repp is a tool for DNA assembly. It takes a target plasmid and finds the least expensive combination of fragments from DNA repositories to create it via Gibson Assembly.

Biologists profit when they can re-use DNA during plasmid design: it enables cheaper designs and faster builds. But parsing through all re-usable DNA is completely infeasible. For example, there are over 75,000 plasmids in Addgene -- the likelihood of knowing the best combination and ordering of sub-sequences from Addgene for a given plasmid design is low.

repp does such plasmid design. It turns specifications into assembly plans that use the least expensive combination of existing (PCR) and newly synthesized DNA fragments. Target plasmids are specifiable using their target sequence, features, or sub-fragments.

Features

• fragment selection: given a plasmid, repp finds the least expensive assembly of fragments including:
  • PCR fragments: you provide the databases of sequences that you have access to for PCR
  • synthetic fragments: the cost of fragment synthesis is configurable for both fixed and variable length pricing
  • synthetic plasmids: repp recommends synthesizing the entire plasmid if it is the cheapest option
• primer selection: repp chooses primers via Primer3 that have:
  • minimal off-target binding
  • minimal Primer3 penalty scores (secondary structure, uneven Tms, etc)
• assembly efficacy filters: when choosing assemblies, repp filters for those with desirable Gibson Assembly characteristics:
  • limits hairpin structures in fragment junctions/overlaps
  • minimum and maximum fragment junction/overlap lengths
  • limits the total number of fragments
• configurability: every setting above is configurable

Publication

We published a paper about repp in PLOS One: Timmons, J.J. & Densmore D. Repository-based plasmid design. PLOS One. We used it to build thousands of plasmids from iGEM and Addgene and showed that it reduced the cost of plasmid design as compared to synthesis.

Examples

See /examples to see input/output from repp.

Documentation

See the docs or --help for any command.

Installation

From Docker

Run repp via Docker:

mkdir -p $HOME/.repp
alias repp="docker run -i --rm --mount type=bind,src=$HOME/.repp,dst=/root/.repp jjtimmons/repp:latest"
repp --help

From Source

Note: repp depends on:

• Go >= 1.18.0 for compilation
• BLAST+ for sequence alignment
• Primer3 for hairpin detection and off-target primer-binding detection

git clone https://github.com/Lattice-Automation/repp.git
cd repp
make install

Sequence Databases

repp uses sequence databases for plasmid assembly. These are added as FASTA files along with the name and cost per plasmid from that source.

Some existing FASTA files are maintained in our S3 bucket repp. Below is a snippet for downloading and adding each to repp:

for db in igem addgene dnasu; do
  curl -o "$db.fa.gz" "https://repp.s3.amazonaws.com/$db.fa.gz"
  gzip -d "$db.fa.gz"
done

# add sequence DBs with the cost of ordering a plasmid from each source
repp add database --name igem --cost 0.0 < igem.fa
repp add database --name addgene --cost 65.0 < addgene.fa
repp add database --name dnasu --cost 55.0 < dnasu.fa

Plasmid Design

Sequence

To design a plasmid based on its target sequence save it to a FASTA file. For example:

>2ndVal_mScarlet-I
CAACCTTACCAGAGGGCGCCCCAGCTGGCAATTCCGACGTCTAAGAAACCATTATTATCA...

Then call repp make sequence to design it. The following example uses Addgene and a local BLAST database parts_library.fa as fragment sources:

repp make sequence --in "./2ndVal_mScarlet-I.fa" --dbs "addgene,parts_library.fa"

Features

To design a plasmid based on the features it should contain, specify the features by name. By default, these should refer to features that are in repp's feature database (~/.repp/features.tsv). Features can also refer to fragments, as in the following example where a plasmid is specified by its constituent list of iGEM parts:

repp make features "BBa_R0062,BBa_B0034,BBa_C0040,BBa_B0010,BBa_B0012" --backbone pSB1C3 --enzymes "EcoRI,PstI" --dbs igem

Fragments

To design a plasmid from its constituent fragments, save them to a multi-FASTA.

>GFP
ATGAGTAAAGGAGAAGAACTTTTCACTGGAGTTGTCCCAATTCTTGTTGAATTAGATGG...
>backbone
TACTAGTAGCGGCCGCTGCAGTCCGGCAAAAAAGGGCAAGGTGTCACCACCCTGCCCTT...

And call the file from repp make fragments:

repp make fragments --in "./fragments.fa" --out "plasmid.json"

Configuration

The default settings file used by repp is in ~/.repp/config.yaml. The maximum number of fragments in an assembly, the minimum overlap between adjacent fragments, and cost curves for synthesis are all defined there. Editing this file directly will change the default values used during plasmid designs.

To overwrite some repp settings on a per-design basis, create another YAML file:

# custom_settings.yaml
fragments-min-junction-length: 25
synthetic-fragment-cost:
  1800: # max length in bp
    fixed: false
    cost: 0.07 # per bp cost

And reference it during plasmid design:

repp make sequence --in "./2ndVal_mScarlet-I.fa" --dbs addgene --settings "./custom_settings.yaml"

Backbones and Enzymes

The plasmid sequence in the input file is designed as a circular plasmid by default. In other words, repp assumes that the sequence includes an insert sequence as well as a backbone. To use the sequence in the input file as an insert sequence but another fragment as a backbone, use the --backbone and --enzymes command in combination. This will lookup --backbone in the fragment databases and digest it with the enzyme selected through the --enzymes flag. The linearized backbone will be concatenated to the insert sequence. For example, to insert a GFP_CDS sequence into iGEM's pSB1A3 backbone after linearizing it with PstI and EcoRI:

repp make sequence --in "./GFP_CDS.fa" --dbs addgene,igem --backbone pSB1A3 --enzymes "PstI,EcoRI"

The largest linearized fragment post-digestion with all enzymes is used as the backbone in the Gibson Assembly.

Output

repp saves plasmid designs to JSON files at the path specified through the --out flag. Below is an abbreviated example of plasmid design output:

{
  "target": "2ndVal_mScarlet-I",
  "seq": "CAACCTTACCAGAGGGCGCCCCAG...",
  "time": "2019/06/24 11:51:39",
  "solutions": [
    {
      "count": 2,
      "cost": 236.65,
      "fragments": [
        {
          "type": "pcr",
          "cost": 94.67,
          "url": "https://www.addgene.org/103998/",
          "seq": "ACAAATAAATGTCCAGACCTGCAG...",
          "pcrSeq": "ACAAATAAATGTCCAGACCTGCAG...",
          "primers": [
            {
              "seq": "ACAAATAAATGTCCAGACCTGCA",
              "strand": true,
              "penalty": 4.406456,
              "pairPenalty": 18.046225,
              "tm": 58.594,
              "gc": 39.13
            },
            {
              "seq": "CATATGTATATCTCCTTCTTAAATCT",
              "strand": false,
              "penalty": 13.639769,
              "pairPenalty": 18.046225,
              "tm": 52.36,
              "gc": 26.923
            }
          ]
        },
        {
          "id": "103998-103998-synthesis-1",
          "type": "synthetic",
          "cost": 129,
          "seq": "AGGAGATATACATATGGTGAGTAA..."
        }
      ]
    }
  ]
}

Contact Us

Do you have a feature request? Do you wish there were better documentation, examples, or a web-server to run repp against? Please create a new issue in this repo, and we will improve the tool.