PCR Primer Design (11/28/2014)

Thu, 11/27/2014 at 2:10 am

Customer: I would like to the designs of a little more than 500 RT-PCR primers to make a cDNA library for sequencing. The basic workflow will be: RNA extraction from human cells --> cDNA library creation and amplification --> sequencing on Illumina. Ideally, these primers should be around 30 bases long. The product length to be around 250-500bp from the 5' end when possible. I understand that this will not be possible for all genes, and longer transcripts are acceptable, but not optimal. Since we amplify the cDNA, it will be double stranded DNA and the second strand (=identical to RNA) will be our target, e.g. in the following target transcript (gene sequence below) , a primer for the sequence 5' AGCCTCAGACCTGCTGGGGG that would be reverse & complementary would be the following: 5' CCCCCAGCAGGTCTGAGGCT. Here is what needs to be done: 1. Get all human mRNA from NCBI FTP site (ftp://ftp.ncbi.nlm.nih.gov/genomes/H_sapiens/RNA). 2. Make in-memory map of all 30-mers present in all mRNAs. 3. Select all unique 30-mers, i.e. ones that are found only in target sequence and nowhere else for all targets. Avoid ones that are similar to and may prime on non-target sequences as well. 4. From these select ones that produce cDNA of appropriate length (250-500).

Thu, 11/27/2014 at 3:36 PM

AccuraScience LB: Some questions are centered around the quality of primers designed. (1) The procedure you described does not involve checking of melting temperature or G/C content, and it does not check for low complexity sequences (e.g., simple repeats such as AGAGAGAG). Are you sure for your purposes, we do not need to worry about these factors? (2) The procedure also does not check for intra-primer homology. Would it be a good idea to avoid, e.g., >3 bases complementary within the same primer? (3) How do we determine two primers are too close? Should we use a simple percentage homology cut-off, or more sophisticated criteria are needed?

Fri, 11/28/2014 at 5:39 AM

Customer:1.Regarding the GC content low complexity regions and melting/annealing temperatures: Please keep the GC content within the 40-60% range. I would prefer to stay under 65% GC. In our hands, amplification has been less efficient above 65%.

Please avoid low complexity regions.

I would prefer them to be within a tight range given that they will be used together. I don't know how tight a range you will be able to get, but I would prefer the temperature to be between 57 and 63C with a range of less than 5C, if possible.

2. Regarding your question of self-complementarity: I would like the values of that to be below 3 on the scale I explain below (taken directly from Primer3 Batch's web site):

The maximum allowable 3'-anchored global alignment score when testing a single primer for self-complementarity, and the maximum allowable 3'-anchored global alignment score when testing for complementarity between left and right primers. The 3'-anchored global alignment score is taken to predict the likelihood of PCR-priming primer-dimers. The scoring system is as for the Max Complementarity argument. In the examples above the scores are 7.00 and 6.00 respectively. Scores are non-negative, and a score of 0.00 indicates that there is no reasonable 3'-anchored global alignment between two oligos. In order to estimate 3'-anchored global alignments for candidate primers and primer pairs, Primer assumes that the sequence from which to choose primers is presented 5'->3'. It is nonsensical to provide a larger value for this parameter than for the Maximum (local) Complementarity parameter because the score of a local alignment will always be at least as great as the score of a global alignment.

3.We can start by using sequence homology, but if you can think of a weighting system that works better, I am open to suggestions.