Ask HN: Getting started in biology with a software background

Having a PhD is not essential -- I know many successful VCs, founders and operators without PhDs -- but you absolutely need an appreciation of science. People like to hire PhDs because they have a fundamental knowledge of biology and or chemistry and they have deep experimental experience in a specific domain. Without that experience doing experiments it's hard to really understand the challenges of science and rigor required. That said not all PhDs give you that, and academic science tends to be less rigorous than industry science

Also, developing a drug requires PhD level experience in many domains and no one person can do it all. A neuroscience PhD won't necessarily qualify you to review tox data. You need people who can quickly get up to speed on various technical fields, find experts and have productive conversations with them

You need to understand the drug development process at a basic level. There are lots of articles about this online, here's one I wrote [0]

You need to understand how to evaluate and critique scientific and clinical data. It is easier to start by looking at clinical data. You can learn to analyze this data without a PhD if you spend time with it and ideally have a mentor. Here is a case study I wrote on basic concepts in evaluating clinical data [1]

Evaluating preclinical or scientific data is much harder. In nonhuman studies you are measuring more endpoints with less direct relationship to human disease in more contrived systems. The experiments also tend to be less rigorously designed, executed and documented (at least in academia) so there are all kinds of pitfalls to avoid that you can't really know without experience. This is where a PhD really helps

To start learning this just struggle through papers. Find a paper that interests you and read it in depth. Learn what each experiment and instrument does. Learn what each molecule does. Here's an example I wrote translating a synthetic biology paper into layman studies terms to give you a sense of what goes into reading a paper [2]

If you have a scientist friend who can walk you through papers that speeds up the process by orders of magnitude and you can learn so much

At first read for comprehension. Then read with a critical eye. Why did they do this experiment instead of this other one? Are they missing an important control here? Is this model robust? Are the conclusions they draw stronger that what the data suggest?

Learn how drugs are valued. Valuation is different in biotech than any other sector. Drugs don't have revenue or users for years. A drug can. A worth $10B before FDA approval. Value in biopharma comes from reducing technical risk. I wrote a post about that here [3]. Most biotech founders lack either the science or business knowledge needed. It is easier to learn the business stuff so that can let you add value to a company while learning more about the science

[0] https://www.baybridgebio.com/blog/drug_dev_process.html

[1] https://www.baybridgebio.com/blog/aducanumab-analysis

[2] https://www.baybridgebio.com/blog/synbio-laymans-terms.html

[3] https://www.baybridgebio.com/drug_valuation.html

I mean, by all means, try it. But life sciences are not computer science. The approach is entirely different and quality of the work you need to do is different. It looks much, much easier than it is. (Which is, on a different note, why I believe the whole pseudoscience crap such as anti-vaxxers is gaining so much traction)

Not all of the places on that list will have biology labs, but some will! It's a good way to get some experience at the bench without going the degree route.

It is important to have a fundamental understanding of biology and chemistry, and ability to critically evaluate data, but once you have that, its more important to be able to find people with the right domain expertise and have productive interviews with them about data. If you are a neuroscience PhD you should not be evaluating a cancer assay or a GLP tox study on your own, you should read enough to have a productive interview with someone with more expertise, then go have that interview

You need a rigorous science background to start a company but that is not sufficient. Many projects are just not fundable. For non scientists, id recommend learning how the industry works and how value is created, because that complements the expertise of scientists who know the science but dont know the right clinical applications