Ask HN: Getting started in biology with a software background • Hacker News Board

Re: Ask HN: Getting started in biology with a software background

#71

Post by goatlover » Thu, Oct 31, 2019, 4:50 PM UTC

notus wrote: post #24

codingdave wrote: post #21

Earlier quoted context omitted.

> you are going to need a PhD if you want to truly understand all of the background of the field Is that the need, though? Few SaaS CEOs truly understand all the background of computer science. They do understand the problem space, but they leave it to the experts in the organization to get into the details of applying computer science to the problem. Is biology different, where the level of expertise needs to be at…

There are plenty of executives running biotech companies who have barely any biology expertise. Pretty much everything they know is contained to their product and its competitors. There are PHD's and MD's in SV who basically specialize in working as consultants for these types of products and back up their potential. You see it all the time with health startups when they have a medical advisor or something to that ef…

Doesn't give me much confidence in their product.

Re: Ask HN: Getting started in biology with a software background

#72

Post by dekhn » Thu, Oct 31, 2019, 4:53 PM UTC

iongoatb wrote: post #31

I'm curious who all these sarcastic bio "experts" are that are suggesting getting a PhD or hiring one. I'm a former bio major and researcher/scientist that transitioned to software engineering years ago. I've published papers in bio and worked with many PhDs. Many of them were idiots. Being a PhD doesn't mean anything, it's the independent work that you put in yourself (in an academic lab or by yourself) that determi…

Saying that many PhDs you worked with are idiots says more about you, than it does about the PhDs.

PhD process is only one path, but it has a number of useful attributes, such as being very close to the active state of the art research, feedback from experts in the field, handholding through the paper and grant process, and introduction to a large social network. Those are all very hard to do with home labs and biohacking. Things like journal clubs with other grad students often help people learn how to evaluate the literature with the appropriate context. Independent work is important, but teamwork and learning from others is far more important.

I've worked with some very smart people (famous software engineers with long track records of innovation) that wanted to help with bioinformatics, and they did do some cool things, but their lack of deep context (the sort of thing you can get from a PhD program or working in the field for many years) ultimately led to problems such as premature optimization for the wrong distribution of data.

Nonetheless, I have see independents who came to the field with no background, absorbed the ground knowledge, and made major contributions, but that's absurdly rare compared to PhDs.

Re: Ask HN: Getting started in biology with a software background

#73

Post by esel2k » Thu, Oct 31, 2019, 4:54 PM UTC

I lead projects with software and bioinformaticians in one of the worlds top bioinfo/genetics company. If you want my opinion: Salary sucks in such companies as there are way too many phds out there. Big egos and never built a company... And bios have no clue about good engineering, TDD or CI. It is all copy paste scripts.

So don’t compete with bio phds, the money is not worth it. I have studied molec bio and comp science separately and can work in both industries. We talk about 150k vs 100k salary differences. Find an area that interests you like data analytics software in healthcare or something like this and go to a tech company you will thank me.

Re: Ask HN: Getting started in biology with a software background

#74

Post by mindcrime » Thu, Oct 31, 2019, 4:55 PM UTC

Odenwaelder wrote: post #36

iongoatb wrote: post #31

I'm curious who all these sarcastic bio "experts" are that are suggesting getting a PhD or hiring one. I'm a former bio major and researcher/scientist that transitioned to software engineering years ago. I've published papers in bio and worked with many PhDs. Many of them were idiots. Being a PhD doesn't mean anything, it's the independent work that you put in yourself (in an academic lab or by yourself) that determi…

1) Not enough for starting a biotech startup as OP would like to do. Having a PhD means that you've spent years on a certain subject. You cannot just read a book and be up to speed on biomedical research and CRISPR. 2) Having hands-on experience in wet labs is useful and relatively easy to learn. People can learn wet lab skills sufficient to carry out experiments (i.e. pipetting stuff together) in under a year. This…

You cannot just read a book and be up to speed on biomedical research and CRISPR.

FWIW, the OP didn't specifically say ze wanted to do "biomedical". Biotech is bigger than just medical applications. Biotech could skew more towards materials science, or environmental engineering, or any number of areas besides "treatment for diseases in humans" or whatever.

Re: Ask HN: Getting started in biology with a software background

#75

Post by mindcrime » Thu, Oct 31, 2019, 4:59 PM UTC

dekhn wrote: post #72

iongoatb wrote: post #31

I'm curious who all these sarcastic bio "experts" are that are suggesting getting a PhD or hiring one. I'm a former bio major and researcher/scientist that transitioned to software engineering years ago. I've published papers in bio and worked with many PhDs. Many of them were idiots. Being a PhD doesn't mean anything, it's the independent work that you put in yourself (in an academic lab or by yourself) that determi…

Saying that many PhDs you worked with are idiots says more about you, than it does about the PhDs. PhD process is only one path, but it has a number of useful attributes, such as being very close to the active state of the art research, feedback from experts in the field, handholding through the paper and grant process, and introduction to a large social network. Those are all very hard to do with home labs and bioha…

Saying that most PhDs you worked with are idiots

The person you are replying to did not say that.

Re: Ask HN: Getting started in biology with a software background

#76

Post by sansnomme » Thu, Oct 31, 2019, 5:02 PM UTC

unearthed wrote: post #67

cryoshon wrote: post #38

my background is in biotech and i've worked at a small handful of biotech startups in a scientific capacity. here's my advice: unless you have friends who can set you up with the right VCs and ensure that they will be willing to overlook your lack of experience and IP, don't bother. you're not going to get up to speed working in the lab on your own in any short amount of time. learning the theoretical stuff that you…

Why are Phds so cheap to hire if their skills are so valueable?

There is not enough demand (and relatively large amount of supply). Biotech isn't like software where it adds immediate value. There are hoops to jump through (unless you are only doing it to cure a disease affecting yourself). The liability is high if things fail. People have accepted bugs as part of computing, they are not so generous when it comes to civil engineering, and even less so when it comes to the life sciences. And most importantly, too many wannabe researchers and not enough money to go around.

To use an analogy, game developers. Plenty of supply because of everyone wants to be game developers. The work is difficult, pay is bad, work-life balance non-existent. Not to mention the economics are brutal. Are individual game developers valuable? Certainly. Are they cheap too? Yes. The problem is structural (big AAA firms as gatekeepers on the high end, tremendous amount of competition on the indie end).

The problem with Bio/Med/Pharma is also structural. Med school supply is capped by forces like professional associations (fancier terms for doctor unions/lobbies), hospital supply is also capped by the similar forces, with perhaps some contributions by the pharma industry. As for pharma, I think enough people has complained about it that it's not worth elaborating here. There has never been a shortage in bio talent or consumer demand. The bottleneck is due regulatory and policy reasons.

HN treats computing and software like how certain groups treats guns. This is the exception not the norm. Most other industries have tremendous oversight/interference by authorities that the move fast break things method is difficult to apply. In practically any other field, the barrier of entry is artificially high and information is locked-in, not open and shared.

Imagine if you create a new Kubernetes load balancer and to get it deployed you need to have it "approved and certified". You pay perhaps 4 figures to the Cloud Native Computing Foundation who is "accredited" by the Association for Computing Machinery to certify software. Sounds ridiculous? This is the way of life for practically everything that is not generic software (embedded automobile/medical/aerospace software is the exception)

Need a leftpad library? Be prepared to discuss licensing for "intellectual property". No fixed pricing on page, a salesperson will contact you and negotiate.

Re: Ask HN: Getting started in biology with a software background

#77

Post by refurb » Thu, Oct 31, 2019, 5:03 PM UTC

unearthed wrote: post #67

cryoshon wrote: post #38

my background is in biotech and i've worked at a small handful of biotech startups in a scientific capacity. here's my advice: unless you have friends who can set you up with the right VCs and ensure that they will be willing to overlook your lack of experience and IP, don't bother. you're not going to get up to speed working in the lab on your own in any short amount of time. learning the theoretical stuff that you…

Why are Phds so cheap to hire if their skills are so valueable?

Supply > demand

Re: Ask HN: Getting started in biology with a software background

#78

Post by goatlover » Thu, Oct 31, 2019, 5:03 PM UTC

unearthed wrote: post #61

aaavl2821 wrote: post #57

Earlier quoted context omitted.

I think for non scientists looking to work in biotech, it is critical to also understand how the industry works and how value is created. That's the failure point for most biotech startups -- they dont have a useful application of their tech, or the useful application takes costs too much to develop compared to the value it creates A practical way to do this is to invest in biotech stocks. This will expose you to cli…

> "That's the failure point for most biotech startups -- they dont have a useful application of their tech." This is precisely the failure of every academic ever. We can examine this through the Theil-lens where they're trying to create a narrative of uniqueness where their research applies + revolutionizes everything ( when it does not ). But everyone knows this is not the case. This is the central plague of all aca…

> This is the central plague of all academic research, that its the pursuit of novel understanding before useful application.

Uh, not all knowledge is about making products. I hope you meant "all academic research" only in context of biotech startups. And applied science is only part of the scientific endeavor, which is to understand the world, regardless of whether that can be monetized.

Re: Ask HN: Getting started in biology with a software background

#79

Post by aaavl2821 » Thu, Oct 31, 2019, 5:03 PM UTC

iongoatb wrote: post #31

I'm curious who all these sarcastic bio "experts" are that are suggesting getting a PhD or hiring one. I'm a former bio major and researcher/scientist that transitioned to software engineering years ago. I've published papers in bio and worked with many PhDs. Many of them were idiots. Being a PhD doesn't mean anything, it's the independent work that you put in yourself (in an academic lab or by yourself) that determi…

Even better than a home lab is finding a friend who works in a research lab, and shadowing them while they do experiments.

Safety is an obvious benefit to this approach, but it also allows you get exposure to more interesting and complex experiments that are more cutting edge and comparable to the kinds of work you'd actually do at a company. These labs have access to equipment and reagents that you cant get anywhere else. From what Ive seen the at home stuff you can do is incredibly limited

This approach also gives you exposure to proper experimental design and technique, which is really hard to learn on your own because biology experiments take so long. Biology experiments take long enough even if you know how to do them

Re: Ask HN: Getting started in biology with a software background

#80

Post by tikej » Thu, Oct 31, 2019, 5:05 PM UTC

unearthed wrote: post #61

aaavl2821 wrote: post #57

Earlier quoted context omitted.

I think for non scientists looking to work in biotech, it is critical to also understand how the industry works and how value is created. That's the failure point for most biotech startups -- they dont have a useful application of their tech, or the useful application takes costs too much to develop compared to the value it creates A practical way to do this is to invest in biotech stocks. This will expose you to cli…

> "That's the failure point for most biotech startups -- they dont have a useful application of their tech." This is precisely the failure of every academic ever. We can examine this through the Theil-lens where they're trying to create a narrative of uniqueness where their research applies + revolutionizes everything ( when it does not ). But everyone knows this is not the case. This is the central plague of all aca…

> This is the central plague of all academic research, that its the pursuit of novel understanding before useful application.

Plague of all academic research?! Isn't the point of academic research to understand before application (and quite often application is not the goal at all and that's ok)?

I get that it may seem trivial to apply something, but with high-cost risks of mistakes when they happen in biology/medicine it is not.

The pharma industry learned this the hard way: https://en.wikipedia.org/wiki/Thalidomide

It seems to me that some software engineers/computer scientist think that other fields are slow/full stupid people because the progress is not fast. Well the progress is not fast because cost to start is often huge, wetware can't be moved to cloud, stakes are higher than unhappy customers etc.