Paul Gardner
@ppgardne
Bioinformatics, RNA, benchmarking & educating to fund kayaking and fishing. Ōtepoti/Dunedin, Aotearoa/New Zealand. From Whāngārā.
Thanks. This group has certainly managed to contaminate the nutrition space, but also several others with this sort of thing. Some of the authors are publishing 60+ articles a year!
3/8 One repeated error: A PRISMA flow diagram should calculate: Box4 = Box3 − Box3a Instead, every paper calculates Box4 = Box2 + (Box3 − Box3a) The same incorrect formula appears in all 12 papers.
2/8 Across these reviews & diverse I found: • the same unusual phrases ("Humanly excluded", "Too odd") • similar numerical patterns • identical arithmetic errors • no reported Boolean search strings These aren't subjective criticisms—they're directly observable.
1/8 Systematic reviews and meta-analyses are at the top of the evidence hierarchy. They influence clinical practice, policy and future research. So I was surprised to find a series of 12 reviews with remarkably similar PRISMA literature-search figures.
Predicting the impact of variation in ncRNAs is difficult. With little training/test data it remains an under-developed area. This SGE screen is one of the few datasets that can meaningfully benchmark approaches. Here’s my quick attempt after an afternoon (& 20+ years working on ncRNAs):
More sketchy articles discovered by my 4th year class. Image duplication and the same errors in each of these reviews: link.springer.com/article/10.1... www.sciencedirect.com/science/arti... www.sciencedirect.com/science/arti...
I set my 4th year biochemistry class to work screening some papers being published by a fishy group of authors. Check out the numbers in these "PRISMA" style plots:
Anyone that thinks email is a secure means of communication is kidding themselves.