Establishing Locus Boundaries: First you have to assess how you are going to build your payloads:
Starting from BAC: This is the simplest case for both payload assembly and LP delivery since, in all likelihood, you can use the same pair of gRNA sequences to transfer your payload from a BAC to an assembly vector as well as to target/integrate your LP into the genome.
Amplifying region of interest by PCR: This might be more complicated as there are primer design constrains. However, in most cases you should be able to first identify the pair of gRNAs for LP integration into the genome, and then use those gRNA binding sites as the primer binding sites for your first and last assembly pieces. If your gRNAs are 'pointing outwards' (the 5' gRNA binds the minus strand and the 3' gRNA binds the plus strand), then you should have 17 bps of each gRNA binding site included in your payload, which should be sufficient for primer design.
Design gRNAs flanking the desired locus: Using www.guidescan.com online design tool, display all possible gRNAs flanking the target locus and identify candidate gRNAs according to the considerations below. Note: you can design more than one gRNA for each locus terminus and test different gRNAs combinations in cells before choosing the best pair. To this end, it is helpful to design the different candidate gRNAs to bind close to each other (<300 bps ideally), so that all gRNA combinations can be assessed using a single PCR assay with primers binding outside the most distal gRNAs on each side.
Considerations:
High efficiency and specificity scores
Mappability and repetitive sequences: avoid repetitive sequence, especially distal to your gRNA binding sites (in the nearest ~300-500 bps). Repetitive sequences would hinder primer design for genotyping and homology arms amplification, as well as sequence mapping. You can inspect the mappability of your locus flanks for human (hg38) in the "Mapping and Sequencing" Hub > "Mappability" track. If you are working with mouse (mm10), you will need to add this link to your hub collection in the IGS Genome Browser (My Data > Tack Hubs > My Hubs): https://bismap.hoffmanlab.org/trackhub/hub.txt. In both cases, you should be looking at Umap Tracks (Bismap corresponds to bisulfite-converted DNA mapping), preferably 36bp single-read mappability (Umap S36).
Orientation (relevant primarily when assembling payloads using PCR amplicons. See above)
If working with hybrid cell lines you can design allele-specific gRNAs to target your LP to only one allele. You can filter your list by gRNAs overlapping at least 1 SNP/indel, or better yet, gRNAs with PAM overlapping at least 1 SNP/indel. Note: You can use the following script to help you build the UCSC tracks to visualize the gRNAs in the genome browser.
If working with hybrid cell lines, choose the locus boundary in a region that is rich in variants. This will facilitate designing allele-specific genotyping primers and identifying sequencing reads that overlap variants.
Design homology arms (HAs):
HAs are needed when integrating landing pads (LPs). Considerations:
Size: the longer the HA, the better the efficiency of integration. However, the longer the HAs, the harder it is to genotype the junction across the HAs (especially with crude gDNA as a template) and the less likely it is that short-read sequencing (Illumina) will provide reads that span the junction. Generally, we had good success with HAs of ~250 bp, but we've also used HAs as short as 100 bp and as long as 1 kb.
Repetitive elements: it's probably best to avoid repetitive elements (REs) within the HA, but sometimes this is unavoidable. It's definitely important to avoid REs flanking (distally) the HAs, as they would interfere with genotyping primers and sequencing mapping. A good practice is BLATing the HAs to make sure it is genomically unique.
BsaI site: when cloning HAs into the pLP vectors we'll be using a BsaI-based Golden Gate reaction. Ideally there would not be any BsaI sites in the HA region. There's a UCSC browser track for all restriction enzyme called Restr Enzymes.
Design PCR primers flanking the target locus to validate deletion and/or landing pad integration:
Considerations:
Use NCBI's Primer Blast to validate the specificity of your genotyping primers. In the field 'Database' select 'Genomes for Selected Organisms' and input your organism in the next field.
If possible, design allele-specific primers.
Order BAC(s): you want to order a BAC clone (or several) that cover the entire engineered locus. This BAC will be used as a template for amplifying homology arms as well as a source to prepare bait for capture sequencing. BACs take a while to ship, so you want to do that early as possible.