How to Build a Biotech | Hacker News Reader

How to Build a Biotech

celinehh.com

133 pointsapsec11224 comments

Hi! I'm Celine - a little embarrassed this is trending because it is very much still a draft - please forgive any typos!

I hope this is helpful and always happy to chat biotech startups - just email me - celinehh.com/about

Thanks for this great draft resource, Celine! I wonder if you see an opportunity for a "sell shovels" approach to BioTech -- for example, visualizations for drug creation, simulation and collaboration?

how is the biotech scene going ?

Celine Halioua is an incredibly talented founder. We share investors and have spent a good amount of time together at investor working sessions, and she's building something really cool. This set of resources is quite valuable and provides good insight on how tech and biotech are different.

Some of the reasons they're different are real:

In tech you're pretty sure you can build something, you're just not sure if anyone will buy it. In biotech you're pretty sure people will buy it, just not sure you can build it.

The challenges are far more complex (biology is a bitch) and require a LOT more domain specific knowledge to get traction.

It's necessarily highly regulated.

But things like people not giving younger founders a chance, and the lowballing on valuation don't need to be how things work. Hopefully people like Celine help turn that around.

> In tech you're pretty sure you can build something, you're just not sure if anyone will buy it. In biotech you're pretty sure people will buy it, just not sure you can build it.

Beautifully stated. There are countless brilliant founders in biotech who toiled for 5+ years to only find out their drug which had a credible mechanism of action failed phase 2 clinicals and now isn't worth a damn. In software, founders of this caliber would have since pivoted and more than likely created a product of some value.

In my opinion, the most important driver to unlock a revolution in biotech is enabling technologies that can create a rapid feedback loop from disease model -> drug hypothesis -> verification of safety & efficacy in humans. Right now, this is a 10 year process. Getting this process down to 1 year and <$50M would be a revolution worthy of several Nobel Prizes.

> There are countless brilliant founders in biotech who toiled for 5+ years to only find out their drug which had a credible mechanism of action failed phase 2 clinicals and now isn't worth a damn. In software, founders of this caliber would have since pivoted and more than likely created a product of some value.

That sounds like it's specific to pharma-biotech. In non-pharma biotech there seems to be more freedom to pivot (e.g. this story[0] from a few days ago).

[0]: https://news.ycombinator.com/item?id=23749296

Decelphalized monoclonal humans.

Turn off neural tube development genes so they never develop brains. Grow them in test tubes, then hook them up to life support.

They can never be conscious beings, so you can do anything with them. They're essentially large-scale tissue cultures.

You can harvest them for blood and organs, create knockout lines for different studies, and speed drug development and verification tenfold. Population studies on everything imaginable becomes tractable.

We have to get out of the biological stone age and develop real automation primitives. We don't have a real test kit. Decephalized monoclonal humans is the reverse salient we need.

I hope someone does this. If nobody does it and I wind up becoming a billionaire, I'll do it. It needs to be done.

From a bio perspective, it sounds like a great idea.

From a squimish perspective... it sounds like something from a horrendous distopia. I just have visions of rows of human shaped tissue cultures.

A less horrendous version could be cultured cells for each organ, with some kind of circulation between them which one could test theraputics for cytotoxicity. I've seen "labs on a chip"[0] for things like PCR, I wonder if you could do that for drug testing in vitro.

[0]: https://en.m.wikipedia.org/wiki/Lab-on-a-chip

You know how after the last big civilization all-out-war it turned out that the losers had secretly engaged in gross scientific experiments on humans? I think it would be naive to imagine that the winners didn't do so as well, they just didn't have their file cabinets put on display to retroactively justify the extreme violence that the world had just witnessed. Governments do scary things in basements, and they keep those basements as secret as possible -- we only get to see behind the veil when a government falls and an opponent successfully takes their files/scientists and that opponent has an incentive to make those files public. Not having brains attached would be a huge improvement over the current state of the art, in my opinion; I believe that we're probably already living in a world far more dystopian than the imagined one being discussed.

While it's a draft and very cursory, resources like this are sorely needed and I hope encourages others to post more. Unlike tech, there's few place where you can learn how to build a biotech. It's a black art that most people still learn by doing.

Decades ago, tech was like this and then Paul Graham, Venture Hacks, Fred Wilson, etc. started a movement to democratize startups and VC. It would be wonderful if the same thing happened with biotech.

Not sure what else is out there, but the TMCx accelerator program is specifically designed around biotech startups and was created to build up the (bio)tech startup ecosystem around the texas medical center (largest medical center worldwide)

https://www.tmc.edu/innovation/innovation-programs/tmcx/

There are a number of other incubators that are serving some of that role. The community is growing, but it's still fairly small.

In the biotech space part of that incubation/culture can come in the form of shared lab space. Here are a few (in the Bay Area):

- MBC (née QB3) https://mbcbiolabs.com/

- JLabs https://jlabs.jnjinnovation.com/

- StartX https://startx.com/med

- IndieBio https://indiebio.co/

- Bonneville Labs https://bonnevillelabs.com/

There are a number of other governments and entities that are trying to replicate and improve on the model.

Portland has OTRADI https://www.otradi.org/

(help add to the list?)

cheers, good to know.

to note, TMCx is also a JLabs site, fwiw

https://jlabs.jnjinnovation.com/locations/jlabs-tmc

It looks like there's still a lot in a draft form here, but the kind (and honestly, the breadth) of information here is difficult to come by outside of experience. And it is really valuable to have some of those bullet points articulated in one place.

Small community around here too - there's a lot we can learn from each other.

Completely agree, and feeling most of these pain points in the medical device / diagnostics industry as well. There are so many black boxes and unknown-unknowns in this space, especially when it comes to developing indications and navigating regulatory interactions. That's a main driver of why we built Essenvia (https://www.essenvia.com/#/), which automates the creation and submission of all regulatory docs required for medical device approval. We're expanding next into other aspects of the industry (complaint handling, compliance testing) as well as adjacent industries (environmental, pharma, etc.).

I too see there is still a lot in draft form, but it states that "this is currently a draft and will launch formally Jan 2019." So its over a year and a half old and lots left to fill in. Maybe they should update that line to address that issue? Its a good start for resources.

I was going to make a "I'd like one Biotech, please" joke, but this is actually an interesting website! All the images are a bit confusing, but the information is great.

Edit: yeaaaah, enable JavaScript if you have NoScript, the website looks and works much better.

Does biotech only have to do with drug development? What about other areas like water cleaning etc, wouldn't that be considered biotech as well?

One thing I have noticed and felt is that there is a very steep energy well that slides you into drug development if your technology is anywhere even near - as drug development has financial returns that are an order of magnitude (or two or three) greater than almost any other problem that bio-technology can be utilized for (right now).

I predict that as the value of those other projects rise, and as the capital required to produce a drug decreases, that those other projects will really start to blossom. And I look forward to it.

And medical devices. I've spent my whole career working on them.

They would, but the purchasers of such things are really different. For a drug biotech, you are trying to get to a point in trials where you are acquired by a large pharmaceutical maker, and the large pharmaceutical maker then produces it and uses their distribution network to get it into circulation.

Something like water cleaning goes through really different routes. You may be selling direct to a consumer segment, or to a contractor segment that chooses the product as part of a design for a consumer, or to a municipality or utility. Those are all really different from each other, and really different from drug development. They also tend to involve a very different skillset, much more chemical and mechanical engineering.

If your definition of "biotech" (an old word that seems less used nowadays) includes fields like bioinformatics and biomedical engineering, then there is a wide range of different technologies besides drugs.

Wow! There's a lot of amazing stuff here. I think a lot of it can be abstracted to apply to any field really.

Having founded and run businesses in software, hardware, and small molecule pharma, I have to say that the life sciences are a completely different universe and that the differences can trip you up. Sure, there are commonalities, just as both baseball basketball are played with a spherical ball.

So a resource like this can be quite valuable. But if it’s good it will have less to say about, say, SAAS startup than you might think.

Reply on news.ycombinator.com