Understanding and Engineering Biological Complexity
This summer, myself, John Love and Markus Gershater (Synthace) are organising a session at the SEB AGM in Prague. The session is called Understanding and Engineering Biological Complexity. More information can be found on the SEB website. Confirmed speakers include Rachel Haurwitz (Caribou Bioscience), Phil Ramsey (University of New Hampshire), Claes Gustafsson (DNA2.0) and Srividya Suryanarayana (Cytovance Biologics). Registration and abstract submission is open until 24th April.
The session will address the novel approaches to genetic manipulation that synthetic biology offers, and how coupling these tools with effective experimental design strategies can untangle complex biological phenotypes.
DNA cloning and molecular engineering is no longer a significant barrier to experimental design: the cost of DNA synthesis continues to fall; synthetic prokaryotic and eukaryotic chromosomes have successfully replaced their natural counterparts and genome editing and refactoring continues apace. Successfully designing ever larger and more complicated constructs however requires the full knowledge of how each of the component genetic parts work individually, in unison, and in different experimental or environmental conditions. If our aim is to move beyond natural sequence arrangements to synthetic designs, to test our hypotheses and build new products, the obstacle is not how to write the code, but what code to write.
We aim to examine the state-of-the-art of bioengineering, including advances in DNA read/write/editing technologies as well as developments in other enabling technologies (e.g. automation and/or high-throughput analytical procedures). However, despite the increased capabilities conferred by these enabling technologies, biological ‘design space’ is vast. We will therefore consider how to best implement a rational interrogation of this space. We will focus in particular on statistical design of experiments (DoE) methodology for bioengineering. Finally, we wish to discuss future directions and constraints for understanding and engineering biology and biological complexity. We strongly encourage abstracts from a wide range of researchers from academia and industry, and from disciplines from molecular and cell biology, synthetic biology, statistics and computer science, as well as from social sciences and technology watchers.
Importantly, we will also depart from the normal session structure by giving a half day workshop on Modern Design and Analysis of Experiments for Biological Applications Using the JMP® Statistical Software. This will be led by Phil Ramsey (Department of Mathematics and Statistics University of New Hampshire) and will focus on how statistics is an invaluable tool when manipulating something as complex and noisy as biological systems. Figure 1 shows a scenario many experimental biologists will recognise. And this only involves a two-factor interaction. Thankfully, statistical design of experimentation is our friend and can make sure we perform experiments in the most efficient and informative way possible.
This course is designed for those involved with biological research and/or development of bio-processes and systems. Previous exposure to designed experiments and statistical model building is not required.
Characterizing and optimizing complex biological systems requires the use of experimental designs that are capable of estimating main effects, factor interactions, and nonlinear effects. Traditional approaches to characterization have used rather large response surface experimental designs that are costly, time consuming, and sometimes prohibitive in size. In this workshop participants will learn how to use newer, more efficient designs, such as Definitive Screening Designs for characterization and optimization. In addition, participants will also learn how to use more modern approaches to the analysis of experimental data that provide better predictive models for characterization and optimization. All of the topics covered will be motivated by the use of actual biotechnology case studies.
The workshop will be conducted using the JMP 12 statistical software and instructions will be provided in the notes on how to use JMP to create designs, build statistical models, and perform optimization. However participants need not have JMP to participant in the workshop and benefit from the coverage. Please note however that there is a free 30-day trial version of JMP that participants can download. Those who have JMP 10 or higher available should feel free to bring a laptop to the workshop and work along with the instructor as the topics and case studies are discussed – copies of all datasets will be provided.
- Introduction to design of experiments and modern statistical modeling.
- Definitive Screening Designs for efficient experimentation with biological systems.
- Using modern statistical modeling techniques to building more effective models for characterization and optimization.
- Optimizing a single response or multiple responses with experimental data.
- Quantifying process factor importance.
Each participant will be given a complete set of course notes in PDF form. In addition, copies of all datasets in JMP format will be provided.