Smut Clyde found a couple of helpless and defenceless journals to ravish and ransack. They belong to a very dodgy Chinese publisher, the editors are mostly Chinese, but not only. The content seems to be exclusively provided by papermills, or rather a papermill. Sometimes the sewage spills into other, more respectable journals, those by Taylor & Francis, Elsevier, Wiley etc.
To run away is not an option for Smut Clyde, so he faces the beast head-on.
With Random Precision
By Smut Clyde
Diverted by so many other exciting developments in fake research, I tried to get away from chronicling and taxonomising papermills… but it is like trying to quit show business. Just when I thought I was out, they drag me back in. Papermills continue to appear on the scene (or rather, they have been there all along, waiting to be noticed) and once again we find ourselves with shovels and brooms, cleaning up after the performing elephants.
The Master of the String-of-Sausages
“I am open to the possibility that they both outsourced their Western Blot production to a single, independent Wurst-Meister specialist.” – Dr Smut Clyde, art historian of the Chinese Papermill Renaissance.
The source of today’s elephant doody first showed up in 2017-2018 and is still an active generator of counterfactuals (spreadsheet here). Production seemed to peak in 2020-2021, thanks to Annals of Translational Medicine and Translational Cancer Research (both from AME Publishing). So no special detection skills were required to find manifestations of this mill, only an aimless wander through those journals’ Tables of Contents – though I hasten to add that ATM and TCR were far from the only outlets (43 and 11 papers respectively), while both are actively targeted by other papermills. For the sake of completeness I’ll mention Journal of Thoracic Disease, also from AME. It’s only known to have published one of the papermill’s products but a systematic search might find more.
Fig 3 from “Effects of wogonoside on invasion and migration of lung cancer A549 cells and angiogenesis in xenograft tumors of nude mice” (Yan et al 2020).
Fig 3b from “MicroRNA-377-3p targeting MMP-16 inhibits ovarian cancer cell growth, invasion, and interstitial transition” (Wang et al 2021).
Fig 1d from “Induction of lncRNA NORAD accounts for hypoxia-induced chemoresistance and vasculogenic mimicry in colorectal cancer by sponging the miR-495-3p/ hypoxia-inducible factor-1α (HIF-1α)” (Zhang et al 2020).
Fig 3A from “Vitexin suppresses the proliferation, angiogenesis and stemness of endometrial cancer through the PI3K/AKT pathway” (Liang et al 2023).
But wait; AME also publishes Journal of Gastrointestinal Oncology. It’s affected too.
[right] Fig 1E from “Enhancing the chemotherapy effect of Apatinib on gastric cancer by co-treating with salidroside to reprogram the tumor hypoxia micro-environment and induce cell apoptosis” (Zhang et al 2020).
At some point the editors of ATM must have realised that the ballooning page counts were not entirely due to a surge in their reputation, for they throttled down the flow. Always exert caution when throttling the flow of papers! I posit that they decided against transforming from a generalist predatory journal into a specialist papermill pipeline.
This belated realisation of the formulaic production-line quality of the submissions did not translate into retractions of the garbage they’d published already; the only retraction was forced by the request of an unwilling author.
I can’t blame the author for not wanting to be involved in this farrago of fabulation.
[right] Fig 2A from “Artesunate induces apoptosis and inhibits the proliferation, stemness, and tumorigenesis of leukemia” (Chen et al 2020b)
For comparison:
Back to the retracted paper. Intuition tells me that there is a technician in a laboratory somewhere whose day-job involves excising xenograft tumors from their living nude-mouse incubators and lining them up in a fashion-show cat-walk to be photographed.
[right] Fig 5A from Chen et al 2020b.
[below] Fig 5A from “Crocin induces autophagic cell death and inhibits cell invasion of cervical cancer SiHa cells through activation of PI3K/AKT” (Zhang et al 2020).
The technician harbours higher ambitions, and feels that his or her photographic skills deserve wider recognition, so there are multiple versions of each cat-walk where the body-parts pose with different rulers and other props.
[right] Fig 4D from “MIER3 suppresses the progression of non-small cell lung cancer by inhibiting Wnt/β-Catenin pathway and histone acetyltransferase activity” (Zhang et al 2020).
Ranging up to at least five versions in one egregious case. Two appeared within a single paper. “That looks fine”, thought the Editors and peer-reviewers of Clinical and Experimental Pharmacology and Physiology.
[right] Fig 5a from “Synergistic anticancer effects of everolimus (RAD001) and Rhein on gastric cancer cells via phosphoinositide-3-kinase (PI3K)/protein kinase B (AKT)/ mammalian target of rapamycin (mTOR) pathway” (Gao et al 2022).
One more pair:
[right] Fig 6B from “Propofol modulates the proliferation, invasion and migration of bladder cancer cells through the miR‑145‑5p/TOP2A axis” (Du et al 2021).
I was relieved to learn that not all the immortalised cancer cells – variously treated by exposure to non-coding RNAs, or repurposed drugs, or obscure fungal metabolites like ligustrazine* – are proliferating within the bodies of mutilated animals. Some proliferate within culture medium to become “spheroids”. Spheroids do not always approximate a sphere; in many cases they are closer in shape to the West Bromley Fighting Haddock.
Fig 6D from “Vitexin suppresses the proliferation, angiogenesis and stemness of endometrial cancer through the PI3K/AKT pathway” (Liang et al 2023).
[right] Fig 3A from “Effects of knockout of long-chain non-coding RNA LSINCT5 on proliferation, apoptosis, epithelial-mesenchymal transition, and p38MAPK pathway of pancreatic cancer PANC-1 cells (Dai et al 2020).
For the sake of variety, in other cases the cells proliferate into clones or colonies on a plate of gel, in which case it is a Clonigenic assay, named after the colons and other punctuation marks formed by all the little dots. We saw some of those above but there are many more. If they put you in mind of chocolate-chip cookies then it is probably time for a snack.
Fig 3H,I from “The Down-Regulation of TrkB Alleviates the Malignant Biological Behavior and Cancer Stem-Like Property of Laryngeal Cancer” (Hu et al 2020).
Fig 2A from Li et al (2020).
Fig 1C from “Tanshinone I attenuates the malignant biological properties of ovarian cancer by inducing apoptosis and autophagy via the inactivation of PI3K/AKT/mTOR pathway” (Zhou et al 2020).
Fig 4B from Chen et al (2020a).
Fig 1B from “The anti-tumor effect of taxifolin on lung cancer via suppressing stemness and epithelial-mesenchymal transition in vitro and oncogenesis in nude mice” (Wang et al 2020).
Fig 4C from Wang et al.
Fig 2A from “Vitexin attenuates epithelial ovarian cancer cell viability and motility in vitro and carcinogenesis in vivo via p38 and ERK1/2 pathways related VEGFA” (Zhou et al 2020).
Fig 1C from “Wogonoside Attenuates Cutaneous Squamous Cell Carcinoma by Reducing Epithelial– Mesenchymal Transition/Invasion and Cancer Stem-Like Cell Property” (Wang et al 2020).
Fig 1C from Zhou et al (2020).
Fig 1A from “Lung cancer A549 cells suppressed with overexpressed HNF1B or PCDHA13 inhibited PI3K/AKT phosphorylation” (Kang et al 2020).
Sometimes the cells display their contortionist skills by forcing themselves through the pores of Transwell filters in an Invasion assay, lured by the promise of nutrient-rich culture medium on the other side… only to be fixed and stained. Biologists revel in playing this kind of practical joke upon unsuspecting cells.
Fig 2F from “Long noncoding RNA LINC00460 conduces to tumor growth and metastasis of hepatocellular carcinoma through miR-342-3p-dependent AGR2 up-regulation” (Hong et al 2020).
Fig 4A from “MiR-511 mimic transfection inhibits the proliferation, invasion of osteosarcoma cells and reduces metastatic osteosarcoma tumor burden in nude mice via targeting MAPK1” (Wu et al 2019).
Fig 3A from Chen et al (2020a).
Fig 3A from Chen et al (2020a) again.
Fig 4G from “LINC01094 promotes the invasion of ovarian cancer cells and regulates the Wnt/β‑catenin signaling pathway by targeting miR‑532‑3p” (Chen et al 2021).
Cells are induced to glow a fluorescent green. When I become Benevolent Dictator of my country, I will order the Post Office to issue a set of fluorescent postage stamps in these designs.
Fig 3C from “Crocin induces autophagic cell death and inhibits cell invasion of cervical cancer SiHa cells through activation of PI3K/AKT” (Zhang et al 2020).
Fig 4A from “Puerarin 6″-O-xyloside suppressed HCC via regulating proliferation, stemness, and apoptosis with inhibited PI3K/AKT/mTOR” (Li et al 2020)..
Fig 4F from “MicroRNA-153-3p sensitizes melanoma cells to dacarbazine by suppressing ATG5-mediated autophagy and apoptosis” (Hou et al 2020).
Who had “duplicate use of immunohistochemistry-stained tissue slices” in their Bingo card?
[right] Fig 6H from.”Nanocomplexes loaded with miR-128-3p for enhancing chemotherapy effect of colorectal cancer through dual-targeting silence the activity of PI3K/AKT and MEK/ERK pathway” (Liu et al 2020).
[right] Fig 6C from “Co-delivery of plantamajoside and sorafenib by a multi-functional nanoparticle to combat the drug resistance of hepatocellular carcinoma through reprograming the tumor hypoxic microenvironment” (Zan et al 2019).
These repeating elements are proof of shenanigans but they are not immediately helpful for classification or differential diagnosis, as repetitions of similar form feature in the oeuvres of many other papermills. Classification comes first and without it, we would be no better than mere beasts such as the West Bromley Fighting Haddock. Yet friends and family are dismissive and unsympathetic when I complain about the difficulties of papermill taxonomy.
The present millers are more sympathetic. They helpfully invented a form of flow-cytometry apoptosis plot and often use it to advertise their identity.
[right] Fig 2C from Li et al (2020).
As an example I focus on Fig 2a of “Liquiritigenin exerts the anti-cancer role in oral cancer via inducing autophagy-related apoptosis through PI3K/AKT/mTOR pathway inhibition in vitro and in vivo” (Ji et al 2021). A separate population of dots has been airbrush-tooled neatly into each quadrant, with trails of points between each quadrant and its neighbours 
to suggest the circulation of cells between the healthy state (lower left quadrants) and apoptosis. It is the Cycle of Life! The appeal of ‘Liquiritigenin’ – other than the millers’ choice of International Klein Blue for plotting Fig 2a – is its Fig 5a, an excuse to show more xenograft fashion parades. Or they could just be wads of used chewing-gum.
[right] Fig 7B from “Long noncoding RNA CASC9 promotes the proliferation and metastasis of papillary thyroid cancer via sponging miR-488-3p” (Chen et al 2020).
Alternative title: This Gum’s for Hire
The other diagnostic attribute of this papermill is a style of generated or hand-drawn Western-Blot-like images. They are meant to be generic fungible commodities, free from distinguishing features that could become the target of critique, but the nature of their non-distinctiveness is itself characteristic. Fuzzy soft-focus streaks march down across a grey untextured background like the alien ships in a game of Space Invaders.
Fig 2c from Ji et al (2021).
Fig 3C from “Bupivacaine inhibits the malignant biological behavior of oral squamous cell carcinoma cells by inhibiting the activation of ERK1/2 and STAT3” (Wang et al 2021).
With immunohistochemistry slices and tumours and spheroids and Transwell panels, recycling is inevitable when the number of the millers’ customers outstrips the image archives that they managed to acquire from actual laboratories. In contrast, these artistic western blot creations do not draw on actual experiments as a source, and you might think that the millers can grind out unique examples of the gamut without ever duplicating a snowflake. But there you underestimate their laziness! René Aquariusin PubPeer comments showed some of the replications in a helpful diagram.
As noted, the papermillers made the most of the ATM / TCR window of opportunity; while it lasted they pumped these blurry streaks into the journals like foie-gras manufacturers force-feeding geese. They spread themselves around beyond the AME stable.
- Eight of their fabrications appeared in Oncotargets & Therapy from Dove (now part of Taylor & Francis)
- Six in Pharmaceutical Biology from Taylor & Francis.
- 15 in the six journals that comprise the Spandidos stable.
- Seven in Journal of Oncology, once a Hindawi journal, now Wiley.
- As the window closed, the papermill scraped further down below the bottom of the barrel in search of sub-barrels. We find 10 papers in Heliyon, a Cloaca Maxima created by Elsevier to house material that couldn’t even reach the minimal standards of longer-established Elsevier spigots.
- Readers should be familiar with Bioengineered, a journal that currently lacks an Editor-in-Chief after Taylor & Francisstaged an intervention earlier this year. Under the old regime, that journal became home for nine of our papermill’s inventions.
- It’s not always the pay-to-publish model at fault! The Journal of Biochemical and Molecular Toxicology (from Wiley) is a subscription journal.
The credibility and prestige and impact of the recipient journals were not major concerns. Though they receive all the imprimatur of being indexed and mirrored by PubMed. As do ATM and TCR.
Dead horses won’t flog themselves
“This is electrophoresis porn, readers, a phrase that I never expected to find myself writing.” -Smut Clyde
These skeezy fraud-friendly journals do have their uses. Once a papermill finds them, the fakes are concentrated in one place, undiluted by genuine research results, allowing a clear picture to crystallise and delineate its papermill style.
Four or so products even grace the pages of RSC Advances. These were retracted as part of the Colludin’ Massacre of January 2021, when the editorial staff of Advances struck back against papermills who had abused their trusting nature. I wrote “four or so” because the Advances papers are part of a a larger tranche of ambiguous transition. Here are some soft-focus streaks:
Fig 4F from “Protective effect of FOXP3-mediated miR-146b-5p/Robo1/NF-κB system on lipopolysaccharide-induced acute lung injury in mice” (Zhu & Chen 2020).
Fig 4E from “Salvianolic acid B inhibits inflammatory response and cell apoptosis via the PI3K/Akt signaling pathway in IL-1β-induced osteoarthritis chondrocytes” (Zhu et al 2018).
Within the right-hand paper (Zhu et al 2018), Fig 4E coexists with Figs 2A, 3B,D, which do have a background texture (albeit a repeating one) while the streaks, while fuzzy, are more rectangular and resemble processionary caterpillars.
To cut a long story even longer, half the Western-Blot-like images of “Salvianolic acid B…” (Zhu et al 2018) are recognisably from the present papermill, and half belong to a now-deprecated style associated with the “Travelling Tumor” papermill, which fell off the radar in a flurry of retractions for which I cheerfully claim responsibility. It’s the same papermill. Just because they wore out their welcome and were caught, millers do not abandon the lucrative income-stream of fraudulence for sale, to pursue new careers in more honest forms of employment such as piracy or organ harvesting. That would be like cancelling a popular series like ‘Dr Who’ just because the lead actor’s contract expired. They simply steal or buy fresh image archives, or create new image-creation algorithms, and regenerate themselves. We taught them how to do fraud better. Well done us! Making the world a better place!
The Adventures of a Mouse Malignancy Group Portraitist
Smut Clyde was busy with yet another Chinese paper mill. A plastic ruler was deployed.
Below at right, an example of the canonical tumours that gave the older papermill its name. In contrast, at left, Fig 6A from Wu et al (2019) displays novel body-parts… but they’re arranged on the same bench-protecting fabric sheet! The lab technician was busy. Meanwhile the western blots in Wu et al are of the new form.
[right] Fig 8A from “lncRNA CCAT1 contributes to the growth and invasion of gastric cancer via targeting miR-219-1” (Li et al 2019) [retracted].
A suite of overlapping IHC-stained slides unites three papers from the reinvented papermill with one retracted paper, Shi et al (2018), that otherwise displays the Western Blot hallmarks of its earlier Travelling-Tumour incarnation.
Fig 4C from “MiR-320d suppresses the progression of breast cancer lncRNA HNF1A-AS1 regulation and SOX4 inhibition” (Shi et al 2018).
Fig 5B from “Icariin Mitigates the Growth and Invasion Ability of Human Oral Squamous Cell Carcinoma via Inhibiting Toll-Like Receptor 4 and Phosphorylation of NF-κB P65” (Lei et al 2020).
Fig 6C from “Astragaloside IV (AS-IV) alleviates the malignant biological behavior of hepatocellular carcinoma via Wnt/b-catenin signaling pathway” (Jiang & Mao 2019).
We encountered Zan et al (2019) already. Here is Fig 6C again, annotated to show more overlaps:
[right] Fig 5D from “Salidroside inhibits the proliferation and migration of gastric carcinoma cells and tumor growth the activation of ERS-dependent autophagy and apoptosis” (Yan et al 2019).
“MiR-320d suppresses…” (Shi et al 2018) and “Salidroside inhibits…” (Yan et al 2019) both belong to that RSC Advances transitional tranche. They are linked:
The Flow Cytometry apoptosis plots from “Salidroside inhibits…” are not of the recent, artisanal, hand-stippled form. They come from an earlier suite that was stolen and used extensively by the Travelling Tumour papermill.
To avoid disappointment, I hasten to add that the scatterplots in this stolen suite do not always resemble Thidwick the Big-Hearted Moose. Some are more like West Bromley Fighting Haddocks goldfish.
Below at left is a November 2020 Correction for “MIER3 suppresses the progression of non-small cell lung cancer by inhibiting Wnt/β-Catenin pathway and histone acetyltransferase activity” (Zhang et al 2020): “The Figure 3B should be replaced and corrected as the below one.” Pay no attention to the overlap of two Transwell panels for different cell lines, or the duplication with another paper (right).
This cack-handed correction was an attempt to scrub off some of the stigmata of the Travelling Tumour period. Here is the original Fig 3B.
The purple-stained invasive cells are very familiar indeed.
Fig 2D from “Down-regulated LncR-MALAT1 suppressed cell proliferation and migration by inactivating autophagy in bladder cancer” (Qi et al 2018) [retracted].
Fig 2C from “Apatinib inhibits macrophage-mediated epithelial–mesenchymal transition in lung cancer” (Liu et al 2018).
None of these papers attracted many citations (only occasionally literature reviews), or many readers (not even the authors or readers of those literature reviews). But there are so many of them! 241 papers in the spreadsheet so far (including 85 from the Travelling Tumour phase that started the millers on their road to fame and fortune), with ??? to go. Think how long this post could have been.
* Found in nattō, fermented cocoa beans and sourdough, so it must be good for you.
“We are extremely guilty and distressed”
“Now no-one wants ForBetterScience to become an all-Papermill channel. And we cannot really expect to shame or inspire scriveners in the academic-ghostwriter industry to seek out more constructive applications for their talents, so the point of exposing them is not immediately obvious.” -Smut Clyde
CODA
Another spreadsheet, another papermill. Or maybe not. I suspect that the production-line featured there is a reinvented incarnation of the prolific ‘Contractor’ papermill, with alternative, plausible generative sources of Western Blots and flow-cytometry scatterplots. If it is the Contractor mill in a new disguise, it’s probably polluted the literature with a thousand papers already because of that “prolific” part, and the 145 entries in the spreadsheet are just the tip of the fatberg.
That probably deserves a separate blogpost, though. This is just a teaser.
Donate to Smut Clyde!
If you liked Smut Clyde’s work, you can leave here a small tip of 10 NZD (USD 7). Or several of small tips, just increase the amount as you like (2x=NZD 20; 5x=NZD 50). Your donation will go straight to Smut Clyde’s beer fund.
NZ$10.00
For Better Science 
Pingback: Science Integrity Digest, August 2024 – Science Integrity Digest