Abstract
High-throughput crystallisation requires the rapid and accurate dispensing of protein and precipitating agent solutions at nanovolumes, but does not end there. The choice of the initial screens is very important, especially with respect to the availability of protein material. Data from previous crystallisation experiments that are scattered in the literature and only partially available in databases have to be analysed in efficient ways that will maximise their utility for designing new screens. A larger portion of crystallisation parameter space should be made accessible to screening, through the use of nucleants and seeding. Observation, assessment and scaling up of the crystallisation trials should be efficiently performed and, finally yet importantly, optimisation of conditions must also be adapted to the high-throughput environment. The above requirements are briefly addressed in the following paper.
Keywords: Crystallisation robots, crystallisation screening, high-throughput, nucleant, seeding, purified protein, structural biology, multi-parametric space, crystallisation setup techniqu, sparse-matrix approach