Is it possible that these proxies you mention could weed out compounds incorrectly? For example, if we have a liver proxy that fails on drug A, is it possible that drug A will work fine in the human body provided that there’s a lot of other processes that could contribute to the liver?

Absolutely. That's the flip side of in vitro and animal testing. You might see a negative signal and stop development, only to find out that it can't be replicated in humans.

The cell line assays replicate much of what happens in humans, but not all of it, in particular the interactions between different tissues. For example, a drug might be toxic in a liver assay because a reactive metabolite damages the cells. But in an actual human, you might never see those high of concentrations in the liver or there may be a process (like glutathione adducts) that deactivates the metabolite and prevents cell damage.

right. i want to go beyond cell assays, i want high throughput organ slices that preserve important morphologies.

it’s not a perfect model, but i think it’s a better proxy than a mouse.