecules at scale.
There is a need for deeper fundamental research on the relationships between protein sequence, structure, functionality, and ultimately performance in plant-based food products. While several plant-based companies have claimed a competitive advantage from building databases of functional properties and applying machine learning to inform protein selection and formulation, these capabilities remain proprietary and the efforts duplicative. An open-access database could provide functional and characterization data using standardized methods to facilitate direct performance comparisons among proteins and train predictive algorithms.
The availability of more open-access formulations will provide a foundation to enable both academic researchers and startup companies to develop their own customized formulations with far less effort and cost.