We use to have a multi-talented postdoc in the lab. He was dynamic, imaginative, and very productive. He is directly responsible for my current project (or rather my "official" project), and for the projects being undertaken by the three graduate students in the lab. He has published in Nature, PNAS, and had 3 back-to-back-to-back J.B.C. articles. For the non-scientific readers, this is quite a resume. He is now at Harvard.
We have recently had cause to revisit the proteins he researched. When you publish a paper on a protein you place the protein sequence in a database for everyone to access (or reference the sequence if it is already there). This sequence represents the exact sequence of the protein you investigated.
We have just re-checked the full-length sequences of the protein he used. From the full-length protein he made several smaller fragments. There are 5 mutations spread across the fragments he made. These are bad mutations and all should have been noticed and fixed.
Three of the fragments have been sent to the lab of a collaborator. All of these must be repaired, re-made and re-sent (it will actually be quicker to start these from scratch than to repair his fragments). The whole lab will stop their work and work solely on this for the next three weeks to fix this part of the problem.
The project of one graduate student involves looking at one of the fragments and changing the first two amino acids (the fragment has 246 amino acids) in a number of combinations to monitor the effect. All of the fragments with the changes at positions one and two are from the same flawed fragment. All 20 are useless and it will also be quicker to make them again from scratch.
We use to have a multi-talented, but sloppy postdoc in the lab...
3 comments:
Now that's what I call disappointing....bugger!
Sloppy science bad, bad, bad!!!!!
If you want to read about some horrendous sloppiness, if not outright fraud, in my area of expertise, google Homme Hellinga. Look for a very recent Nature report for a succinct summary. Bad, bad, bad.
"very productive" and "very sloppy" start us down the road of suspicion. these properties occasionally add up to "data faker". sad but true.
Post a Comment