When Will Citing Blog Posts be a Norm in Bioinformatics Publishing?

When Will Citing Blog Posts be a Norm in Bioinformatics Publishing?


For many years, bioinformaticians were defining the publishing trend in biology. This started with the influx of physicists around the completion of human genome project. I remember from early 2000s, when my papers with physicists went straight to preprint servers before publication, whereas the papers with biologists had to go through military-level secrecy. Biologists were not ready to share their papers even with close friends due to the fear of “getting scooped”.

Biological preprint server biorxiv is gaining far more acceptance these days, and the researchers are ready to cite preprints not anointed by “official” journals. One classic example is the BWAMEM paper that Heng Li decided to leave as preprint after a row with an anonymous reviewer. Based on Google-search estimates, it currently has 2190 citations.

Recently we are noticing a new trend of citing blog posts and github projects directly. For example, the Marijon et al paper on scrubbing tools mentioned yesterday cited two blog posts from Eugene Myers. Our blog posts also got cited in the past (e.g. by the Polish group that wrote the fantastic kmc tools).

We believe this is a healthy trend. Bioinformatics sits half way between biology and computer science. In fact, it is more applied than traditional computer science, and nobody in the programming world developing practical tools is writing papers to explain them. People find the useful projects through word of mouth, web-search and github scans. The best original reference for the popular Ruby on Rails was a video by its creator David Heinemeier Hansson (this is a copy).

There were skeptics, when the physics preprint server (arxiv) was gaining acceptance in early 1990s. They often asked - “How do we know if a paper is bad if no peer reviewer checks it?” There was fear the preprint server would be filled with so much noise that signals would be undetectable. The answer turned out to be rather simple. Bad papers got rejected just because everyone ignored them, and good papers built solid citation history.

Clearly bioinformatics is heading in the same direction of peer review through use by the actual peers, not some “peer” pulled up from the journal database. Sooner we can get there, the better.



Written by M. //