4 ms·

Ask HN: Startup opportunities in bioinformatics

12 points by volodia 14y ago

I'm a first-year computer science PhD student currently working in optimization and machine learning. For my PhD, I have a long-term goal of building a commercializable piece of technology (an algorithm, a piece of software, etc.) and hopefully turn it into a company. One potential area for this that seems exciting is bioinformatics. I was wondering if anyone here had any insight into developing bioinformatics technologies and turning them into a startup. Is bioinformatics a good field for this? Are there any examples of comp. bio. technologies that have been successfully commercialized? Right now, most bioinformatics research that I see in universities seems to be more on the pure science side. I feel like in this kind of research, progress is extremely incremental, and one individual's work will have a noticeable impact only within a small community of academics. Also, let me know if there are other promising areas besides bioinformatics that I should consider for what I would like to do. Thank you!

aheilbut 14y ago
There seems to be quite a lot of interest right now in companies trying to build platforms for cloud-based 'genetic-analysis-as-a-service', such as DNAnexus. The data management/storage and scaling problems in genomics are very real, but to a large extent these issues are not really specific to bioinformatics. There are also companies working on software for NGS, such as Real Time Genomics and Novocraft. Historically, pure-play bioinformatics startups have had a pretty tough time, perhaps because the primary users have been academics who are loathe to pay for software. I think many of the successful ones also involve proprietary databases (eg. Ingenuity) or have deep scientific experience and expertise (eg. OpenEye). The real opportunities at the moment are more likely in companies that are tightly integrating genomics technologies with diagnostics and services. There are many, but for example, look at Foundation Medicine in cancer, or Counsyl or Natera in prenatal diagnostics.
dkural 14y ago
I run one of these companies, Seven Bridges Genomics. I'd encourage you to learn more about genomics, biology and the problems these fields are trying to solve. Technology is useful only insofar it solves a problem in a big market -- so instead of starting with a technology / picking an area; I'd start with a problem.
kirk21 14y ago
Don't know if you consider it as bioinformatics but tissue 3D printing is an interesting field. http://forums.reprap.org/list.php?194 http://forums.reprap.org/list.php?194 If you want to keep working on machine learning, your progress also depends on which databases you can access. Other ideas: - Predicting how cells evolve and grow over time. - Predicting how viruses spread
johncoogan 14y ago
I personally don't know much about bioinformatics but what I do know about potential commercial applications has come from hearing speakers from Singularity University discuss what they are working on. You might be interested in looking into http://singularityu.org/ http://singularityu.org/ or their FutureMed conference http://futuremed2020.com/ http://futuremed2020.com/ Good luck!
subrat_rout 14y ago
First thing you can do is to 1. Contact few bioinformatics professor/researcher from your university (If your university has a bioinformatics dept) or other universities and ask what are the problems nagging them for a long time. You will have an idea. You can also contact industries (23andme). For more info go to http://bionet.awardspace.info/company.html http://bionet.awardspace.info/company.html and see what bioinformatics companies do and what interests you. 2. In healthcare it is just an open field. You can take open data from data.gov and analyze and if you can develop an algorithm which can predict the likelihood of hospitalization post surgery in a patient then insurance companies will be drooling over that.

12 points by volodia 14y ago

aheilbut 14y ago

There seems to be quite a lot of interest right now in companies trying to build platforms for cloud-based 'genetic-analysis-as-a-service', such as DNAnexus. The data management/storage and scaling problems in genomics are very real, but to a large extent these issues are not really specific to bioinformatics. There are also companies working on software for NGS, such as Real Time Genomics and Novocraft. Historically, pure-play bioinformatics startups have had a pretty tough time, perhaps because the primary users have been academics who are loathe to pay for software. I think many of the successful ones also involve proprietary databases (eg. Ingenuity) or have deep scientific experience and expertise (eg. OpenEye). The real opportunities at the moment are more likely in companies that are tightly integrating genomics technologies with diagnostics and services. There are many, but for example, look at Foundation Medicine in cancer, or Counsyl or Natera in prenatal diagnostics.

dkural 14y ago

I run one of these companies, Seven Bridges Genomics. I'd encourage you to learn more about genomics, biology and the problems these fields are trying to solve. Technology is useful only insofar it solves a problem in a big market -- so instead of starting with a technology / picking an area; I'd start with a problem.

kirk21 14y ago

Don't know if you consider it as bioinformatics but tissue 3D printing is an interesting field. http://forums.reprap.org/list.php?194 http://forums.reprap.org/list.php?194 If you want to keep working on machine learning, your progress also depends on which databases you can access. Other ideas: - Predicting how cells evolve and grow over time. - Predicting how viruses spread

johncoogan 14y ago

I personally don't know much about bioinformatics but what I do know about potential commercial applications has come from hearing speakers from Singularity University discuss what they are working on. You might be interested in looking into http://singularityu.org/ http://singularityu.org/ or their FutureMed conference http://futuremed2020.com/ http://futuremed2020.com/ Good luck!

subrat_rout 14y ago

First thing you can do is to 1. Contact few bioinformatics professor/researcher from your university (If your university has a bioinformatics dept) or other universities and ask what are the problems nagging them for a long time. You will have an idea. You can also contact industries (23andme). For more info go to http://bionet.awardspace.info/company.html http://bionet.awardspace.info/company.html and see what bioinformatics companies do and what interests you. 2. In healthcare it is just an open field. You can take open data from data.gov and analyze and if you can develop an algorithm which can predict the likelihood of hospitalization post surgery in a patient then insurance companies will be drooling over that.