Non-functional DNA sequences
Background
Sometimes when designing DNA parts (e.g. promoters, RBSs etc.) there is a need to include spacer DNA that has no function other than to separate two other DNA sequences. I've often been unsure how to go about designing such a spacer sequence. To increase the likelihood of designing an adequately non-functional piece of DNA, I thought it would be a good idea to have a checklist of all the types of sequences you should make sure are not in the spacer sequence. This list is unlikely to be comprehensive so please add anything else you can think of. In addition if anyone has any sequences that they know to be entirely non-functional they might post them here.
Sequences to avoid
Don't forget to consider both the spacer DNA and the flanking DNA lest you create an undesirable sequence overlapping your DNA and the flanking DNA sequences.
Operator sequences
Repressor/activator binding sites, σ-factor binding sites and promoters should all be avoided if possible.
Start codons
e.g ATG
RBSs
There is a lot of variability in RBS sequences. The E.coli consensus RBS sequence can be found here
The Shine-Dalgarno sequence is -
TAAGGAGGT
Even a part of this sequence can induce translation if there is a start codon nearby. See Chen et al.[1] for more details.
Restriction Sites
Depending on your construction strategy, you may want to avoid certain restriction sites in your spacer sequence. If you are using the BioBricks standard assembly method for cloning, you need to avoid these restriction sites - EcoRI, XbaI, SpeI and PstI.
Methylation Sites
Depending on your cellular chassis, you may need to avoid sequences that could be methylated. For example, GATC will be methylated in some E. coli strains.