Flaw cytometry, FACsimiles and fly-speck thunderclouds

This post was earlier cross-posted at Leonid Schneider's site, hence the unfrivolous tone. The version there is improved by Leonid's editing, background details and frame-story. OTOH, it does not have any references to the role of fly-specks in literature... in particular, the opening chapters of Osudy dobrého vojáka Švejka za světové války when Palivec (hapless landlord of The Chalice) is arrested on suspicion of treason because flies had shat upon the pub's mandatory portrait of Emperor Franz-Joseph.

The post is based on Elizabeth Bik's examination of some papers by Academician Xuetao Cao, as a follow-up / elaboration on Leonid's own summary of Bik's concerns. The story was picked up by other media and became a cause célèbre, with the Chinese academic establishment promising to fully exonerate investigate the adumbrated Xuetao Cao. Also, Paris, Rome, Munich and New York are burning.

When a paper has five diagrams to illustrate the various techniques that were used (and to reassure the readers that experiments did actually take place), is it the same paper when four of those Figures are replaced with less problematic versions? This is a new phrasing of the conundrum of continuity and identity over change, taking its place in the menagerie of Metaphysics alongside the Ship of Theseus, Trigger's Broom and Locke's Sock. And evidently the paper's Conclusions at least are unchanged by the substitution of data, at least in the view of the editors of Clinical Cancer Research. The resulting spectacular megacorrection (Wang et al. 2005 [3]) featured in a recent post at Science Integrity Digest.

[Figure 3B is no longer operative]

[Figure 3D]

One of these now-deprecated Figures consisted of Flow Cytometry plots, which have come up before. Here, measurements of surface protein expression in a succession of cells become log-scale coordinates of a FACS plot. There are lasers involved, and fluorescent antibody markers, but all that concerns us now is that a raw data file of sequential cell measurements can be run through the plotting software repeatedly, while twiddling the settings of the gating parameters that define which cells to plot and which to omit (as outliers or irrelevant). Thus multiple slightly-different FACSimiles can be produced, and if the researchers are sufficiently sloppy in how they file the printouts then these might turn up in a paper mistakenly presented as different experiments.

Alternatively, one might fake a FACSimile in 'Visual Enhancement Software' by cut-&-pasting groups of points, though the results can leave the impression that a child ran amok with a potato stamp. If one is caught out it is hard to blame sloppy filing, and it is time to find a junior author or graduate student to through to the wolves. Or under the bus. Or to the wolves living under the bus.

Now FACSimiles and "flaw cytometry' feature heavily in the oeuvre of Dr Cao across the 16-year span 2003-2019, at least in the fraction of that oeuvre that has so far provoked discussion in threads at PubPeer. In many cases the use of a duplicate or almost-duplicate plot was an innocent mistake, quickly corrected (Qian et al, 2013 [12]) or subsequently acknowledged (Liu et al., 2018 [15]).



Also Wang et al. (2014) [14], Fig 5 from Liu et al. (2018) [16], and probably more by now, for this is a developing story.

In other cases, concerns about copy-pasted images were conveyed to journal editors five years ago but progressed no further (Wang et al. 2004 [2]; Li et al. 2007 [5]). Perhaps the Journal of Biological Chemistry at the time had not tasked any staff to find gambling going on here ensure image integrity; alternatively, the staff-member in question saw all the glitches and many-colored boxes in the emailed notifications and thought that they were invitations to an art exhibition... perhaps a Frank Stella collaboration with Nam June Paik.

[Thunderclouds with color-coded rainfall]

I should emphasise at this point that these are all multi-author papers, and it is above my non-existent pay-grade to single out any one person as the originator of the fabrications. Dr Cao was the principal author of many but not all of them, but he is clearly a man of many roles and responsibilities competing for his time and attention. In addition, email addresses for Dr Cao were only recently uploaded to the PubPeer database to ensure that he received invitations to take part in the discussions. Now that he is aware of the concerns being raised and has responded in some cases, and given his role as China's Integrity Watchdog, there is every reason to believe that he or the corresponding authors will correct or retract other papers.

Much of the time, alas, any error was unlikely to be innocent, given the levels of shenanigans and jiggery-pokery. Non artifact, sed artifice. In Wang et al. (2014) [13], Figures 2G and 2H are rife with refracted, 90°-rotated constellations, like star-gazing through a kaleidoscope:

Figure 4C (below at right) is a Transwell migration assay of some kind, but it has kaleidoscope problems of its own.



Moving along to Han et al. (2009) [10], Figure 4 is a patchwork:

One of the squares from the patchwork, having appeared three times in 4A, makes a fourth reappearance in Figure 7B of Li et al. (2009) [11].



Now [11] is a kettle of red herrings of another color. It is hard to know where to start. Reused FACS confections form a delicate web that binds all the illustrations together. The rubber-stamping is subtle (except when it's not)...



...so an easy way to demonstrate how patches of random half-tone are replicated between Figures is to overlay them, with one image black/white reversed, to see if and where patterns of identical dots cancel out.



Two more examples because why not? Figure 6C from Yang et al (2004) [1]...

...and Figure 1B from Wang et al. (2006) [4].

Here is one that was Withdrawn (not retracted) in 2015: Wang et al (2008) [9].

VOLUME 283 (2008) PAGES 12076–12084

This article has been withdrawn by the authors.

The Withdrawal notice is terse and cryptic, but it sounds as if someone on the authorship list belatedly paid more attention than the editors and reviewers had, and decided not to enrich the scientific literature with legacies like these:

Yes, I know... this is a case where a Western Blot was caught in the matter-transporter malfunction, rather than a FACSimile, for this diagrammatic corpus is versatile and not exclusively flow-cytometric. Here are some more Western Blot examples. Figure 2 from [5], which we just encountered:

And returning to the beginning of our time-line, [1], Figure 2:

At left, Fig 4A/B from He et al (2007) [6], with a procession of Loch Ness monsters. At right, tying together Sun et al. (2008a) [7] and Sun et al. (2008b) [8].



Of course readers are really here for the flyspeck thunderclouds. Fortunately those last two examples provide a segue back to the main theme. So here are are Dr Bik's findings in Figures 3B and 6 of [7], before she ran out of colors:

Figure 2E of [8] has a little less going on.

Now the lower-left frame of 2E is not one of the "twenty-seven 8 x 10 colored glossy photographs with circles and arrows and a paragraph on the back of each one explainin' what each one was", but all the same it is scaring me, because what are those TWO BLACK-RINGED EYES creeping down from above into the cut-off frame?

Perhaps it is a hommage to a 1991 Rik Mayall movie.

Even more than usual, I should credit Dr Elisabeth Bik and other less-nonymous contributors to PubPeer for providing all the images I am curating here. Interested readers should visit the source material to see how much I left out!

SOURCES:

[1]. "Cyclin L2, a novel RNA polymerase II-associated cyclin, is involved in pre-mRNA splicing and induces apoptosis of human hepatocellular carcinoma cells", Lianjun Yang, Nan Li, Chunmei Wang, Yizhi Yu, Liang Yuan, Minghui Zhang, Xuetao Cao (2004)
Journal of Biological Chemistry [PubPeer].

[2]."A novel human phosphatidylethanolamine-binding protein resists tumor necrosis factor alpha-induced apoptosis by inhibiting mitogen-activated protein kinase pathway activation and phosphatidylethanolamine externalization", Xiaojian Wang, Nan Li, Bin Liu, Hongying Sun, Taoyong Chen, Hongzhe Li, Jianming Qiu, Lihuang Zhang, Tao Wan, Xuetao Cao (2004).
Journal of Biological Chemistry doi: 10.1074/jbc.m405147200 [PubPeer].

[3]. "Silencing of human phosphatidylethanolamine-binding protein 4 sensitizes breast cancer cells to tumor necrosis factor-alpha-induced apoptosis and cell growth arrest", Xiaojian Wang, Nan Li, Hongzhe Li, Bin Liu, Jianming Qiu, Taoyong Chen, Xuetao Cao (2005).
Clinical Cancer Research doi: 10.1158/1078-0432.ccr-05-0879 [PubPeer].

[4]. "Induction of allospecific tolerance by immature dendritic cells genetically modified to express soluble TNF receptor", Quanxing Wang, Yushan Liu, Jianli Wang, Guoshan Ding, Weiping Zhang, Guoyou Chen, Minghui Zhang, Shusen Zheng, Xuetao Cao (2006).

Journal of Immunology [PubPeer].

[5]. "hPEBP4 resists TRAIL-induced apoptosis of human prostate cancer cells by activating Akt and deactivating ERK1/2 pathways, Hongzhe Li, Xiaojian Wang, Nan Li, Jianming Qiu, Yuanyuan Zhang, Xuetao Cao (2007).
Journal of Biological Chemistry doi: 10.1074/jbc.m609494200 [PubPeer].

[6] "TLR4 signaling promotes immune escape of human lung cancer cells by inducing immunosuppressive cytokines and apoptosis resistance", Weigang He, Qiuyan Liu, Li Wang, Wei Chen, Nan Li, Xuetao Cao (2007).
Molecular Immunology doi: 10.1016/j.molimm.2007.01.022 [PubPeer]

[7]. "Rapamycin suppresses TLR4-triggered IL-6 and PGE(2) production of colon cancer cells by inhibiting TLR4 expression and NF-kappaB activation", Qiaoling Sun, Qiuyang Liu, Yuanyuan Zheng, Xuetao Cao (2008a).
Molecular Immunology doi: 10.1016/j.molimm.2008.01.025 [PubPeer].

[8]. "Rapamycin reverses TLR4 signaling-triggered tumor apoptosis resistance by disrupting Akt-mediated Bcl-xL upregulation", Qiaoling Sun, Yuanyuan Zheng, Qiuyan Liu, Xuetao Cao (2008).

International Immunopharmacology doi: 10.1016/j.intimp.2008.08.009 [PubPeer].

[9]. "IPP5, a novel protein inhibitor of protein phosphatase 1, promotes G1/S progression in a Thr-40-dependent manner", Xiaojian Wang, Bin Liu, Nan Li, Hongzhe Li, Jianming Qiu, Yuanyuan Zhang, Xuetao Cao (2008).
Journal of Biological Chemistry doi: 10.1074/jbc.m801571200 [PubPeer].

[10]. "CD69+ CD4+ CD25- T cells, a new subset of regulatory T cells, suppress T cell proliferation through membrane-bound TGF-beta 1", Yanmei Han, Qiuli Guo, Minggang Zhang, Zhubo Chen, Xuetao Cao (2009).
Journal of Immunology doi: 10.4049/jimmunol.182.1.111 [PubPeer].

[11]. "Cancer-expanded myeloid-derived suppressor cells induce anergy of NK cells through membrane-bound TGF-beta 1", Hequan Li, Yanmei Han, Qiuli Guo, Minggang Zhang, Xuetao Cao (2009).
Journal of Immunology doi: 10.4049/jimmunol.182.1.240 [PubPeer].

[12]. "Fas signal promotes the immunosuppressive function of regulatory dendritic cells via the ERK/β-catenin pathway", Cheng Qian, Li Qian, Yizhi Yu, Huazhang An, Zhenhong Guo, Yanmei Han, Yongjian Chen, Yi Bai, Qingqing Wang, Xuetao Cao (2013).
Journal of Biological Chemistry doi: 10.1074/jbc.m112.425751 [PubPeer].

[13]. "Intracellular NAMPT–NAD+–SIRT1 cascade improves post-ischaemic vascular repair by modulating Notch signalling in endothelial progenitors", Pei Wang, Hui Du, Can-Can Zhou, Jie Song, Xingguang Liu, Xuetao Cao, Jawahar L. Mehta, Yi Shi, Ding-Feng Su, Chao-Yu Miao (2014).
Cardiovascular Research 10.1093/cvr/cvu220 [PubPeer].

[14]. "The STAT3-binding long noncoding RNA lnc-DC controls human dendritic cell differentiation", Pin Wang, Yiquan Xue, Yanmei Han, L. Lin, Cong Wu, Sheng Xu, Zhengping Jiang, Junfang Xu, Qiuyan Liu, Xuetao Cao (2014).
Science doi: 10.1126/science.1251456 [PubPeer].

[15] "HSP70L1-mediated intracellular priming of dendritic cell vaccination induces more potent CTL response against cancer", Shuxun Liu, Lin Yi, Ma Ling, Jinxia Jiang, Lijun Song, Juan Liu, Xuetao Cao (2018).
Cellular and Molecular Immunology doi: 10.1038/cmi.2016.33 [PubPeer]

[16]. "MicroRNA in vivo precipitation identifies miR-151-3p as a computational unpredictable miRNA to target Stat3 and inhibits innate IL-6 production", Xiang Liu, Xiaoping Su, Sheng Xu, Huamin Wang, Dan Han, Jiangxue Li, Mingyan Huang, Xuetao Cao (2018).
Cellular and Molecular Immunology [PubPeer].