The Scandals Rocking Cancer Science

In case you wondered why cancer is so hard to cure.

Interesting insights from Bloomberg reporting about The Scandals Rocking Cancer Science and how “there have been accusations of hyped findings, sloppiness and even fraud.” - "...hype is the norm. Researchers found they could only reproduce 50 out of 193 experiments. And in those that did replicate, the second try showed much smaller effect sizes — only 15% as big as what had originally been claimed."

So, let's do the math. 15% of the 50 out of 193 experiments is not more than 3.9%. Meaning, that only 3.9% of the experiments were reproducible AND their authors did not inflate the results. Then, considering that 90% of drugs tested on animals fail in humans—that is, only 10% of the experiments made on animals correlated in humans—we are left with less than 0.4%. This means that only about 1 paper over 250 that are published in research journals leads to anything useful for cancer treatment. All the rest is either hyped or low quality which doesn't lead to any tangible results.

And the naive hope that AI will fix everything is part of the same hype and delusion. I expect people to say that we must double down and need more of the same.

While, for the root causes of this, see my take on why ‘Big Science’ is a big flop.