lattice-automation / repp Goto Github PK

Some goals to make REPP easier to install and use. Breaking changes.

Easier install

• Dockerize
• Prune git history a bit, it's too large, doesn't fit with external db model discussed below
• Stop including iGEM and other databases with REPP install

Easier use

• Remove embedded blast/primer3 binaries (lookup in PATH)
• Set defaults for all client settings
• Embed enzymes, no internal database
• Expose all client config options thru CLI flags (ie don't require users to change the config file to change their settings)
• Actual f'ing logging

Local sequence databases

• install sequence databases from FASTA files
  • all databases are used by default during sequence assembly
• Remove --igem and --dnasu flags, etc, from the repp make cmds
• Move iGEM, DNASU, Addgene sequence databases to S3 for now

Right now, the only supported assembly method is Gibson Assembly. BsaI and/or BpiI restriction enzymes could be added on the outsides of primers so each fragment anneals with its neighboring fragment through digestion and ligation rather than homology arms.

Functionality:

• addition of BsaI and BpiI to the ends of the primers
  • consider the added BsaI and BpiI sequence when designing/selecting for primers
  • add additional seq on the 5' end, beyond the BsaI and BpiI sites
• data-driven selection of overhangs between sets of fragments, eg (Potapov, 2018)
  • avoid duplicate or cross-specific overhangs
  • select for best binding overhangs, avoid weak overhangs like TATA
  • avoid assemblies where a set of valid and desired overhangs is impossible to select
• reduce the allowable overlap between fragments to just the 4bp of the neighboring overhang

Changes:

Most of these changes will be made in the primer3.go file: https://github.com/Lattice-Automation/repp/blob/master/internal/repp/primer3.go

How overlap between fragments is calculated and how the overlap is selected will need to be included and overwrite the current consideration about overlap lengths.

Maybe create a separate Golang SBOL parser in https://github.com/SynBioDex

Interesting project! I'm curious how you sourced the addgene, igem, and dnasu databases in the s3 bucket.

Thanks for making this very useful tool. Right now, getting the installation right is not trivial. It would be better to have a more detailed installation instruction based on platform (Mac, Windows, Linux). A web server of this tool would probably be the best, so that users who do not wish to deal with installation can benefit from using this tool.