# sbc-assembly

For initial environment setup:

1. Install Vienna RNA (see https://anaconda.org/bioconda/viennarna).
2. Install NCBI Blast (see https://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Web&PAGE_TYPE=BlastDocs&DOC_TYPE=Download).

LCR2

This script assumes that PlasmidGenie has been used to add the necessary entries into the ICE database.

1. Amend the script lcr2.sh, updating ICE URL, username and password, as well as ICE ids to assemble. The optional three-letter code for naming the worklist, currently with a default of LCR, may also be updated.

2. Ensure that all necessary source plates are in the data/plates directory. This is a key issue: SYNBIOCHEM have never managed to develop a searchable database of plates and samples, therefore any plasmids, parts, dominoes, primers, etc. required for assembly need to be specified, along with their well location, in csv files in this directory. Plates that are required include those containing plasmid parts (i.e. individual parts within the plasmids, typically that which is returned from Twist), and those containing primers (these primers are not held in ICE, and must be specified here, using the naming convention of [PART_PLASMID_ID]_P and [PART_PLASMID_ID]_NP for phosphorylated and non-phosphorylated primers respectively. Plates can be defined in one of two formats (see data/plates/12135272.csv and data/plates/11276738.csv) for examples of each format. The filename typically follows the id / barcode of the plate, although any id can be used as barcode reading is not used.

3. Run the script through the command: bash lcr2.sh.

4. Worklists and plates for each of the six steps in the full LCR assembly process with be found in the ./out directory, in a subdirectory named [date]LCR.

The worklists cover each of the six sequential steps in a full automated LCR v2 assembly, and are generated in numbered directories. The steps are:

1. Part PCR. The PCR step required to PCR the part from its parent plasmid (typically, the plasmid ordered by Twist). Note that both parts and plasmids should have their own ICE id if they have been generated correctly through PlasmidGenie). See for example https://ice.synbiochem.co.uk/entry/10041 and https://ice.synbiochem.co.uk/entry/10040. SBC010040 corresponds to the Part, and SBC010041 corresponds to the parent Plasmid. This step essentially PCRs Part SBC010040 from its parent Plasmid SBC010041.

2. Part digest. This step performs a digest on the Parts generated in Step 1 (e.g. SBC010040) to remove unnecessary flanking sequence at the 5' and 3' ends.

3. Part QC. This step takes the digested parts from Step 2 and prepares them for (optional) analysis on the fragment analyser.

4. Domino pooling. This step pools together all dominoes required for the LCR of each pathway. If, for example, a design contains two pathways, two domino pools will be generated.

5. Part digest pooling. Depending on the design, certain Parts are needed in increased volume (for example, a promoter may be needed numerous times). This step pools an appropriate number of digested parts from Step 2, ensuring that there is enough for the LCR stage (Step 6).

6. LCR writer. This step performs the LCR assembly of each pathway in the design, using pooled digested parts from Step 5, and domino pools from Step 4.

LCR3

Very early code to begin to implement new LCR version 3 method.

1. Run the script through the command: bash lcr3.sh.

2. Output is in the form of a number of csv files in the .out/lcr3.

Four files are generated:

1. design_parts.csv. This shows each design, along with the individual parts which need to be LCRed together to make the full design.

2. part_primers.csv. This shows each individual part and the primer required to PCR the part from its parent plasmid (i.e. the plasmid which is purchased from Twist).

3. pair_dominoes.csv. This shows the part pairs (each pair of parts that need to be LCRed together) and the domino that it required for the LCR.

4. summary.txt, which summarises the previous three files in a single text file.

The files part_primers.csv and pair_dominoes.csv essentially contain the "shopping list" for what needs to be bought from Twist.

As mentioned, this is very early code, and will have to be much further developed to integrate with ICE and generate worklists, etc., as is currently done for LCR v2. Depending upon the priorities of the centre, this work may or may not be worth doing.

ENZYME SCREENING

This script reads a manually-generated screening list, and from which generates appropriate worklists.

The worklist essentially just mixes the appropriate cell lysate(s) with the compounds being screened. All screens are run in duplicate by default.

See data/enz_scr/enz_scr.csv for an example of such a screening list.

1. Amend the script enz_scr.sh to update the location of the enzyme screening list, and the optional three-letter code for naming the worklist, which currently has a default of ENZ.

2. Run the script through the command: bash enz_scr.sh.

Worklists are output in the out directory, and are organised according to project. (The project is defined by the Project column in the enz_scr.csv file.)