Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: PlasCAD: Open-source plasmid editor (github.com/david-oconnor)
101 points by the__alchemist 15 days ago | hide | past | favorite | 24 comments
Hey! This is an open source plasmid and vector (Short, often circular segments of DNA) editor, with features related to primer quality checks, PCR cloning, and protein analysis. I plan to add more restriction-enzyme-based features in the near future. It has some extras like a solution-mixing helper, automatic feature annotation.

From a technical standpoint, this is a standalone binary written using the EGUI library in rust. A project goal is performance, with small memory footprint, and small application and file sizes.

This is a continuous work-in-progress, and I'm open to any and all feedback, criticism, and requested features.




Cool tool!

It brings back memories: one of my favourite high level undergraduate course finals involved being given some instructions, printouts of sequence/primers/enzymes cleavage sites/other plasmids and a rough list of my resources and told to make a specific product. The exam was long answer essay writing out step by step instructions with rationale.

I did the whole thing and double checking realized I misread an initial sequence in the first few steps making the whole thing void and probably more or less difficult than intended. I added an oh shit disclaimer in my solution. The prof, a researcher, to his immense credit presumably took off some points but the point of the exercise was testing a skillset and I did excellent in the course. Basically how higher education should be and night and day difference from 1000/2000 level courses!


In college, I took one lab course different from all the others after a senior told me it was life changing.

In this lab the class was broken into teams and you were handed a smelly pile of bacteria. You were told the species and then had to read the literature, to learn how to purify specific restriction endonuclease (an enzyme that cuts DNA at a specific location based on its sequence). You were given access to basic molecular biology lab equipment. At the end, each group shared its RE and everybody used everybody's RE to map a plasmid. This is 1970s state of the art, but by the 90s was pretty much routine.

The class kicked my ass. I struggled at every step and barely managed to reach the end. Lab work is quite hard and it can be challenging to master the techniques. A coworker was moved to tears during a challenging step involving a purification using a membrane bag, but reasoned that since tears are salty, they probably helped the purification when they landed in the solvent.

I passed, but barely, and it took me ages to learn even the basics of doing stuff in the lb. it's one of the things that convinced me robots will do a better job in the lab than any human, although in retrospect, it's actually quite hard to build a robot that's better than an average lab tech.


I had a few professors who would actually give full points in these situations because knowing that you’re wrong is often way more important than being right. If you went through a process mostly correctly and then identified that you were wrong and explained it well enough you got nothing taken away.


This is cool! I've worked on a tool like this for a major biotech, its nice to see an open source version with a decent feature set.

Looks like your other work is related to UAVs. How did you find this problem space? What was the inspiration for this tool?


Great question. I'm on a molecular biology deepdive (Have a reasonably-stocked at-home lab etc) Overall, I love technology, and biology is some very sophisticated technology! I think Drexler is correct that it is key to opening a world of nanotechnology.

This looks very nice. I don't do cloning, but various amplicon sequencing things. I had often used AliView, which is nice for alignments but can't mark regions or primers. I also like the Benchling sequencing editor, but that's online and may go away/charge money.

A few questions:

Can I make the UI light instead of dark?

The 'simulate PCR' works, but seems to unload the original sequence. Can I get that back? This is a bit unexpected.


I appreciate the feedback!

A: Eventually I think I'll add a light mode

B: Great call on the PCR. The cloning, btw, has similar behavior. I'm adding a multi-tab system that should hopefully fix this. I also noticed several bugs related to PCR amplification that need fixing too; should have a new version out later this week.


Update: The latest version (0.7.2) includes the updated PCR (and cloning) behavior, opening the amplicon/product in a new tab instead.

Looks cool, and definitely like the standalone binary format.

Tangentially, glad to see more biotech stuff finally making the front on HN! :D


Can't wait until I can 3D print my own proteins.


It would be cool to have a compact, cost-effective automated recombinant DNA + protein expression/purification workflow! To do this currently involves a series of steps, reagents, equipment etc.


One prion disease, please!


This is basically what this is. It is very cheap to get plasmids manufactured which you can then put into e coli or whatever which will build your proteins for you.


You can! You're doing it right now!


Exciting!


Sorry is this is a lame Q:

Could you take the sequences and python up a blender script that will model the thing?

Also, on solutions-mixer- can I use this to map proteins to feed to stem cells to get them to present in a certain way?


Not a lame question at all. For modeling, are you referring to protein products, or something different? If so, the answer is yes for known proteins; it includes a workflow to download structure coordinates from PDB based on an amino-acid sequence query. You can currently click a button to download these coordinates, and another to open a browser with the PDB sequence viewer open to it.

It would be possible, as you said, to add functionality to open a script using Python/Blender or similar to view these, or even integrate 3D graphics directly in the program. Drawing color-coded molecules would be straightforward, but cartoon drawings etc would take more work. Modelling unknown/custom proteins would presumably require something like AlphaFold integration.

The solutions mixer currently is just for molarity calculations. Like, you are making a buffer that contains 300mM of this, 100mM of that. It will tell you how many grams or mL to add of each reagent. That's a great idea re more practical workflows like presenting in stem cells! Should add more to that effect.


There was a research from UCSF that I met when we were building that DOgPatch facility in Sf...

She had written a program in python to pipette tiny amounts of proteins via tubular matrices valve thingy she made which allowed her to feed various proteins to stemcells via her little 3D printed tubes and then was researtching what each proteins made the stemcells behave like, what they presentd, her quote was "what do we need to ffed the stemcell to become a heart muscle."

Ive alwasy been intriugued by this as I want to experiment with my own stemcells.... specifically in ocular regen.

But having AI craft blender snippets has been super simple to get it to draw geometry - so a GPT that can translate between some chemistry constraints to 3D- molocule constraints - plus there are a bunch of 3D chem libraries (there was an interesting one on HN a while back that I commented on a bit..)

What would be interesting would be a version of 'meta clay' (balls) -- that had attraction rules that werent meant to merge Nurbe surfaces, but rather it would know the attractions of various proteins/molocules/biochemicals such that it will know which 'covalant bonds' (I am puling that phrase out of my non-chem butt) -- work together.


[flagged]


Are you sure they're not the same? Plasmids strike me as the ideal medium for implementing plasmids.


As MatrixMan pointed out, these plasmids works somewhat like that in Bacteria - they take up one of these plasmids, and gain superpowers. It's really cool! (Examples: Antibiotic resistance, virulence, producing light etc. Not sure about the shock plasmid yet though)


This install part REALLY needs to improve. Give me screenshots of how to use this on Windows. Explain step by step. Does installation requires administrative credentials or I can run this as a lowly tech on the locked down computer the laboratory has?

Overall the readme should cater for biologists, biochem and others that don't necessarily are well versed into computer science. Just explain the basics to get things going.

Also a video would help too - in addition, not as sole resource.

Looks interesting but too out of reach for a regular biochem grad to pickup as is.


Hey! This is a standalone executable; you download the zip file, unzip it, and double-click to run. I was thinking about instead making it into a minimal installer that would move the program to C:\Program Files\PlasCAD or similar, then set up shortcuts and file associations. (There is something similar for Linux).

Did you hit a snag in that process? This sort of feedback is helpful, as you may have hit issues I haven't when testing.


I think it’s just someone being overly critical in a way that’s more than a bit silly.


1. download 2. extract 3. double click to run




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: