sequencing primer lengths

Linan Chen lichen at
Thu Sep 20 13:29:25 EST 2001


The following guide is from University of chicago. You may go this link for
reference. Also, for your convenience, I copy and paste some content here.


One of the single most important factors in successful automated DNA
is proper primer design. It is important that a primer has the following
·       A melting temperature (Tm) in the range of 52 C to 65 C 
·       Absence of dimerization capability 
·       Absence of significant hairpin formation (>3 bp) 
·       Lack of secondary priming sites 
·       Low specific binding at the 3' end (ie. lower GC content to avoid
Primers designed according to these criteria will generally be from 18 to 30
bases in length and have %GC of 40 to 60. Try to avoid using primers with Tm's
above 65-70 C, especially on high GC templates, as this can lead to secondary
priming artifacts and noisy sequences. We strongly recommend the use of
computer software to design primers with these characteristics. Examples of
such software are: LaserGene (DNAStar), Oligo (National Biosciences, Inc.),
MacVector (Kodak/IBI) and the GCG suite. In addition, there is a web site
available for designing PCR primers using the Primer program (Whitehead
Institute). In lieu of software, the following equation can be used to roughly
estimate Tm: 
    Tm = 59.9 + 0.41*(%GC) - 600/length
If designing a primer based on existing sequencing data, choose a priming site
that is greater than 50 nucleotides away from the position where new sequence
is needed. Avoid designing primers using regions of poorer quality sequence,
such as areas beyond single peak resolution of a chromatogram (typically
600-700 bases). Avoid primers where alternative priming sites are present with
more than 90% identity to the primary site or that match at more than seven
consecutive nucleotides at the 3' end. 
Finally, be aware that no set of guidelines will always accurately predict the
success of a primer. Some primers may fail for no apparent reason, and primers
that appear to be poor candidates may work well. 


More information about the Methods mailing list