Friday, October 18, 2013

Journal club: On data generation and analysis

This week's paper: Characterization of the early events leading to totipotency in an Arabidopsis protoplast liquid culture by transcript profiling by Chupeau et al. (2013)

I think the experiment is a good idea, and the data could be informative; unfortunately I don't think the authors had the expertise to analyze the data they generated. They probably would have been better off adding another author or two and passing the data to a collaborator with more bioinformatic experience; they may have been able to make a more cogent story.

My observation has been that some (many?) life scientists are quite antagonistic towards bioinformaticians/computational scientists; I've also met some on the computing side of the fence who are quite antagonistic towards the life scientists. The life scientists think the computional folk just push buttons and use meaningless statistics or models. The computional folk think the life scientists are just observational scientists who don't appreciate scientific rigor.

My comment:
"I think the experiment behind this paper is worthwhile and that the methods and data will be useful. However, I also think that the paper demonstrates a common issue in modern genomics research, which is that our data generation capacity frequently outpaces our technical ability to analyze the data in meaningful ways."

No comments:

Post a Comment