Sunday, May 5, 2013

Summary of inappropriate image manipulation by Issei Komuro's group.

Complement C1q activates canonical Wnt signaling and promotes aging-related phenotypes.

Nature. 2012 May 10;485(7397):251-5.
Mitochondrial DNA that escapes from autophagy causes inflammation and heart failure.

Patients with CD36 deficiency are associated with enhanced atherosclerotic cardiovascular diseases.

A crucial role for adipose tissue p53 in the regulation of insulin resistance.

Nature. 2008 Jul 17;454(7202):345-9.
IGFBP-4 is an inhibitor of canonical Wnt signalling required for cardiogenesis.

J Biol Chem. 2004 Mar 19;279(12):11384-91.
Adult cardiac Sca-1-positive cells differentiate into beating cardiomyocytes.

Biochem Biophys Res Commun. 2004 Sep 10;322(1):310-9.
Atrial natriuretic peptide inhibits cardiomyocyte hypertrophy through mitogen-activated protein kinase phosphatase-1.

Nat Cell Biol. 2004 Jun;6(6):499-506.
Mechanical stress activates angiotensin II type 1 receptor without the involvement of angiotensin II.

Circulation. 2003 Dec 16;108(24):3024-30.
Heat shock transcription factor 1 protects cardiomyocytes from ischemia/reperfusion injury.

Circulation. 2002 Sep 10;106(11):1397-402.
Detection of experimental autoimmune myocarditis in rats by 111In monoclonal antibody specific for tenascin-C.

Circulation. 2002 Mar 12;105(10):1240-6.
Peroxisome proliferator-activated receptor gamma plays a critical role in inhibition of cardiac hypertrophy in vitro and in vivo.

J Biol Chem. 1999 Mar 19;274(12):8231-9.
Context-dependent transcriptional cooperation mediated by cardiac transcription factors Csx/Nkx-2.5 and GATA-4.


Coverage of the media on this matter
Forbes: Suspicions Raised About Another Japanese Cardiovascular Researcher (copy

Saturday, May 4, 2013

Circulation. 2002 Sep 10;106(11):1397-402.

Circulation. 2002 Sep 10;106(11):1397-402.

Detection of experimental autoimmune myocarditis in rats by 111In monoclonal antibody specific for tenascin-C.

Sato M, Toyozaki T, Odaka K, Uehara T, Arano Y, Hasegawa H, Yoshida K, Imanaka-Yoshida K, Yoshida T, Hiroe M, Tadokoro H, Irie T, Tanada S,Komuro I.

Source

Department of Cardiovascular Science and Medicine, Graduate School of Medicine, Chiba University, Chiba-shi, Chiba, Japan.



In figure 3, the image of 3-c is the same image of 165 degrees-clockwise rotated and size-reduced version of 2-c.



国立大学法人千葉大学学術国際部研究推進課長様 

千葉大学大学院医学研究院循環病態医科学グループの研究活動の不正行為の疑惑に関して申立てましたXXXXXXX と申します(平成25年5月30日付けで受理)。 

前回送付した申立書において、千葉大学に所属していた小室一成氏らのグループから発表された14報の論文について研究不正の疑義を申立てていましたが、さらに同グループから発表された下記の別の論文1報に関して実験画像の流用(捏造・改ざん)の疑いが見つかりましたのでお知らせいたします。この論文の疑義に関しても調査していただけないでしょうか。論文名および不正疑惑の具体的内容は下記の通りです。このメールに添付した不正疑義の解説図(Figure3.jpg)もご参照ください。 

論文名: 
Circulation. 2002 Sep 10;106(11):1397-402. 
Detection of experimental autoimmune myocarditis in rats by 111In monoclonal antibody specific for tenascin-C. 
Sato M, Toyozaki T, Odaka K, Uehara T, Arano Y, Hasegawa H, Yoshida K, Imanaka-Yoshida K, Yoshida T, Hiroe M, Tadokoro H, Irie T, Tanada S,Komuro I. 
Source 
Department of Cardiovascular Science and Medicine, Graduate School of Medicine, Chiba University, Chiba-shi, Chiba, Japan. 

不正疑惑の具体的内容: 
Figure 3において、2-cの画像を時計回りに165度回転させ、サイズを90%に縮小させたものに、3-cの画像が類似しています。 
2-cと3-cの画像は互いに異なる個体から由来するものであるはずですが、互いに類似しているのは極めて不自然です。 
これが、不適切な画像流用だとすると、元の画像を回転させたり、サイズを変更したりするなどの操作を加えていることから、意図的に研究不正を働いた可能性が推測されます。2-cの画像は2-aと2-bに輪郭がほとんど一致していることから、2-cの画像がオリジナル画像と推測され、その場合、3-cが流用画像となりますが、つじつまを合わせるために、3-cの輪郭も3-aや3-bの画像の輪郭に近くなるように、画像の回転角度やサイズが調整され流用されています。よって、意図的な不正である可能性が極めて高いと考えれます。 
不正の有無について調査をよろしくお願いいたします。 









ブルースin東京/Blues in Tokyo・Yumemi より



  
  
「あ・・。
  
 Wow! It sure is jaw-dropping. 
  
 先に解析した画像トリックについて、もうちょっと突っ込んでみましょう。
  
 ひとつの画像が別の画像に捏り変えることができるということは、その逆も成り立ちます。
  
 すなわち、3-c2-cなら、2-c3-cも成り立つということです。
 

 
  
 二枚しかなかったら、どっちがどっちかわかります?わからないでしょ。
  
  
 「こんなことして何になる」と言われるかもしれませんが、捏られた画像の特徴をもっとよく観察してみましょう。
  
 「アングリ開いたアンコウの口」に見えてきませんか?実写、動物、かわいい のデコメ絵文字
 
  
  
  
  

  

 もともと口がないはずの3-a 3-bの左端に、それらしく削って捏っているのがお キャラなど のデコメ絵文字 かりでしょうか?

  
  
【補足的考察】
 原図2-cから3-cを捏り、その後、口を合わせるために、3-b3-aの左端を巧妙に削った。。。そなの?
【語彙】
 捏る」はもともと「こねる」の意味。近頃は「ツクる」「ネツる」と訓読みするのがウケる。
 
 

  

訪問者:「ちっ!開いた口が。。。」

  
 さらに悪い冗談がその後のペーパーに・つ・づ・く。。。  かわいい のデコメ絵文字
  
   
訪問者:「もう知らんぞ。どーなっても。かわいい のデコメ絵文字
  
  
 Wow! It sure is jaw-dropping. 
  
 Delve a little bit further into the trick previously analyzed. When an image is cooked up into another image, the opposite is true. That is, image processing from 3-c to 2-c is possible, and vice-versa from 2-c to 3-c.
  
 If there were only two images, could you guess which is which? You can’t, I bet. Though you may say what is the point of delving it, but let’s see a faked image a little bit closer. It begins to seem like an O-sign mouth of an anglerfish, isn’t it? 実写、動物、かわいい のデコメ絵文字
 Did you notice that each left edge of images (3-a, 3-b) without mouths were trimmed to look more real?
    
  
 Supplementary Consideration: The original image 2-c was possibly cooked up to 3-c, and then for making mouths, each left edge of 3-b and 3-a was cut off in an artful way. Is that so?
  
 Glossary: Japanese word ‘捏るwith an original meaning of ‘Kone-ru’ (v. pug in English) in kun reading, however, is more popularly pronounced ‘Tuku-ru’ or ‘Netsu-ru’ in these days.
  

VISITOR

 No shit! It sure is jaw-dropping.

  
 To be continued for more funny jokes in later papers. かわいい のデコメ絵文字
     
VISITOR
 I don’t care anymore what happens to you!かわいい のデコメ絵文字

Cell. 2012 Jun 8;149(6):1298-313.


Cell. 2012 Jun 8;149(6):1298-313. doi: 10.1016/j.cell.2012.03.047.

Complement C1q activates canonical Wnt signaling and promotes aging-related phenotypes.

Naito AT, Sumida T, Nomura S, Liu ML, Higo T, Nakagawa A, Okada K, Sakai T, Hashimoto A, Hara Y, Shimizu I, Zhu W, Toko H, Katada A,Akazawa H, Oka T, Lee JK, Minamino T, Nagai T, Walsh K, Kikuchi A, Matsumoto M, Botto M, Shiojima I, Komuro I.

Source

Department of Cardiovascular Medicine, Osaka University Graduate School of Medicine, Japan.

Erratum in http://www.cell.com/retrieve/pii/S009286741200877X

The article above reported a role for complement C1q in promoting age-related phenotypes. Since publication, we have noticed that the upper panels shown inFigure 7F were incorrectly selected and labeled. The upper-right panel portrayed an image from aged C1qKO mice, not aged wild-type mice treated with M241, as originally labeled. The upper-left panel was correctly labeled as aged wild-type mice treated with PBS, but the magnification of both images in the top panels was incorrectly labeled and did not match the magnification of the bottom two panels. The corrected figure, with the upper-left panel now portraying an image from aged wild-type mice treated with PBS and the upper-right panel portraying an image from aged wild-type mice treated with M241, both taken with a 4× magnification (scale bars, 150 μm), is shown on the following page. We sincerely apologize for these mistakes and emphasize that the correction of these errors does not impact our conclusions



Nature. 2012 May 10;485(7397):251-5.


Nature. 2012 May 10;485(7397):251-5. doi: 10.1038/nature10992.

Mitochondrial DNA that escapes from autophagy causes inflammationand heart failure.

Oka T, Hikoso S, Yamaguchi O, Taneike M, Takeda T, Tamai T, Oyabu J, Murakawa T, Nakayama H,Nishida K, Akira S, Yamamoto A, Komuro I, Otsu K.

Source

Department of Cardiovascular Medicine, Osaka University Graduate School of Medicine, Suita, Osaka 565-0871, Japan.

In Fig. 4d, the panel showing CD3 staining for ODN2088 control-treated TAC-operated Dnase2a+/+ mice is a duplicate of that showing Ly6G staining for ODN2088-treated TAC-operated Dnase2a−/− mice. 
































大阪大学大学院医学系研究科 循環器内科学 Nature誌論文 Fig.4d 最下段(CD3)の一番左の画像が、上から3段目(Ly6G)の一番右の画像と同一。 ケアレスミスか?念のため、他にも調査の必要性があるか?

 (その後、Corrigendum が出される。指摘の件以外にも、多数の不適切な画像操作が判明)
↓ 

Erratum in


Nature. 2012 Oct 11;490(7419):292.


In this Letter, several images were mistakenly switched or duplicated during preparation of the artwork. In Figs 1f and 2a, the sham-operated Dnase2a−/− and TAC-operated Dnase2a+/+ mice panels were switched. In Fig. 4d, the panel showing CD3 staining for ODN2088 control-treated TAC-operated Dnase2a+/+ mice (now shown correctly as black-bordered panel in Fig. 1 below) is a duplicate of that showing Ly6G staining for ODN2088-treated TAC-operated Dnase2a−/− mice. The panel showing CD45 staining for ODN2088-treated TAC-operated Dnase2a+/+ (now shown correctly as blue-bordered panel in Fig. 1 below) was prepared from the original picture of ODN2088 control-treated TAC-operated Dnase2a+/+. In Supplementary Fig. 4c, sham-operatedDnase2a−/− and TAC-operated Dnase2a+/+ mice panels were switched. Finally, in Supplementary Fig. 10d, the panels showing CD3 and Ly6G staining for sham-operated Tlr9+/+ mice were switched. These corrections do not alter any of the conclusions of this Letter, and the authors apologize for any confusion these errors may have caused.

J Atheroscler Thromb. 2012;19(3):263-75.


J Atheroscler Thromb. 2012;19(3):263-75. Epub 2011 Nov 10.

Patients with CD36 deficiency are associated with enhanced atherosclerotic cardiovascular diseases.

Yuasa-Kawase M, Masuda D, Yamashita T, Kawase R, Nakaoka H, Inagaki M, Nakatani K, Tsubakio-Yamamoto K,Ohama T, Matsuyama A, Nishida M, Ishigami M, Kawamoto T, Komuro I, Yamashita S.

Source

Department of Cardiovascular Medicine, Osaka University Graduate School of Medicine, Osaka, Japan.


Fig.2  (A is the same as C)





















今話題の阪大循環器内科グループの論文:J Atheroscler Thromb. 2012;19(3):263-75. Fig.2A (RCA) と Fig.2C (RCIA) の写真が同一。

Nat Med. 2009 Sep;15(9):1082-7.


Nat Med. 2009 Sep;15(9):1082-7. doi: 10.1038/nm.2014. Epub 2009 Aug 30.

A crucial role for adipose tissue p53 in the regulation of insulin resistance.

Minamino T, Orimo M, Shimizu I, Kunieda T, Yokoyama M, Ito T, Nojima A, Nabetani A, Oike Y, Matsubara H, Ishikawa F,Komuro I.

Source

Department of Cardiovascular Science and Medicine, Chiba University Graduate School of Medicine, Chuo-ku, Chiba, Japan.


Surprisingly, In Jan, 2014, Nature Publishing Group replaced the following misused image data in Supplemental Fig. 2c with new one without any notice in the Nature Medicine article (2009 Sep;15(9):)!
Please compare the following old images and new images.
(New image: http://www.nature.com/nm/journal/v15/n9/suppinfo/nm.2014_S1.html).


"WT (HF/HS)" and "G4 (HF/HS)" in Supplemental Fig. 2c and "Pre operation" in Supplemental Fig. 2d are similar CT scan images. These images may be derived from the identical patient.





Supplemental Fig. 2cの一番右、Lean tissue/BWの棒グラフのG4 (HS/HS)G4 Trp53+/- (HS/HS)は、meanSEMが完全に一致。遺伝型が違うのにn=5meanSEMが完全一致するのは極めて稀?CT scanのこれだけ似た画像を元に作ればあり得る? 



Nature. 2008 Jul 17;454(7202):345-9.


Nature. 2008 Jul 17;454(7202):345-9. doi: 10.1038/nature07027. Epub 2008 Jun 4.

IGFBP-4 is an inhibitor of canonical Wnt signalling required for cardiogenesis.

Zhu W, Shiojima I, Ito Y, Li Z, Ikeda H, Yoshida M, Naito AT, Nishi J, Ueno H, Umezawa A, Minamino T, Nagai T,Kikuchi A, Asashima M, Komuro I.

Source

Department of Cardiovascular Science and Medicine, Chiba University Graduate School of Medicine, Chiba 260-8670, Japan.

(The above PDF file has higher resolution images compared with the html file。)




Concern#1
In the beta-actin panel of Fig.1C, a left band (OP-9) is similar to a right band (OP-9 + anti-BP4).








Concern#2
In the beta-Actin panel of Fig.3B, lane 1 and lane 2 bands from the left are similar.









Concern#3
In Fig.3e, lane 3 and 5 bands from the left of a "alpha-MHC" panel are inappropriately erased, and Lane 1, 4 and 6 of the "alpha-MHC" panel and lane 5 of a "GATA-4" panel are similar to each other.





Concern#4
In the "GATA-4" panel of Fig.3e, lane 1 and lane 6 bands from the left are similar to each other"










Concern#5
In Fig.3e, Lane 1 and 2 image from the left of a GATA-4 panel are a vertically flipped and resized version of lane 1 and 2 image from the left of Nkx2.5 panel.




Concern#6
In Supplementary Figure S3e, the left band of "Nkx2.5 (-)" panel is the same as the right band of "GATA4 BP4-2" panel.






Concern#7
In Supplementary Figure S3f, the left band of beta-actin panel is a resized version of that of Nkx2.5 panel.