4 ms·

Show HN: Web app for DNA sequence manipulation and PCR Primer design

40 points by folli 10y ago

kevinwuhoo 10y ago
This is really cool! In my experience, few scientific tools are designed with modern interfaces. Most are a webpage with many, many inputs and textareas and some expectation of the user to understand or trust the defaults (looking at you BLAST [1]). I haven't designed a primer in quite a while, but doing a cursory Google search for "primer design" confirms that things haven't changed much. I know that writing and maintaining user authentication code can be quite a slog, but have you though of making a private/paid version of this? I can see a few times where something like this could be useful. Ex: assembling a number of sequences, annotating them, and having all that metadata inline to share with colleagues? Maybe having some kind of Jupyter Notebook inoperability would be enough? [1]: https://blast.ncbi.nlm.nih.gov/Blast.cgi https://blast.ncbi.nlm.nih.gov/Blast.cgi
- whistlerbrk 10y ago
  About ten years ago I started down this path myself, it just seemed overwhelming though to recreate all the functionality of the desktop systems like Vector NTI on the web. With today's tools it is likely much better. It does seem as you're alluding to though that a Jupyter like format would work really well. I wonder what the state of the art is today at least on desktop.
  - jfarlow 10y ago
    Unfortunately, software has not much advanced beyond where it was ten years ago. There have been a lot of attempts, but the fact is that it's very difficult to see returns on mass-market scientific software - the market is relatively small compared to the likes of Snapchat. The software tooling certainly makes it easier to make the scientific tools - but you still have to earn enough to support a company (and compete against 'decent' freeware). Desktop software - outside of very custom huge pharma-style companies, or free-ware academic python libraries is a non-starter - as no one will download it. The nicest Desktop software I know of is SnapGene: http://www.snapgene.com/ http://www.snapgene.com/ which has to compete against the ancient (and pretty good) freeware, APE: http://biologylabs.utah.edu/jorgensen/wayned/ape/ http://biologylabs.utah.edu/jorgensen/wayned/ape/ On the web there is YC's own Benchling: https://benchling.com/ https://benchling.com/ Twist (Startup DNA synthesizer) just bought Genome Compiler: http://www.genomecompiler.com/ http://www.genomecompiler.com/ Teselagen for those hardcore genetic engineers: https://www.teselagen.com/ https://www.teselagen.com/ Our own Pinecone: https://serotiny.bio/ https://serotiny.bio/ (styled to be a 'simple' abstraction above the DNA to produce functional proteins while bypassing cloning) And Desktop Genetics (for CRISPR design): https://www.deskgen.com/ https://www.deskgen.com/ The older players (VectorNTI, and-the-rest-I-can't-rememeber) are never updated and likely yet exist to service contracts from huge institutional customers. It's a relatively small (and dispersed) community of users right now. We need a place that scientists actually go to to show off all of our nice shiny new toys. Each comes with their own style and with their own strengths - it'd be fun to showcase them all together.
    - jfarlow 10y ago
      Just because it's funny - direclty from Vector NTI's FAQ on their webpage, today, in April 2017: 1. When is Vector NTI Advance™ 11.5 available? Version 11.5 was launched on October 15, 2010. 2. Is Vector NTI Advance® 11.5 compatible with Microsoft® Windows 7? Yes. Version 11.5 is compatible with Windows 7 on 32-Bit Systems - 64-Bit compatibility is currently in development. 3. Is Vector NTI Advance® 11.5 compatible with Mac OS? Yes, this requires Mac OS® X running Windows® under Parallels Desktop® 5 or Boot Camp. https://www.thermofisher.com/us/en/home/life-science/cloning/vector-nti-software/vector-nti-advance-software/vector-nti-frequently-asked-questions.html https://www.thermofisher.com/us/en/home/life-science/cloning...
      - kevinwuhoo 10y ago
        That's absurd! Still the same version I was using when I was in grad school. I've heard that Benchling is a decent alternative. Although few, if any, companies have the marketing and sales resources as Thermo.
    - sndean 10y ago
      > The nicest Desktop software I know of is SnapGene: http://www.snapgene.com/ http://www.snapgene.com/ Have to agree with that, though it's a bit expensive.
      - toufka 10y ago
        You're (not even you, your lab is) paying for niche, specialized, professional software - it's not "accounting software" or "photo software" or "chat software". Out of curiosity, how much do you expect professional "genetic engineering software" to cost? Between $270-700 for a permanent license seems like a pretty good deal to me.
        sndean 10y ago
        > (not even you, your lab is) Well, I had to pay for it out of pocket. Got an annual license. But yeah, I get your point.
- folli 10y ago
  I'm not yet sure in which direction I want to take thia project. I don't think there's a big chance to monetize something like this, given the multitude of software that already exists in this field. The Jupyter notebook is a really cool idea, I'll have to look into this.
JohnJamesRambo 10y ago
I can't wait to use this to create phylogenetic trees when I need some. Thanks for posting!
- folli 10y ago
  Be aware that this is a quick-and-dirty way to get a tree. It's based on the MUSCLE guide tree to get a quick glimpse. It doesn't replace a real phylogenetic tree algorithm such as RAxML.
jfarlow 10y ago
Nice work. Getting all of these tools in one place, in an easy to use way is not always easy. I'm well aware of what goes on behind the scenes to make something like your alignment there as straightforward to use. The tricky part is what next :) I'm curious what your plans are - interoperability, private infrastructure, plugins/extensibility for others or just simply nice tools for people to use. I'd also be curious how you intend to approach the differences and distinctions between protein and DNA sequences. It doesn't look like you currently have much to do with an amino acid sequence. At Serotiny we've chosen to focus on the amino acid sequences for manipulation and leave the DNA in the background. It feels like it's often a trade-off - either/or. It's a challenge to get both into the same working environment without overwhelming the user. I'd be curious your thoughts on what we've built: https://serotiny.bio https://serotiny.bio btw - it's nice to see some ShowHNs show up for biological software.
- folli 10y ago
  Thanks for the encouraging words! I don't have any future plans yet, depends on the feedback I get and if there's enough interest for further features. I think you're touching a valid point, but it goes even further. There are many different use cases for a bio-software; a bench-worker has different requirements than a geneticist, microbiologists vs cell biologists etc, WGS and RNA-seq need an even more different toolset. It's impossible for a single software to cover all possible aspects of biology. The only thing you can do is focus on one group of users and try to fullfill their requirements.
  - jfarlow 10y ago
    yep :) Unfortunately, the dearth of tools means everyone (coming from as many directions) feels like they must rebuild every tool - so there is a bit of a convergence right now - which spreads every tool thin. But the market for 'RNA-seq tools' is but a fraction of 'biochemical scientists', which itself is still a 'tiny' market to produce high-tech software for.
etimberg 10y ago
Cool! I built something similar to the find tool for a friend. However, it's for RNA and specifically for looking for a region that could base pair as part of an RNA secondary structure interaction. It allows you to search for all possible locations that could pair given a certain number of mismatches. It can also correlate against MFOLD predicted RNA secondary structures to eliminate possibilities that are not in single stranded regions. Tool: http://everttimberg.io/app/rna/ http://everttimberg.io/app/rna/
- folli 10y ago
  Nice tool! I also was planning to integrate Mfold into the toolset. However I put it on hold due to lack of time.
bhickey 10y ago
Thanks, better tools are a godsend. Back in school I spent too much time using PrimerX (http://www.bioinformatics.org/primerx/ http://www.bioinformatics.org/primerx/).

40 points by folli 10y ago

This is really cool! In my experience, few scientific tools are designed with modern interfaces. Most are a webpage with many, many inputs and textareas and some expectation of the user to understand or trust the defaults (looking at you BLAST [1]). I haven't designed a primer in quite a while, but doing a cursory Google search for "primer design" confirms that things haven't changed much. I know that writing and maintaining user authentication code can be quite a slog, but have you though of making a private/paid version of this? I can see a few times where something like this could be useful. Ex: assembling a number of sequences, annotating them, and having all that metadata inline to share with colleagues? Maybe having some kind of Jupyter Notebook inoperability would be enough? [1]: https://blast.ncbi.nlm.nih.gov/Blast.cgi https://blast.ncbi.nlm.nih.gov/Blast.cgi

whistlerbrk 10y ago
About ten years ago I started down this path myself, it just seemed overwhelming though to recreate all the functionality of the desktop systems like Vector NTI on the web. With today's tools it is likely much better. It does seem as you're alluding to though that a Jupyter like format would work really well. I wonder what the state of the art is today at least on desktop.
- jfarlow 10y ago
  Unfortunately, software has not much advanced beyond where it was ten years ago. There have been a lot of attempts, but the fact is that it's very difficult to see returns on mass-market scientific software - the market is relatively small compared to the likes of Snapchat. The software tooling certainly makes it easier to make the scientific tools - but you still have to earn enough to support a company (and compete against 'decent' freeware). Desktop software - outside of very custom huge pharma-style companies, or free-ware academic python libraries is a non-starter - as no one will download it. The nicest Desktop software I know of is SnapGene: http://www.snapgene.com/ http://www.snapgene.com/ which has to compete against the ancient (and pretty good) freeware, APE: http://biologylabs.utah.edu/jorgensen/wayned/ape/ http://biologylabs.utah.edu/jorgensen/wayned/ape/ On the web there is YC's own Benchling: https://benchling.com/ https://benchling.com/ Twist (Startup DNA synthesizer) just bought Genome Compiler: http://www.genomecompiler.com/ http://www.genomecompiler.com/ Teselagen for those hardcore genetic engineers: https://www.teselagen.com/ https://www.teselagen.com/ Our own Pinecone: https://serotiny.bio/ https://serotiny.bio/ (styled to be a 'simple' abstraction above the DNA to produce functional proteins while bypassing cloning) And Desktop Genetics (for CRISPR design): https://www.deskgen.com/ https://www.deskgen.com/ The older players (VectorNTI, and-the-rest-I-can't-rememeber) are never updated and likely yet exist to service contracts from huge institutional customers. It's a relatively small (and dispersed) community of users right now. We need a place that scientists actually go to to show off all of our nice shiny new toys. Each comes with their own style and with their own strengths - it'd be fun to showcase them all together.
  - jfarlow 10y ago
    Just because it's funny - direclty from Vector NTI's FAQ on their webpage, today, in April 2017: 1. When is Vector NTI Advance™ 11.5 available? Version 11.5 was launched on October 15, 2010. 2. Is Vector NTI Advance® 11.5 compatible with Microsoft® Windows 7? Yes. Version 11.5 is compatible with Windows 7 on 32-Bit Systems - 64-Bit compatibility is currently in development. 3. Is Vector NTI Advance® 11.5 compatible with Mac OS? Yes, this requires Mac OS® X running Windows® under Parallels Desktop® 5 or Boot Camp. https://www.thermofisher.com/us/en/home/life-science/cloning/vector-nti-software/vector-nti-advance-software/vector-nti-frequently-asked-questions.html https://www.thermofisher.com/us/en/home/life-science/cloning...
    - kevinwuhoo 10y ago
      That's absurd! Still the same version I was using when I was in grad school. I've heard that Benchling is a decent alternative. Although few, if any, companies have the marketing and sales resources as Thermo.
  - sndean 10y ago
    > The nicest Desktop software I know of is SnapGene: http://www.snapgene.com/ http://www.snapgene.com/ Have to agree with that, though it's a bit expensive.
    - toufka 10y ago
      You're (not even you, your lab is) paying for niche, specialized, professional software - it's not "accounting software" or "photo software" or "chat software". Out of curiosity, how much do you expect professional "genetic engineering software" to cost? Between $270-700 for a permanent license seems like a pretty good deal to me.
      - sndean 10y ago
        > (not even you, your lab is) Well, I had to pay for it out of pocket. Got an annual license. But yeah, I get your point.
folli 10y ago
I'm not yet sure in which direction I want to take thia project. I don't think there's a big chance to monetize something like this, given the multitude of software that already exists in this field. The Jupyter notebook is a really cool idea, I'll have to look into this.

JohnJamesRambo 10y ago

I can't wait to use this to create phylogenetic trees when I need some. Thanks for posting!

folli 10y ago
Be aware that this is a quick-and-dirty way to get a tree. It's based on the MUSCLE guide tree to get a quick glimpse. It doesn't replace a real phylogenetic tree algorithm such as RAxML.

jfarlow 10y ago

Nice work. Getting all of these tools in one place, in an easy to use way is not always easy. I'm well aware of what goes on behind the scenes to make something like your alignment there as straightforward to use. The tricky part is what next :) I'm curious what your plans are - interoperability, private infrastructure, plugins/extensibility for others or just simply nice tools for people to use. I'd also be curious how you intend to approach the differences and distinctions between protein and DNA sequences. It doesn't look like you currently have much to do with an amino acid sequence. At Serotiny we've chosen to focus on the amino acid sequences for manipulation and leave the DNA in the background. It feels like it's often a trade-off - either/or. It's a challenge to get both into the same working environment without overwhelming the user. I'd be curious your thoughts on what we've built: https://serotiny.bio https://serotiny.bio btw - it's nice to see some ShowHNs show up for biological software.

folli 10y ago
Thanks for the encouraging words! I don't have any future plans yet, depends on the feedback I get and if there's enough interest for further features. I think you're touching a valid point, but it goes even further. There are many different use cases for a bio-software; a bench-worker has different requirements than a geneticist, microbiologists vs cell biologists etc, WGS and RNA-seq need an even more different toolset. It's impossible for a single software to cover all possible aspects of biology. The only thing you can do is focus on one group of users and try to fullfill their requirements.
- jfarlow 10y ago
  yep :) Unfortunately, the dearth of tools means everyone (coming from as many directions) feels like they must rebuild every tool - so there is a bit of a convergence right now - which spreads every tool thin. But the market for 'RNA-seq tools' is but a fraction of 'biochemical scientists', which itself is still a 'tiny' market to produce high-tech software for.

etimberg 10y ago

Cool! I built something similar to the find tool for a friend. However, it's for RNA and specifically for looking for a region that could base pair as part of an RNA secondary structure interaction. It allows you to search for all possible locations that could pair given a certain number of mismatches. It can also correlate against MFOLD predicted RNA secondary structures to eliminate possibilities that are not in single stranded regions. Tool: http://everttimberg.io/app/rna/ http://everttimberg.io/app/rna/

folli 10y ago
Nice tool! I also was planning to integrate Mfold into the toolset. However I put it on hold due to lack of time.

bhickey 10y ago

Thanks, better tools are a godsend. Back in school I spent too much time using PrimerX (http://www.bioinformatics.org/primerx/ http://www.bioinformatics.org/primerx/).