Technology Behind Lattice Automation
Alberto Sangiovanni-Vincentelli. Quite a mouthful. Particularly for someone from Michigan. Good thing he let me just call him “Alberto”. Alberto was my advisor at UC Berkeley. He was (is) at the forefront of “Platform Based Design” which is an approach to the design of embedded electronic systems. Lattice was started with this philosophy and in that sense is very uniquely positioned to think about problems in a very modular, formalized way. Lattice has very strong ties to academic research (I am a professor at Boston University at the end of the day). Research is in Lattice’s DNA. I probably should have called the company Lattice Research instead of Lattice Automation.
I have a lot of people ask me about our technology and I inevitably have to forward them a bunch of papers. This post is my way of capturing the top 10 papers from my research that represent what Lattice is all about. Without further ado…
Honorable Mention — Automated robotic liquid handling assembly of modular DNA devices. JOVE, 2017
This paper actually has two Lattice authors (Joshua Timmons and Lloyd McCarthy). A good overview of what it takes to get liquid handling robotic automation to work in a very practical manner and presents a very prototype design application built to user specifications. Good example of a computational and experimental collaboration between organizations.
10. Bio-design automation: software+ biology+ robots. Trends in Biotechnology, 2014
A position paper. Discusses my philosophy on bio-design automation. Explicitly exposes my view that “Specification” is an equal part of the design-build-test cycle. This separation is a key part of Platform-based Design. Credit to Swapnil Bhatia for the catchy title.
9. Developer’s and User’s Guide to Clotho v2. 0: A Software Platform for the Creation of Synthetic Biological Systems. Methods in Enzymology, 2011
In general Clotho was poorly published (sigh…). This continues to be a challenge for me with large scale software engineering efforts (hence the birth of Lattice). This is not a great paper and is already VERY dated but it gives you some idea of the concepts Clotho was trying to establish. The most important part is that it exposes the data types present in Clotho’s data model and how they work together. You can see data elements that SBOL has incorporated as well. Having this data model explicitly exposed was Clotho’s key contribution to the field.
8. DeviceEditor visual biological CAD canvas. Journal of Biological Engineering, 2012
In 2009 the UC Berkeley Software iGEM team created a tool for Clotho called “Spectacles”. This was a drag-and-drop environment to design composite DNA part based designs. This has been done “better” now by programs like SBOL Designer. This work also seeded some of the tools created by Teselagen. It is unique in that it has a visual canvas that can be controlled by a domain specific language (Eugene). I personally do not feel that visual environments like this are the future. I don’t think humans visually designing things is going to 1) scale and 2) be semantically rich enough to specify designs. However this does provide a small window into how this style of approach got started.
7. Merlin: Computer-aided oligonucleotide design for large scale genome engineering with MAGE. ACS Synthetic Biology, 2016
Merlin is a tool to design oligos for Mage (a Church Lab technology). This paper is another demonstration of a collaboration between groups and the formalization of a problem that had a very naive original implementation. I would have liked to follow this up with more experimental work. This remains an open problem and one that several of our customers have been interested in exploring.
6. Double dutch: A tool for designing combinatorial libraries of biological systems. ACS Synthetic Biology, 2016
An example design of experiments based approach. This work has a nice open source set of tools to go with it. The paper does a decent job of also outlining the algorithms in the main body of the paper and presenting the results over a couple of case studies. Really nice work lead by Nic Roehner. The name is a play on words since the the statistical software JMP was used (i.e. jump rope).
5. An end-to-end workflow for engineering of biological networks from high-level specifications. ACS Synthetic Biology 2012
This work is often referred to as the “TASBE project” and was joint work done with BBN Technologies and MIT as part of DARPA work that preceded the 1000 Molecules and Living Foundries programs. Lead by Jake Beal at BBN, this was one of the first end-to-end tool-chains demonstrated for synthetic biology. Most of the software has been replaced with more sophisticated approaches but it was a nice holistic approach for its time.
4. Algorithms for automated DNA assembly. Nucleic Acids Research, 2010
Elegant dynamic programming formulation for binary assembly of DNA parts. Evan Appleton’s follow up paper (“ Interactive assembly algorithms for molecular cloning”) is likely more complete and has a software tool called “Raven” to back it up. Also this work was surpassed in some regards by “Heuristic for Maximizing DNA Reuse in Synthetic DNA Library Assembly”. However, this paper laid the foundation and likely is superior in its simplicity and clarity. This also is my last real first author technical paper so I have a soft spot for it.
3. Genetic Design via Combinatorial Constraint Specification. ACS Synthetic Biology, 2017
This paper is criminally under recognized in my opinion. A truly elegant solution to representing genetic designs by Swapnil Bhatia. The formalism lets you both generate and verify designs. It also is amenable to numerous formal operations that are correct-by-construction. Storage in graph based databases is also natural. This type of approach is the future of design specification. Lattice tools all will support this data model and its associated operations.
2. Eugene–a domain specific language for specifying and constraining synthetic biological parts, devices, and systems. PLoS One, 2011
Lots of great concepts here. The idea of a programming language that allows you to specify genetic constraints while providing the control flow statements that allow you to make decisions on those constraints is very powerful. This work is still ahead of its time. Ernst Oberortner built other publications on top of this which are more formal but this is the original work. The funny thing is that as one of my most cited papers, I often get international students reaching out to me as Dr. Bilitchenko.
- Genetic circuit design automation. Science, 2016
A paper that truly puts it all together. Extremely rigorous experimental work, cutting edge computational work, true inter-university collaboration, and inter-disciplinary expertise. All of this is tied together by an air tight vision of how parts based synthetic biology should be done. This is the paper Chris Voigt and I envisioned when we first met in 2007. This work serves as the foundation as well for the creation of Asimov. The Cello design software represents a true separation of specification and design.
Douglas Densmore
President
doug@latticeautomation.com
