Given the size and scope of the annual J.P. Morgan (JPM) Healthcare meeting (I resisted the temptation to say “diversity”), everyone in town – the minority who actually attend the formal presentations, and the many others who show up in San Francisco to meet and network – comes away with a slightly different experience.
With this caveat (and with the explicit reminder/disclosure that I now work at a life science venture fund, and as always, I’m speaking only for myself), I left the meeting with two fairly pronounced takeaways.
JPM: Two Contrasting Takeaways
First, this feels like an unbelievable, almost magical time in biopharma – a colleague described it (in a good way) as science fiction coming to life. Biological technologies, approaches, and ambitions that might have been dismissed as fantasies only a few years ago now are part of the fabric of the industry – and increasingly, it seems, clinical care. Gene therapy, gene editing, cell therapy, immune modulation – these modalities, alone and in combination, are what many in and around biopharma are contemplating, and the sorts of programs many drug development organizations are hoping to prosecute. It’s hardly surprising many JPM participants emerged with the sense of optimism my Forbes colleague Matthew Herper so accurately captured.
I was equally surprised by what I saw – or more accurately, didn’t see – through the lens of data and technology. As I’ve shared on Twitter, in addition to life science opportunities, I aspire to focus on the elusive middle-ground between tech and life science, and identify and invest in grounded, implementation-focused tech-powered startups that can improve how impactful new treatments are discovered, evaluated, and delivered. However, my overwhelming impression from this year’s JPM is that while data and tech may be embraced at the level of the C-suite, and while everyone professes an interest in AI, these emerging approaches and ways of thinking have generally not penetrated most biopharma organizations at the line/operations level, and have generally not yet impacted how these organizations actually approach their basic work of discovering and developing new therapeutics. Exploratory innovation initiatives, of course, abound, as do data wrangling and integration efforts (see here, eg), but these activities as yet seem to have had minimal impact on how most R&D is actually prosecuted within these organizations.
From what I can gather, it’s not a hostility to technology as much as a sense that it’s not immediately clear to most of those in the trenches how (or even whether) the emerging technologies will meaningfully impact the work they need to do, and many are concerned about, or at least wary of, the additional work it may create. Most acknowledge the possibility that big data and emerging analytics will likely be useful eventually, but few see these changes on the immediate horizon.
One biologist who does believe we’re on the threshold of profound change is Vijay Pande, a computational biologist and investor at the high-profile VC fund Andreessen-Horowitz, best known for their tech prowess, but increasingly looking towards biology; they recently raised their second biology-focused fund ($450M), and have built out their biology/healthcare team. During JPM, Pande published a (characteristically) thoughtful essay that captured what might be described as the tech view of biology.
The gist of Pande’s argument is that biology is especially complex because evolution has left cells with the equivalent of “technical debt” – essentially, old code that isn’t really used but has stuck around; cells can be “refactored,” he argues – the old code removed, the essential stuff retained and perhaps refined. Moreover, our ability to understand biology has been limited by the intellectual capabilities of the human mind; turn AI loose on biology, he argues, and we’ll be able to “go beyond the limits of the human mind and inherited tech debt.” And when “we can go beyond human hunches to really understanding biology,” Pande says, “we get far greater predictive power.” He adds,
“For the first time, the technical debt and ‘spaghetti code’ of biology can be mapped, understood, and even refactored. And given the better-than-Moore’s-Law for bio, this is happening at a time when genomics, proteomics, metabolomics, etc. have become relatively inexpensive to map. Coupled with the advances in AI (which itself are driven by similar cost reduction curves), this all opens the door to new applications of biology for healthcare with unprecedented accuracy. So the question becomes: When you finally understand the spaghetti code of bio, what can you do with it?”Pande’s perspective on our current state of understanding – the view that we’re on threshold of replacing empirical experimentation with in silico analysis — contrasts sharply with the perspective of almost everyone I met at JPM with biopharma domain expertise. Most would likely dismiss Pande’s take as incredibly naïve, disconnected from their lived reality. As outspoken entrepreneur Ethan Perlstein (an impassioned empiricist) put it, rather acidly, on Twitter, Pande’s piece “is like an AI bot imitating a first year CS grad student discovering biology for the first time on Wikipedia.” Well, then.
Tech futurists like Pande typically are not troubled by such critique; many have the view that workers in industries about to be disrupted by tech don’t see it coming, and tend to see such disruption as the sort of thing that impacts other people and other industries, often discovering too late that they were wrong. (Of course, sometimes it’s the futurists who are wrong — flying cars, anyone?)
How Will Tech Find Its Way Into Pharma?
My own view is that we’re still pretty far away from the future Pande describes, and I’ve seen little evidence to suggest technology has brought us anywhere close to meaningfully resolving the complexity of biology and of biological systems, much less turning drug discovery and development into a tidy in silico exercise.
And yet, as fantastical as this tech perspective seems, my sense is that there’s something incredibly robust and exciting behind it. Pande’s optimism, I suspect, is driven by the remarkable progress that occurring on the tech side, especially in the tools available for collecting and analyzing data, often in massive quantities. Simply stated, these tools are too powerful and too important not to impact how we think about biology and drug discovery. The question is how to bring the power of these tools to bear in developing impactful new medicines for patients. (See also this recent post on the vital, often-underappreciated importance of implementation.)
To this point, much of the interaction has taken the form of transactional engagements, what I call “Rumpelstiltskin” projects, designed essentially to spin straw into gold. For example, a pharma company has a dataset (such as from a failed trial), and the tech co is tasked with finding hidden value – a discriminatory biomarker, perhaps. Or, a pharma company is struggling with designing a small molecule against a particular target, and the tech co is tasked with using their proprietary methodology to deliver it.
In the near term, I suspect the main impact of technology on drug development will be through such transactions. A pharma is more likely to license a compound developed through fancy analytics (it knows how to evaluate compounds) than it is to license, successfully implement, and routinely incorporate the analytic methodology itself. Given that the economics for products (especially after early clinical de-risking) tend to be especially attractive, it’s perhaps not surprising that many companies that start as analytic or diagnostic companies ultimately pivot towards therapeutics.
After enough of these successes, however, pharmas will recognize that competence in modern analytics (or big data), is essential for drug development (a must-have, rather than a nice-to-have, in consultant-speak), and at that point will strive with a real sense of urgency to internalize and integrate both the technology and the expertise.
The Nature Of Medical Progress
While such tech-enabled drug discovery may seem remote today, consider how far-fetched gene editing and gene therapy seemed just a few years ago – who would have guessed they would so quickly become must-have biological technologies?
Medical advances rarely arrive as overnight successes, and powerful new approaches tend to fail before they succeed, as Anish Koka has described in this truly magnificent, must-read essay on the history of organ transplantation.
“We need the pessimists,” writes Koka, “because most attempts at progress in medicine will fail. But we also need the relentless optimists, because just maybe, one of them will break through and make the impossible possible.
The point – which Paul Kedrosky described on Twitter as “the “Triumph of the Optimists” argument, or the Enlightenment idea of progress” – is that the naive/optimistic belief that progress just around corner motivates the intense effort often required to deliver change, even though the change generally takes far longer to achieve than the initial champions imagined.
We’ve seen this in gene therapy, an effort that took more than thirty years of hard work, as biotech CEO Cyrus Harmon pointed out on Twitter. Moreover, as VC Vishal Gulati wryly observed, “at no point in those thirty years did we believe it was not imminent in the next five years.” (tweet lightly edited for clarity).
I resonate with the relentless optimism that propels medical science forward – it’s also what I love most about tech and Silicon Valley. And as out of touch as the tech view of biology seems today, I am excited about the potential of emerging technologies to radically redefine the way we approach biology and understand disease.
To be sure, it would behoove tech futurists (and particularly, I suspect, tech futurists who are investors) to have the humility to appreciate the difficulty of taming biology. But it seems equally important for contemporary drug developers to remind themselves of the need for radical improvement, of the possibility of radical change, and of the tendency for disruption, like bankruptcy, to arrive slowly at first, then all at once.
David Shaywitz is a Palo Alto-based VC. His views are his own.