Events Graphic
MIT Media Lab about us . academics . sponsors . research . publications . events . people . contact us


Erik Winfree (California Institute of Technology):
Fault-Tolerance in Biochemical Systems: From Molecular-Scale to Avogadro-Scale Engineering

Monday, March 29, 2004, 4:00–5:00 PM EST

Bartos Theatre, MIT Media Lab (E15)

Biochemistry is messy. It's a miracle any of it works, and yet it does, life making life more beautiful: flowers burst forth each spring, antelope bound across the plains, songbirds from South America twitter in Massachusetts forests. The wonderful diversity and amazing talents of living things derive from the biochemical processes that copy genetic information and use that information as a program to construct a sophisticated organization of matter and behavior—reliably and robustly overcoming insult after insult from the environment. Despite enormous progress during the past few decades in our ability to dissect and understand biological systems at the molecular level and in our ability to modify and design molecular components, the principles of biological complexity and robustness are likely to remain a mystery for many years to come.

In this talk, Winfree will first suggest that some insight is to be gained by considering known techniques for fault-tolerant computing and translating them to the biochemical context. The natural place to start is von Neumann's multiplexing technique for digital circuits, which has a direct analog in kinase cascade circuits. Fault-tolerant cellular automata provide an interesting contrast. However, neither of these approaches translates well to the DNA self-assembly systems studied experimentally in our lab. Winfree will describe a new fault-tolerance approach, which he calls "proofreading tile sets," that exploits physical aspects of the self-assembly process such as reversibility and cooperative binding. Although Winfree and his research team's best experimental attempts at algorithmic self-assembly still have 1% error rates, he will conclude his talk by arguing that these techniques, properly applied, should be sufficient for attaining error rates better than 10-12, thus enabling the algorithmic self-assembly of macroscopic objects.

Thomas Burg:
Suspended Microchannel Resonators for Biomolecular Detection

MIT Media Laboratory Home Page | Events Main Index

  . colloquium series
  . livewire
  . events archive