CODEHOP: COnsensus-DEgenerate Hybrid Oligonucleotide Primers

CODEHOP help

General information

The multiple-sequence alignments should be of amino acid sequences of the proteins and be in the Blocks Database format Proper alignments can be obtained by different methods.

The result of the CODEHOP program are suggested degenerate sequences of DNA primers that you can use for PCR. You have to choose appropriate primer pairs, get them synthesized and perform the PCR.

A CODEHOP primer is degenerate at the 3' core region, with a length of 11-12 bp across four codons of highly conserved amino acids, and is non-degenerate at the 5' consensus clamp region, with a length which depends on its desired annealing temperature, typically between 20 and 30bp:

5'                            3'
--------------------===========

non-degenerate      degenerate
consensus clamp     core
#bases from temp    11-12 bases

The hybrid structure (5' consensus and 3' degenerate) of CODEHOP primers allow the PCR amplification to be specific during the early cycles from the original source DNA and selective during the late cycles from the PCR synthesized products:

Schematic comparison of standard degenerate PCR (left) with the CODEHOP (right), illustrating regions of mismatch in primer-to-template annealing during early PCR cycles and in primer-to-product annealing during subsequent cycles. Vertical lines indicate nucleotide matches between primer (arrow) and template or synthesized product. The overall degeneracy is the product of degeneracies at each nucleotide position, so that the fraction of precisely hybridizing primers is 1/degeneracy.

Obtaining input alignments

You can get a multiple alignment from a group of related protein sequences using the Block Maker or other automated methods (such as ClustalW). The alignments can also be manually made or modified according to your knowledge of the proteins (position of the active sites or post-translational modifications etc.). In any case the alignments passed to the CODEHOP program must be in the Blocks format. BlockMaker blocks need no reformatting. Appropriate parts of Clustal- or FASTA-formatted multiple alignments can be automatically made into blocks by the Blocks multiple alignment processor. Other types of multiple alignment can be semi-manually reformatted with the Blocks formatter. All of the above Blocks programs have links to send the resulting blocks directly to the CODEHOP program and also provide other information for evaluating the blocks (logos, trees, searches).

More information on multiple alignments can be found in the notes for the ISMB97 tutorial on Introduction to making and using protein multiple alignments.

Terms and parameters

• Degeneracy measures how many different sequences are specified by the primer. The degeneracy of each position depends on the sum of the nucleotides appearing in it (1 to 4). The degeneracy of the primer is the product of the degeneracies of all of its positions. The degenerate core region is scored by its degeneracy, with lower degeneracy scored higher. A standard degenerate alphabet is used in the core region.
• Degeneracy strictness specifies how to count nucleotide(s) with low occurrences. It is a parameter between 0 and 1. Strictness of 0 means that all nucleotides that actually appear in the position are counted to calculate the degeneracy. Strictness of 1 means that only the nucleotides with the highest value in the position are counted. Intermediate strictness values give behavior in between. You can look at a detailed explanation for more information on how the strictness is calculated.
• Temperature is computed for the clamp region plus any adjacent non-degenerate positions in the core region. Different temperatures affect only the length of the clamp; higher temperatures result in longer clamps. Temperature is computed using the nearest-neighbor method of Rychlik, et al ( NAR 1990 18:21, 6409-6412). This method takes into consideration the primer and K+ (potassium ions, taken as 50 mM) concentrations.
• Nucleotide runs in the clamp region can be limited (the default is 5 in a row). If longer runs are encountered, they are broken up by taking the next highest- scoring nucleotide in the position with the lowest-scoring run nucleotide.
• The clamp score is a measure of how well the non-degenerate 5' clamp matches the block given a codon usage table. The score is a weighted average of the maximum scores in each column of the DNA PSSM. A perfect score of 100 is obtained if the clamp consists of invariant methionines and tryptophans, and a minimum score of 25 is obtained if A, C, G and T are equally likely at all positions. In computing the clamp score, positions are increasingly down-weighted as distance from the degenerate 3' core increases. The clamp score is intended to estimate how well the clamp will stabilize the core in annealing to the target templates. All else being equal, the higher the clamp score the better. However, the clamp score is not relevant to stability of the clamp/product hybrid, which is estimated by the annealing temperature using the nearest neighbor method.
• The core/clamp boundary can be restricted to always be on a codon boundary.
• Genetic code tables are used to back-translate from protein to DNA.
  
• Codon usage tables are derived from the Codon Usage Database.
  Available tables:: The codon usage table is used to convert the amino acid PSSM constructed from the input blocks into the DNA PSSM. It can also be used to determine the content of the clamp region.
• Clamp residues can be selected as the most common codons of the consensus amino acids are used in the clamp. Else the clamp residues are the ones with maximum weight in the DNA PSSM, which may result in 'artificial' codons. If the clamp residues are taken from the DNA PSSM and more than one residue in a position have maximum weight, one of them is selected at random to keep the clamp non-degenerate.
• The program output can show all possible primers or only the single most degenerate primer in each region.

Your target sequence(s) might be expected to be more similar to some specific sequence(s) in the input blocks. In this case you can bias the primers towards these sequences. Rather than raising the degeneracy or strictness, increase the weights of the specific sequences. The weights are the values following each sequence segment of the block. Usually the highest weight is 100. In the CODEHOP input window increase the weights of your specific sequences (say to 3 or 4 times the original weight or to 200 or 400). You can also remove individual sequences from the input blocks by down-weighing them to 0 (the minimal weight) if they are too divergent and prevent finding primers.

For amplification, we recommend using AmpliTaq Gold with a 9' preheat (this provides an automatic hotstart - a hotstart of some kind is important). We have had success using the time-release feature with the addition of 15-20 extra cycles. The primer finding strategy of the CODEHOP program (Rose, et al, manuscript accepted by NAR) is different from the usual degenerate PCR strategies. It is desirable to keep annealing temperatures high - 60^oC is OK if you have a 60^oC clamp. We recommend trying the highest temperature that yields a clean PCR product. We have used "touchdown" PCR down to Tm-3^oC or lower, say from 63^oC down to a good clamp annealing temperature in -0.5 to -1^oC increments, and the remaining cycles are carried out at the 53-57^oC clamp annealing temperature for a 60^oC clamp. The intent of the touchdown is to give the correct product a head start, because it is likely to anneal at a higher temperature than any failure product. Once the clamp 'takes over', then all primed products, whether correct or not, will be on an even footing, so we try to keep the stringency high in all cycles. With luck, it should not be necessary to gel-purify product, but may rather try cloning directly from the reaction mix if a single band of the expected size is obtained.

We and a few other users were already successful in using the CODEHOP strategy and program to amplify various sequences from complicated and diverged genomes. Please let us know if you have any tips to pass on based on your experience using CODEHOP-predicted primers.

Publication

Genes identified using CODEHOP.

Page last modified Mar 2003