I haven't been following the Cassava Biosciences story, but this Science news report (by Charles Piller) seems pretty damning:
...AFTER ORI REACHED OUT TO CUNY, the school’s investigators examined Wang-authored papers published from 2003 through 2021, a conference poster, and a grant proposal to NIH. Many of the 31 allegations the panel reviewed involved apparently improper alterations of Western blots, a technique to distinguish distinct proteins within a tissue sample. Such manipulation can significantly alter the interpretation and validity of experimental findings. The committee said it “found evidence highly suggestive of deliberate scientific misconduct by Dr. Wang for 14 of the 31 allegations.”
To check for doctoring of those data, Shafer and his colleagues sought raw-data images to compare against published versions. Wang provided none of those, the report said, adding: “It appears likely that no primary data and no research notebooks pertaining to the 31 allegations exist.” The panel also found that Wang “starkly siloed” Western blot preparation in his lab, apparently preparing nearly all such images himself—a highly unusual practice for a lab’s principal investigator.
Among his defenses, Wang told the investigators that “at least one hard drive” containing key data was destroyed by CCNY officials when they sequestered his materials for review. Wang also accused the committee of bias against him, “failing to follow the CUNY guidelines for this investigation, and of lacking a basic understanding of Western blot analysis.” The committee noted in its report, however, that three of its members “routinely conduct experiments involving protein biochemistry and two out of four routinely conduct and publish western blot experiments.”
It seems surprising when a PI solely prepares Western blot data - that genuinely seems pretty alarming.
(After working for 10+ years in chemical manufacturing, that there isn't a basic quality assurance function in academic science is kind of amazing. If I were a PI, I could easily imagine double-signoffs required for providing key spectra and supplemental information...)
I agree with you on the boss doing a basic technique like Westerns being a red flag.
ReplyDeleteHowever, on the QA question... I think at least 19 PIs out of 20 would give you a confused look and ask you why you'd waste resources on something like that. At least that large a fraction of the academic labs I've been in or talked to prioritize efficiency--measured as publishable findings per grant dollar--in a way that would only be degraded by pretty much any serious attempt at QA.
Charlatans in academia = par for the course… unfortunately 99.9% of them get away with it.
ReplyDelete