A major bottleneck for protein structure determination using x-ray diffraction is the difficulty in finding conditions for crystallization and optimizing those conditions for the production of high-quality crystals. Hundreds of crystallization trials are typically necessary to find optimized crystallization conditions, but these are limited by the amount of protein sample available. Our goal was to increase the number of crystallization trials while at the same time reducing the amount of protein sample required, all in a high-throughput mode. These design goals suggested automation of all key protocols.Begun in the spring of 1998 at the request of personnel now at NIFG, the BioInstrumentation Group at LBNL undertook to increase the throughput of protein crystallization trials by an order of magnitude while reducing protein usage per trial by 90%. In order to achieve this goal, two completely automated robotic systems were designed and constructed with initial testing of the completed system in the spring of 1999. Joining a combinatorial approach with parallel processes, protein samples are screened against a coarse array of up to 480 unique solutions at a rate of eight seconds/trial. This compares with a rate of one to two minutes/trial either by commercial robots or manual methods. Along with the increased throughput, protein usage was reduced up to 50-fold where only 20 nanoliters of protein/trial were needed to grow crystals compared to current best practices using one microliter of protein. Based on sub-optimal growth conditions identified by hits from the coarse screen, a second system narrows in on the optimal conditions by synthesizing a new two-dimensional array of solutions out of 72 stock chemicals and new crystallization trials begin. By automating crystal growth trials, effort previously spent on manual trial preparation is freed to work on other aspects of proteomic research.

This music score was submitted for the Kaleidoscope 2020 Call for Scores, an open access collaboration with the UCLA Music Library.

Because the components of genetic networks are present in small quantities, the detailed behavior of network motifs are better described by stochastic models than by macroscopic kinetics. An exact solution is possible for a gene regulated by its own proteomic atmosphere, showing a dependence in the adiabaticity parameter measuring the relative speed of the binding/unbinding and synthesis/ degradation events. This solution deviates from the commonly used approximation based on the equilibrium law of mass action. The dynamics of small networks are shown to depend on the average number of proteins in the system, the repression strength and the adiabaticity parameter. Stochastic simulations of toggle switches show that strongly repressed systems are better candidates for "good switches". The discrete nature of these systems also defines the behavior of systems where the molecular species are on the verge of extinction, the death and resurrection of the species greatly modifying the attractor landscape. Deterministic models and the diffusion approximation to the master equation break down at the limits of protein populations, in a way very analogous to the breakdown of geometrical optics that occurs at distances smaller than one wavelength of light from edges. Our arguments suggest that the stability of lysogeny in the \[lambda\]-phage may be influenced by such extinction phenomena. The role of stochasticity and noise in controlling genetic circuits is investigated in the context of transitions into and from competence in Bacillus subtilis. Recent experiments have demonstrated that bistability is not necessary for this function, but that the existence of one stable fixed point (vegetation) and an excitable unstable one (competence) is sufficient. Stochasticity therefore plays a crucial role in this excitation. We consider the importance of the protein binding/unbinding to the DNA as a noise source. A theoretical model that includes this "nonadiabatic" mechanism appears to produce a better agreement with experiments than models where only the adiabatic limit is considered. The knowledge gained with the study of smaller motifs can be extended to larger networks. The network controlling the Bacillus subtilis decision between competence and sporulation is analyzed in the light of modularity.

Like any doctoral student, I wanted to write a compelling dissertation with meticulous archival research. The archive is often defined as the state’s official depository of administrative documents. In my case, the National Archives in Egypt has somewhat mythic status: an object of perennial desire that can prove frustratingly difficult to access. And yet, there is a frequently expressed fear among students that a dissertation written without this experience is insufficient. I had plans to use the National Archives to reconstruct a history of urban water. Instead, things turned out differently. When Covid-19 hit in 2020 I had a two-year-old son, and my daughter was born in July of that same year. Caring for young children among myriad covid restrictions foreclosed a return to Egypt, official security clearances in hand or otherwise. It was imperative at that stage to make new plans.

This music score was submitted for the Kaleidoscope 2020 Call for Scores, an open access collaboration with the UCLA Music Library.

This music score was submitted for the Kaleidoscope 2020 Call for Scores, an open access collaboration with the UCLA Music Library.