To Automate Or Not to Automate? That is the question

To be or not to be, that is the question. Inspired by Hamlet

Sometimes, I write scripts to automate repetitive actions. Although it sometimes takes as long as doing the task manually, I must say it provides a different level of satisfaction. I heard a story about a student who exported almost 1 000 datasets from a program by repeatedly clicking and manually exporting each one. That sounds like a nightmare to me.

I was confronted with more than a hundred X-ray diffraction datasets, all of which needed to be checked because I had almost a continuous variation of protein domain positions across different crystals. I spent quite some time writing a series of scripts that would automatically process, solve structures, collect statistics, and open the structures with electron density maps for visual inspection. Here I said automation to be, that is the answer.

The amazing team from EMBL Grenoble, led by José Márquez, took automation to another level with the CRIMS system, something I couldn't have dreamed of. I am happy that I was using this system during my PostDoc time at EMBL. Luckily, we have got it installed also in University Geneva.

I am very passionate about details and always afraid of missing something crucial, which could be key to solving a scientific problem. I meticulously recheck everything because I know I could always be wrong. So then no automation, is the answer?

I am trying to find a Buddhist-inspired middle way of combining high-throughput processing with careful revision to avoid missing a key element in solving scientific problems.