Laboratory automation strategies have vast potential for accelerating discovery processes. They enable higher efficiency and throughput for time-consuming screening procedures and reduce error-prone manual steps. Automating repetitive procedures can for instance support chemists in optimizing chemical reactions. Particularly, the technology of DNA-encoded libraries (DELs) may benefit from automation techniques, since translation of chemical reactions to DNA-tagged reactants often requires screening of multiple reaction parameters and evaluation of large numbers of reactants. Here, we describe a portable, automated system for reagent dispensing that was designed from open source materials. The system was validated by performing amide coupling of carboxylic acids to DNA-linked amine and a micelle-mediated Povarov reaction to DNA-tagged hexahydropyrroloquinolines. The latter reaction required accurate pipetting of multiple components including different solvents and a surface-active reagent. Analysis of reactions demonstrated that the robotic system achieved high accuracy comparable to experimentation by an experienced chemist with the potential of higher throughput.

No abstract

Model-assisted prediction methods depict an important tool in the development of digital bioprocess twins. However, this requires reliable predictions as informative as possible to support process design, development of process strategies, and decision making during running processes. Challenges in this context are large time steps between measurements, measurement deviations, and biological variabilities, leading to difficulties in parameter estimation and prediction of reliable outcomes.

No abstract

No abstract