It seems that at some level, that is exactly the purpose of (publicly-funded) basic research on the academic side. For example, the idea of messenger RNA therapy has been around for decades in the academic literature, and of course this idea was only possible with preceding advancements in general biology.

Protein folding helps some. Genomics help some. Computer models help with pharmacokinetics.

But all these things help you get to the expensive part more easily: testing safety and efficacy.

https://www.science.org/content/blog-post/big-problems

https://www.science.org/content/blog-post/ai-and-drug-discov...

https://www.science.org/content/blog-post/another-ai-generat...

Read up on Andrew Lo. He used his background in finance to develop better biotech financing models.

https://www.bloomberg.com/news/articles/2021-02-05/out-of-gr...

Currently, no. Bringing a new drug to market takes several years and a billion dollars. And I haven't seen any compelling evidence that anyone is working on a "solution".

A large part of the cost and complexity is the clinical trials, of which several stages of increasing complexity are required to get FDA approval. While it might seem inefficient, it's also the reason why vaccines work so well and have such small side effects - about half of Phase 3 candidates either fail the clinical trial or are rejected by the government for some other reason.

Each round of clinical trials requires significantly more money than the one before it, but it's also arguably easier to get investors on board in the later rounds as risk is lower. Smaller pharma companies often aim to get bought outright before their product is brought to market because it makes more sense for everyone involved.

Maybe someone will come up with an algorithm that lets us simulate clinical trials with sufficient precision, but that seems like a far-off sci fi fantasy to me.