Antoine Dujon

French scientist living in Australia doing #stats on marine and terrestrial species to understand the links between #evolution and #cancer. Part of @CANECEV.

[Antoine Dujon @AMDujon]

New study in @royalsociety investigating how a short term sublethal exposure to a cancer risk factor affects the behaviour of a freshwater planaria and how it recovers from the damages over time.

Link to the paper:
https://t.co/wQEbyDXlXv

[Antoine Dujon @AMDujon]

@AlexanderMWolf7 @Andrei_Tarkhov Hydra oligactis shows actuarial senescence even in asexual phase we have a paper coming soon on that. It is the only hydra species we know that does ageing. Hydra vulgaris shows constant mortality rates over time (no actuarial senescence). This one doesn't seems to age.

[Antoine Dujon @AMDujon]

Don't really have the choice those days. If you want to show your paper is genuine you need at the very minimum to share the raw data and all the code used to generate the analyses. But soon this won't be enough... and I am not too sure how to demonstrate the work is genuine.

[Mikael Elias @elias_mikael]

⚠️All of the below images were FABRICATED by ChatGPT Images 2.0, each with a single prompt❗️ ⚠️

[Antoine Dujon @AMDujon]

@drdevangm Usually your grant gets rejected because you do not have a Nature paper, not because you have many! Commiserations!

[Antoine Dujon @AMDujon]

@db_econ If we build only the left side of the rail it goes all the way!

[Antoine Dujon @AMDujon]

@anshulkundaje It is part of the job, I totally agree! But the incentives are not here.

[Antoine Dujon @AMDujon]

@AndreaSottoriva @trevoragraham @gcaravagna We understand cancer mostly through metastasis, because metastasis kills humans. But many multicellular species develop benign tumours with major health impacts (hydra, planaria...). Because of anthropocentrism I think there are blind spots in how we think about cancer biology.

[Antoine Dujon @AMDujon]

@Lyonpaul @Nature @CancerGrand @nlbigas @IARCWHO For some reason your post is showed as age restricted on my side.

[Antoine Dujon @AMDujon]

@plasmidsaurus Something is clogging the volcano, and now lava is fountaining out from the sides. Probably not very safe!

[Antoine Dujon @AMDujon]

@jaredrhoads Be ready for your manuscript to be claimed by 50 predatory journals within minutes of it to be submitted to the database.

[Antoine Dujon @AMDujon]

@NeuroAI_Nexus I think this is a direct consequence of journals shifting proof-checking responsibilities onto authors. It would not be particularly difficult for them to implement a tool that automatically cross-checks every reference in a submitted manuscript.

[Antoine Dujon @AMDujon]

@AllenInstitute Perfect mapping of that single brain cell! Congrats for all the hard work! Decades in the making!

[Antoine Dujon @AMDujon]

@RolandDunbrack If one wanted to be cynical: Anything for such a paper! If an AI-written paper gets into Nature or Science, the OA fee is worth paying by the humans. What matters most is the journal brand and the "stamp of approval" for jobs and funding, not so much the science itself.

[Antoine Dujon @AMDujon]

@akoustov One can imagine AIs specifically trained to prompt-inject peer review systems and bypass checks. That’s why humans must stay in the loop. My concern: unpaid, time-poor editors may end up relying too heavily on AI outputs for decisions.

[Antoine Dujon @AMDujon]

@chris_harrod Watch the price of plastic lab consumables explode again, with a six-month waiting list for 6-well plates.

[Antoine Dujon @AMDujon]

@ziv_ravid @thegautamkamath @icmlconf If a reviewer agrees not to use AI to write their review but fails to disclose that they did, this constitutes a breach of trust. Trust is a fundamental pillar of science, and when it is compromised, the credibility and meaning of our work are weakened.

[Antoine Dujon @AMDujon]

@gabriberton It's the guy you train with of course! the reviewer should know that! Joke aside as a quantitative biologist reading the tweets about the CS/AI conferences the selection process looks horrifying.

[Antoine Dujon @AMDujon]

@dashunwang Tried it on my research topic, and it seems to hallucinate a lot. It doesn't seems to capture the body of literature I, collaborators, and colleagues wrote on the topic over the past 10-15 years.

[Antoine Dujon @AMDujon]

https://t.co/ROwdiDdVuh

[Antoine Dujon @AMDujon]

🚨 📰 Evolutionary Applications is launching a special issue on the theme of Evolutionary Medicine for which I am honoured to be an editor. 📰🚨 LINK BELOW 👇 👇👇

[Antoine Dujon @AMDujon]

Both original research and synthetic reviews are welcome. Join us in showcasing how evolutionary perspectives can transform medical sciences.