Wednesday, September 23, 2026

The difficulty of replicating scientific methods - a personal story

 Reposting from LinkedIn.  Someone I followed posted a paper about how hard it is to replicate biology experiments. 

For my master's thesis, I needed to make SSR markers for quinoa and then use them for measuring diversity in the germplasm and breeding populations and for building a genetic map. I used a method from a paper for making my library that used blunt-end cuts and ligation with enzymes, but I didn't know that the temperature needed to be higher or else you get chimaeric sequences in the plasmids.

This was found after I graduated by a reviewer that thought some of the markers were behaving oddly. We retracted the paper and my professors did additional sequencing and found a high percentage of chaemeric sequences, which I couldn't have seen since when I did the work there were no quinoa sequences to compare with. One of my professors was upset that we had to retract two papers and redo my microsatellite markers, in part because he was up for tenure and was not approved. He accused me of covering up the mistake in my thesis and academic dishonesty. I had to return to BYU on my own time and money to redo a significant amount of work and to defend myself. I replicated all of the key experiments in a short time and showed where the mistake came from and the department chair and other professors were on my side. I returned to TAMU and finished my PhD with overall a good relationship with my previous professors. Such a stressful situation that come from a very simple mistake.

It was a huge deal at a tough time in my life, all because I didn't know what temperature to set the water bath. After this I always contact authors for methods I want to replicate from their paper and recommend students to travel and work temporarily in other PI's labs to learn the nuances in techniques.

Biology experiments are not always replicate-able for lots of reasons, but it is tough to squeeze all of the details of experiments that take years into the few hundred words of a materials and methods section of a publication.

No comments: