Establish optimal manufacturing processes from the unit-operation to the integrated-screening scale.

Engineering Biology

Data Science Goal:

Establish optimal manufacturing processes from the unit-operation to the integrated-screening scale.

Current State-of-the-Art

Current state-of-the-art capabilities for generalized biofabrication reside primarily in large, well-established organizations (such as the Broad Institute) and biotechnology companies (such as Zymergen). The state-of-the-art, however, is still a somewhat ad-hoc assemblage of product-oriented tools, customized software local to that institution, proprietary data sets, custom automation solutions, and customized data-logging and analysis systems. Often, industry views its proprietary approach to data flow and informatics as unique and as a large part of their value-proposition and tends to sequester informatics gains to particular institutions. However, to address this, there is a rapidly growing number of public-funded, non-commercial bio-foundries, which has recently resulted in the establishment of the Global Biofoundries Alliance.¹ The aims of the Alliance are to establish open technology platforms that will allow the sharing of automation workflows and protocols, software, reference materials and best practices which may lead to new standards for measurement and data, as well as global capacity for establishing optimal manufacturing processes for synthetic biology.

Breakthrough Capabilities & Milestones

Standardized informatics tools, data, and automation platforms for efficient and collaborative use and integration of data in order to develop novel products more quickly.

Establish communications and networks to develop democratized platforms for data exchange and automation across industry and academia.

Bottleneck/Challenge: Lack of standards for data exchange and communication, lack of standards of automation platforms, and extreme cost of automation for implementation in non-industrial settings.

Potential Solution: Dialogue between industry and academic scientists to develop standardized, cheaper, automated platforms for high-throughput experimentation of commonly used microbes.

Potential Solution: Development of a greater number and better-connected industry-academic consortia to share ideas, equipment, and platforms.

Democratized platform for data exchange related to standard/model microorganisms.

Bottleneck/Challenge: Lack of standards for exchange of design information and communication with automated systems for both build-execution and test-data-acquisition for commonly used microbes.

Potential Solution: Launch new industrial-academic consortia or partnerships that leverage shared automation technology and platforms.

Potential Solution: Start leveraging consortia to develop industry- and academia-wide data-logging, -analysis, and -sharing standards.

Potential Solution: Create standards-based approach to data exchange with access to integrated, generalized databases.

Potential Solution: Incorporation of design-of-experiments strategies that integrate the economics of obtaining data and required standards for data precision and accuracy.

Potential Solution: Collaborations to develop new data analysis tools specific to biology.

Potential Solution: Miniaturization of automated host engineering and analytical systems and integration into desktop machines able to re-engineer microbes. These machines may act genome-wide, in an automated fashion without human interference, i.e., from design to organism in one iteration.

Initial development of a database of organisms beyond E. coli and S. cerevisiae (i.e., a database of non-model organisms), that leverages existing databases, to integrate predicted pathways, and -omics data that confirm specific production of a compound of interest.

Bottleneck/Challenge: There are many organisms that might be useful in a particular environment or for producing a particular chemical; however, identifying the most useful hosts beyond current model organisms is challenging.

Potential Solution: Leverage existing databases (Biocyc, KEGG, etc.) to construct a database of non-model organisms that focuses on the known functionality of the non-model organism.

Democratized platform for non-model organisms and microbial communities.

Bottleneck/Challenge: Lack of standards for exchange of design information and communication with automated systems for both build-execution and test-data-acquisition for non-model microbes.

Potential Solution: Extend industry-academic consortia, and platforms/data sharing solutions for model organisms (5 year milestones) to problems associated with non-model organisms.

Potential Solution: Extend microfluidics approaches to experiments with non-model organisms.

Democratized suite of platforms that can be utilized across different model systems.

Bottleneck/Challenge: Full biological characterization of greater libraries of organisms, including transferability of design, modeling, and engineering strategies between organisms.

Potential Solution: Extend industry-academic consortia, and platforms/data sharing solutions for model organisms (5 year milestones) to problems associated with non-model organisms.

Potential Solution: Extend microfluidics approaches to experiments with non-model organisms.

Full machine learning capabilities and ability of algorithms to run greater than 90% of the DBTL+automation cycle.

Bottleneck/Challenge: Current understanding of the principles behind optimal design for complex systems is still limited.

Potential Solution: Improve predictability of complex systems through many years of iterations of current algorithms; much of this solution should happen organically honing of machine learning and artificial intelligence algorithms continues over time, particularly with increased access to standardized data and democratized automation platforms.

Potential Solution: Work to create new data, coding, and analytical languages that better capture the rules of biology.

Footnotes

Hillson, N., Caddick, M., Cai, Y., Carrasco, J. A., Chang, M. W., Curach, N. C., … Freemont, P. S. (2019). Building a global alliance of biofoundries. Nature Communications, 10(1), 2040. View publication.