We are in the process of moving the development of primer3, its web interface, and it documentation to SourceForge . If you are interested in helping out at any level (for example, C code, perl code, web design, release engineering, enhancements, user documentations), please post e-mail to primer3-mail (at) lists.sourceforge.com or send e-mail to steverozen (at) gmail.com, and thank you!
Department of Bioinformatics at Institute of Molecular and Cell Biology University of Tartu provides a Web-based primer picking interface at http://bioinfo.ut.ee/primer3-0.4.0, which is provided as a service to the research community, and it is not supported by dedicated funds.
Other organizations with which we have no connection also offer the Web interface.
All code and the Department of Bioinformatics at Institute of Molecular and Cell Biology University of Tartu Web service are offered on an "as-is", use-at-your-own-risk basis.
ld: can't locate file for -lcrt0.o
If you check the "Show Debugging Info" box in the Web input form, then the results will include the exact input given to primer3_core.
The nearest-neighbor model of oligo melting temperatures is a good fit for short oligos, but presumably breaks down for longer oligos. Therefore, there is a more obscure parameter, in primer3_main.c, called MAX_TM_NN_LENGTH, which indicates the maximum length for which to use the nearest neighbor model when calculating oligo Tmís. In primer3_main.c as provided, this is set to 36 (the default value of MAX_PRIMIER_LENGTH). For oligos over MAX_TM_NN_LENGTH, the program uses a GC% formula. See oligotm.h, long_seq_tm() for the formula. We do not know the maximum reasonable value for MAX_TM_NN_LENGTH. Also, if you design oligos of various lengths, you will get a discontinuity in oligo Tmís between MAX_TM_NN_LENGTH and MAX_TM_NN_LENGTH+1. If you are thinking about hybridization to oligos attached to a surface, as far as we know, the modeling of this has lagged, so the GC% formula might be just fine.
There has been much interest in recent years in designing oligos for hybridization, for example, for spotted arrays for mRNA expression profiling. There may be off-the-shelf software that is better fitted to your needs. In particular, there are two packages that offer documented models of secondary structure that are more detailed than those in primer3, which may be important for longer oligos:
If you are designing oligos for expression profiling, you may want to check them for cross hybridization to other transcripts (anything in the rest of the transcriptome). OligoArray 2.1 does this. We are not sure about Oligo Design Platform.
Disclaimer: we have no first-hand experience with these packages.