Gibson Assembly is case sensitive

[jonjoet]

Hi!

Thanks for all your work on this excellent software. I've used it for years as a stringent test of my Golden Gate designs. I recently had occasion to apply it to some Gibson assemblies a colleague had designed, and we discovered a bug: Gibson assembly simulations are case sensitive. That is, valid overhangs are ignored if the sequences are not both lowercase or both uppercase. I've attached a notebook and two data files (including some made up sequences) that illustrate the problem. There are four fragments involved in the test. Two uppercase fragments with valid overhangs, and the same two fragments in lowercase. The notebook attempts to simulate four two-part assemblies. Two of these use parts of the same case, and the other two switch the case of one of the two fragments. Only the same case assemblies work.

Maybe this could be addressed by converting everything to uppercase as part of the sequence_to_string function within (line 44) the HomologyChecker module. I hope you're able to fix this bug; it would be great to have DNA cauldron as a final check on our Gibson designs.

Thanks,
JT

case_comparison.txt
toy_case_test.csv
toy_case_assemblies.csv

[veghp]

Thank you! We actually had a similar problem a while ago (#20) and probably should have checked for the other assembly types.

I'll look into this.

[veghp]

Related to that, I'm planning a review and update on the Gibson assembly method, so if you have any feature requests or ideas please let me know here (or as another issue), or via email (see address on main page https://github.com/Edinburgh-Genome-Foundry).