Why being aware of problematic research should form part of a researcher’s DNA
Circles have always fascinated humans. With no apparent beginning or end, circles signify mystery, continuity and endurance. It’s no surprise that people choose circular objects as some of our most important possessions. Circles can signify positive processes- feedback loops, recycling, the circle of life itself. However, circles can also be associated waste and even futility- circular reasoning, arguments, traffic routes. Circles may be beautiful, but the shortest distance between any two points is usually not a circle, but a straight line.
In human genetics, straight lines have long seemed to be more relevant than circles. Genes are linear sequences of 4 chemicals called nucleotides that are written using the letters A, C, G and T. Used in many different combinations, these 4 nucleotides build long strands of DNA, rather like beads of 4 different colors on very long strings. Humans have approximately 40,000 genes that are ordered along DNA strands. These genes serve as linear instruction manuals, where every person has their own copy of these instructions that will be used throughout their lifespan.
Instructions are very important, but to be effective, instructions need to be accurately communicated. Genetic instructions are communicated by copying parts of the gene onto a related linear molecule called RNA. Most of the time, the right RNA messages are sent to the right place at the right time. However, when RNA messages go astray, important biological processes can be derailed. This can lead to a range of genetic diseases, including cancer.
In this complex and highly regulated system, linear RNA messages have long been assumed to communicate genetic information. Researchers were therefore surprised to discover that some genes also produce circular messages. These circular messages appeared to be linear RNAs whose ends had been fused together to form circles [1]. These circular messages were named circular RNAs, or circRNAs for short [1].
circRNAs were discovered largely by accident and were initially thought to represent rare mistakes. However, over time, circRNAs have grown in numbers. The idea that circRNAs are mistakes has been revised, and researchers are increasingly studying circRNAs as possible “master-regulators” of other genes [1]. More and more circRNA papers are being published [2, 3]. This should all be good news for a research field that could shed new light on how genes communicate their instructions. Nonetheless, there are signs that not all may be well in the field of human circRNA research [4, 5].
We first encountered circRNAs through our publication integrity research [4]. We study the identities of the DNA and RNA reagents (or ingredients) that researchers use to study genes [4, 6]. Like genes themselves, DNA and RNA reagents acquire their precise capabilities through their sequences of the letters A, C, G and T. These reagent sequences need to be accurately written down in research publications, so that other researchers know exactly which gene experiments were carried out.
Because gene reagents are written using only 4 letters, their sequences can look very repetitive. Possibly because gene reagents are so hard to read by eye, we and others have found that published gene reagents can be incorrect [4, 6]. We have found many examples of gene reagents that seem to map to the wrong human gene, or to no human gene at all [4, 6].
Wrongly identified gene reagents could indicate simple issues, such as copying the wrong DNA sequence from a database after the experiments were done correctly. However, wrongly identified reagents could also indicate more serious problems that could compromise the validity of the research [4, 6].
We know from experience that it is difficult to check gene reagent identities, so we try to check reagents for gene research topics that we understand. Nonetheless, publication integrity research can lead to unexpected directions. When we screen publications in particular journals [4, 6], we cannot simply choose to study the papers that we already understand. We need to study every paper if we are to understand how many papers in the journal might be reliable.
Our decision to study papers in high impact cancer journals led us to first encounter papers about circRNAs [4]. We knew nothing about circRNAs before my student Pranujan Pathmendra began to study these papers. I realised that we needed help to understand circRNAs and the types of experiments that are required to study them. We were fortunate be able to collaborate with Prof Franscisco Enguita from Lisbon, Portugal, who served as our guide to this rapidly changing field.
Given the decades of research that have focused on linear RNAs, circRNA researchers need to convince other scientists that they are studying circular messages [1]. In particular, circRNA researchers need to show that they are studying messages where the two ends have been joined to form a circle. Researchers do this by selecting gene reagents that should recognize circRNAs, and not linear messages [1]. Despite this requirement, we found circRNA reagents that seemed to target linear messages [4].
Other circRNA reagents need to target the specific sequences that make circRNAs unique [1]. These sequences occur where the two message ends have been joined together. Despite spelling out the exact junction sequences that should be targeted, we found that some authors described circRNA reagents that didn’t match the junction sequences described in their own papers [4]. This is a bit like getting your home address wrong on an important government document. It’s hard to understand how authors could make such important mistakes, even in complicated publications that describe many different experiments.
So why are these mistakes happening? At the level of individual papers, the answer is of course that we simply don’t know. However, when these papers are viewed as a group, we see common features that we and others have seen before [4, 6]. Perhaps some circRNA papers have serious errors because they were not written by circRNA experts, but by people who are imitating this type of research simply to produce publications. From our experience of learning about circRNA research, it’s probably very, very difficult to invent complex manuscripts about circRNAs and get every single detail right.
So why should we worry about errors in papers that describe poorly understood phenomena such as circRNAs? Well, many scientists are attracted to new topics, particularly where there is a clear twist on an accepted narrative, such as gene messages that form circles as well as straight lines. circRNAs have also been proposed to represent a new way of regulating gene activity [1], so researchers in many different fields could be tempted to study them.
Just as when we first started reading circRNA papers, many researchers who are unfamiliar with circRNAs may not understand the specific requirements of circRNA reagents and experiments. Because circRNAs are still a fairly new research topic, there may not be many local experts available to help. Researchers who are new to the circRNA field may therefore be heavily reliant on published descriptions of circRNA experiments when planning their next moves.
Unfortunately, this could mean that researchers who are new to circRNA research might not recognize wrongly designed circRNA reagents. Researchers are also unlikely to expect incorrect reagents in seemingly sophisticated papers that are published in prestigious journals. circRNA papers that describe wrongly designed reagents, or even circRNAs that cannot be found in common databases [4, 5], could easily lead researchers to waste their time and grant money on poorly designed or fruitless experiments.
Fortunately, researchers have every reason to avoid unreliable publications, as their own time, money and even their careers are at stake. We therefore hope that our research will encourage researchers to take a very close look at the gene research papers that they intend to study [4, 6]. Building a new field, however promising this might seem, requires reliable research and publications.
Anything less could see many gene researchers simply going around in circles.
References
- Nielsen, A. F. et al. Best practice standards for circular RNA research. Nature Methods 19, 1208-1220 (2022).
- Zhang, C., et al. Hotspots and development frontiers of circRNA based on bibliometric analysis. Non-coding RNA Res. 7, 77-88 (2022).
- Wu, R. et al. Bibliometric Analysis of Global Circular RNA Research Trends from 2007 to 2018. Cell J. 23, 238-246 (2021).
- Pathmendra P. et al. Verification of nucleotide sequence reagent identities in original publications in high impact factor cancer research journals. bioRxiv, doi: https://doi.org/10.1101/2023.02.03.526922 (2023) (preprint).
- Patop, I. L. & Kadener, S. circRNAs in Cancer. Curr. Op. Genet. Dev. 48, 121-127 (2018).
- Park, Y. et al. Identification of human gene research articles with wrongly identified nucleotide sequences. Life Sci. Alliance. 5, e202101203 (2022).
This post may be cited as:
Byrne, J. (21 June 2023) The trouble with circRNAs- are researchers going around in circles? Research Ethics Monthly: https://ahrecs.com/the-trouble-with-circrnas-are-researchers-going-around-in-circles/