Developing computational tools for NMR: The early days of protein NMR

This article describes the personal memories of the author on the early developments of computational methods for protein structure determination in the laboratory of Kurt Wüthrich from 1977 to the 1990's. The computational problem to solve protein structures from NMR structures needed sophisticated computational and mathematical approaches; some of the technical problems were related to the limited memory capabilities of the computer hardware at that time. Competition and collaboration between X-ray crystallographers and NMR was intensely intertwined in the early days of protein NMR, and provided the driving force to establish the NMR method as a second method for high-resolution structure determination. Kurt's vision on the power of NMR and his faith in theoreticians made his laboratory a fruitful place for co-workers with different scientific background.