Genes like KRAS, Myc, beta-catenin, and p53 now feel within reach.
Three years after its founding, Reed Jobs' oncology venture firm Yosemite has raised a $350 million second fund and grown to nearly 25 portfolio companies — a pace that surprised even its founder. Built on the premise that philanthropy and capital could be fused to pull cancer treatments directly from university laboratories, Yosemite now operates at a moment when pharmaceutical patent cliffs, record industry cash reserves, and the maturation of AI in drug discovery have aligned in ways that make previously undruggable genes feel, for the first time, within reach. The firm's quiet urgency is not merely commercial — it is a wager that the window to rewrite cancer's story may be narrower, and more open, than anyone expected.
- Pharma's largest patent cliff in history is forcing major companies to acquire aggressively, and Yosemite's early-stage bets are suddenly sitting in the path of that capital.
- AI has shattered a decades-old ceiling: where scientists could once target only 15 percent of the human genome, cryptic protein pockets and machine-learning models are now unlocking oncogenes like KRAS, p53, and Myc that resisted every prior approach.
- Two of Yosemite's 25 companies have failed for scientific reasons — a reminder that the firm operates so early that the ground itself can disappear, and that speed and risk are inseparable here.
- Synthetic control arms powered by AI could cut patient recruitment in clinical trials by half, compressing timelines that once stretched across decades into something closer to years.
- The GLP-1 boom has unexpectedly redirected scientific attention toward cancer and neurodegeneration, widening the aperture of what investors and researchers now believe is fundable and solvable.
- Jobs describes the moment as both scarier and more empowering than he anticipated — a rare admission that the stakes of the experiment have grown faster than the experiment itself.
Three years ago, Reed Jobs launched Yosemite as an experiment: fuse venture capital with no-strings-attached philanthropy, and build cancer drugs directly from university research rather than waiting for pharma to discover them. The firm has since grown to nearly 25 companies across two funds, with a second close bringing in $350 million. Jobs says Yosemite is moving faster than he expected — and the reasons are structural, technological, and partly just luck.
The structural tailwind is a pharmaceutical industry entering what Jobs calls its largest patent cliff in history. Blockbuster drugs are losing protection in a compressed window, while major companies sit on record cash reserves accumulated during the pandemic. The result is an acquisitive market: Eli Lilly bought Kelonia for $7 billion, and Revolution Medicines has doubled survival rates for the most common form of pancreatic cancer — from 12 to 24 months — by targeting KRAS, a gene long considered nearly impossible to drug.
The technological tailwind is AI. When Yosemite launched, artificial intelligence was still peripheral to drug discovery. Now it sits at the center of the firm's work — not necessarily producing better drugs, Jobs says, but accelerating the foundational science with reproducible outcomes and opening doors that were previously locked. For decades, researchers could target only about 15 percent of the human genome. That constraint is dissolving. KRAS, which had no natural binding site for a drug molecule, was cracked when Amgen scientists found a hidden pocket in its surface; AI has since mapped every targetable variant.
Yosemite's portfolio reflects this shift. The firm is pursuing p53 — the most frequently suppressed gene across human cancers — through three separate companies and multiple strategies. It backs Tune Therapeutics, which uses epigenetic editing to target hepatitis B, a disease affecting over 250 million people and the primary driver of liver cancer. It funds Histosonics, whose histotripsy device destroys tumors noninvasively by collapsing microscopic air pockets inside tissue. About a third of the new fund goes into companies Yosemite builds itself, often alongside researchers at Yale, Berkeley, and Stanford.
One surprise has been the GLP-1 boom. Early evidence suggests these drugs — which made Eli Lilly the first trillion-dollar pharmaceutical company — may offer protection against neurodegeneration and cancer unrelated to weight loss, because obesity is one of only two pan-disease risk factors, alongside smoking, that elevates risk across nearly every disease category. That has redirected capital and attention toward oncogenes that had gone cold for years.
Two of Yosemite's companies have failed, both for scientific reasons — an expected cost of operating this early. What Jobs did not expect was the speed of everything else. With a team of 17, the firm is taking meetings from anyone with an idea that could affect cancer patients, across every modality from small molecules to gene therapy to digital health. The biggest surprise, he says, is not just that the science is advancing — it's that the moment feels more consequential than he realized, which is both frightening and clarifying.
Three years ago, Reed Jobs launched Yosemite into a biotech industry still nursing its pandemic wounds. The venture firm was an experiment: what if you combined old-fashioned capital with no-strings-attached philanthropy, and built cancer drugs directly from university labs instead of waiting for pharma to discover them? Now Yosemite has grown to nearly 25 companies across two funds, with a second close that brought in $350 million. Jobs, who is easy to like and quick to deflect questions about his surname with a joke, says the firm is moving faster than he anticipated—and the reason is partly structural, partly technological, and partly just luck.
The structural piece is straightforward: pharma is entering what Jobs calls its largest patent cliff in history. A cluster of blockbuster drugs are losing protection in roughly the same window, and the major pharmaceutical companies are sitting on record cash reserves from the pandemic. That combination has made them acquisitive in ways they weren't three years ago. Eli Lilly bought Kelonia for $7 billion. Revolution Medicines, targeting KRAS—a gene long considered nearly impossible to drug—has doubled survival rates for the most common form of pancreatic cancer, from 12 to 24 months. These wins are recent, and they're changing what investors think is possible.
The technological piece is AI. When Jobs started Yosemite, artificial intelligence was still a curiosity in drug discovery. Now it's central to what the firm does. AI isn't necessarily making better drugs, Jobs says, but it's accelerating the grunt work with reproducible outcomes. More importantly, it's opening doors that were locked. For decades, scientists could only drug about 15 percent of the human genome, because they couldn't target proteins interacting with other proteins—the chemistry was too hard. That's changed. KRAS, which for years had no natural pocket for a drug molecule to latch onto, was finally cracked a decade ago when Amgen scientists found a cryptic pocket hidden in its smooth surface. AI has since found all the other variants of KRAS that can now be targeted, and shown creative new ways to block them.
Yosemite's portfolio reflects this shift. The firm is pursuing p53, the most frequently suppressed gene across human cancers, with three different companies and several strategies. It's backing Tune Therapeutics, which uses epigenetic editing—adding or removing chemical tags that act like dimmer switches on genes—to target hepatitis B, which affects over 250 million people and is the primary driver of liver cancer. It's funding Histosonics, a device company that uses histotripsy, a noninvasive therapy that creates small air pockets and collapses them to destroy tissue in specific areas, for pancreatic and liver tumors. About a third of the new fund goes into companies Yosemite builds itself, often alongside academics at Yale, Berkeley, and Stanford. The rest goes into companies other people made that Yosemite wants to join.
One surprise, Jobs says, is how the GLP-1 boom—the best-selling drug class in the world, which made Eli Lilly the first trillion-dollar pharmaceutical company—has redirected attention and capital toward disease areas that had gone cold. Early signs suggest GLP-1s may be protective against neurodegenerative disease and cancer, unrelated to weight loss. That's because obesity is one of only two pan-disease risk factors, along with smoking, that raises your risk across nearly every disease category. The result is that genes like KRAS, Myc, beta-catenin, and p53—the pantheon of oncogenes that have evaded researchers for decades—now feel within reach.
Two of Yosemite's 25 companies haven't worked out, both for scientific reasons. The firm tranches investments against scientific milestones, and since it operates so early, sometimes the science fails. That's expected. What's less expected is the speed. Jobs says he didn't anticipate Yosemite moving this fast. The firm has a team of 17 now. It's taking meetings from anyone with an idea that could affect cancer patients. It's looking at every modality—small molecules, radiopharmaceuticals, gene therapy, immunotherapy, AI, digital health. The biggest surprise, Jobs says, isn't just that the science is advancing. It's that the moment feels more important than he realized, which is both scarier and more empowering.
Notable Quotes
I didn't expect Yosemite to be moving this fast. This time is more important than I realized, which is both scarier and more empowering.— Reed Jobs
We think we found something to hit that exposed marker across all the different ways p53 gets mutated.— Reed Jobs, on Yosemite's p53 targeting strategy
The Hearth Conversation Another angle on the story
When you say AI is accelerating drug discovery, what does that actually look like in practice? Is it writing the drugs, or is it doing something else?
It's mostly doing the grunt work incredibly fast. Historically, chemists would have to manually explore thousands of possible compounds to find one that might work. AI can do that in days. But the real breakthrough is finding pockets we never knew existed. KRAS is the perfect example—it looked like a smooth oval, impossible to drug. Then someone found one tiny cryptic pocket, and that led to the first drug. AI found all the other variants we can now hit.
So it's not replacing the human insight, it's multiplying it.
Exactly. The human insight—finding that first pocket—that's still the hard part. AI just says, okay, now that we know what we're looking for, here are all the other places we can look. It's democratizing science. A smaller lab can now do what used to require massive resources.
You mentioned p53 as the biggest target. Why is that gene so important?
It's the most frequently suppressed gene across human cancers. Almost every cancer has to knock it out to exist. If we could turn it back on, or attack its mutated forms, that's one of cancer's Achilles' heels. We're going after it with three different companies and several strategies. It's never been done before.
And you think it's doable now because of AI?
Partly AI, partly just the convergence of everything—better tools, better funding, better understanding of the protein structures. But yes, AI is a big part of it. We think we've found something to hit that exposed marker across all the different ways p53 gets mutated.
What's the biggest mistake you see founders make when they're pitching you?
Bad storytelling. I've seen companies with great science fail because the CEO couldn't tell the story. Usually the founder is the scientist, and the CEO is a professional operator whose job includes raising capital and telling the story. That division of labor works well. But if you get it wrong, it doesn't matter how good the science is.