Home      Labs      Publications      People      Tools   




This page is a mirror. Original document is located here.

To reference this server, please cite:

Rachlin J., Ding CM., Cantor C., and Kasif S. (2005).
muPlex: multi-objective multiplex PCR assay design.
Nucleic Acids Research. 33(Web Server Issue):W544-W547.



muPlex is a web-enabled multiplex PCR assay design service for high-throughput genotyping and other applications. The application identifies viable primer pair candidates for each locus of interest. The program selects both a primer pair for each locus and partitions the primer pairs into multiplex-compatible sets, each corresponding to a potential multiplex PCR reaction. Sequence data may be entered manually or uploaded from a file using a variety of formats. Multiple algorithms (agents) running across our network work in the background to generate one or more candidate solutions intended to reveal the design tradeoffs multiple assay design candidates. These tradeoffs include average multiplex level, locus coverage (% of sequences assigned to some tube), number of tubes achieving desired multiplex level, and the number of different tube sizes required. The system is designed to support the addition of new optimization algorithms over time. When solutions are produced, they are emailed to the user in plain text format.

Sequence Data Format

Sequences must be provided in standard FASTA format. This format applies for both manually entered data as well as for data uploaded in a file. Everything on the line following the '>' is considered the Sequence name. Sequence data follows on one or more lines. With the release of Version 2.0, MuPlex no longer requires that the input sequences contain SNPs. SNP locations (if included) are identified either by bracketing the wild type and mutant type alleles or by specifying an IUPAC ambiguity code:




Effective as of version 2.1, you can also specify left and right brackets anywhere within the sequence allowing you to guarantee that forward and reverse primers will be positioned so as to flank the bracketed sequence. In this case the 5'/3' distance parameters apply to the left and right extremes of the bracketed sequence (see below for further details.) For example:


Whatever format is used but be used consistently throughout the file. If you choose to use IUPAC ambiguity codes for SNPs, such codes should only occur once in each sequence, i.e., at the position of the SNP itself to avoid ambiguity. Sequence data can cover multiple lines and spaces within the sequence are automatically ignored. Test files containing random SNP data may be found <a href="muplex_samples.html">here</a>.

Input Parameters

Name Your name. Although this information is optional we would appreciate knowing about who is using our server. Feedback is always most welcome.
Institution Your academic institution or company name.
Email Your email address. A valid email is required because solutions

are emailed in plain text format to the specified address. Your

email address is not used for any other purpose.
Assay Name
(New in 2.2)
The name of your assay or experimental design. This is an optional field provided for your reference in your emailed solution report.
File The name of a fasta-formatted text file containing sequence data

for various SNPs to be multiplexed. The file size limit is currently 500k. If your data exceeds this limit, consider either reducing the amount of flanking sequence provided, or chunking your problem into multiple

Manual Entry SNP data may be manually entered. If the user specifies a file for upload,

any data entered here is ignored. See above for detailed formatting instructions.

Sequence Format Input sequence format. (See above for details).
  • Sequence only: No SNPs or brackets. Primers may be designed anywhere within the sequence.
  • Sequence brackets: Left and right brackets ('[' and ']') surround an extended region in the sequence. Forward and reverse primers will be positioned so as to flank this bracketed region.
  • SNP brackets: Brackets surround a SNP. e.g., [C/G]
  • SNP IUPAC Ambiguity Code: Uses an IUPAC code to designate a SNP mutant and wild type.
  • Mask Regions By default, any sequence data containing NNN's or XXX's are ignored during primer selection.

    In addition, you can have the system ignore lowercase DNA sequence sometimes used to designate low-complexity regions.

    Mplex Target The multiplexing level constituting a "full" tube. The program attempts to partition SNPs into tubes of this size.
    Algorithm muPlex implements multiple algorithms for generating one or more solutions designed

    to reveal tradeoffs among different design objectives. To control the execution of these methods, the user may specify one of three possible choices:

  • One Solution: Produce a single solution. This is usually very fast and can be useful for checking input parameters and the feasibility of obtaining a high-quality solution.
  • Limited Tradeoffs: Executes a number of heuristics designed to generate several viable solutions. While this takes a bit longer, it can be useful for understanding key tradeoffs between coverage and the number of tube sizes required for a specified multiplexing level. May take up to 10x longer.
  • Deep Search: Generates hundreds of solutions and returns the best one. This takes up to 50x longer (potentially hours for large problems) and can produce dozens of solutions, but is more likely to produce highly efficient solutions if such solutions exist.
  • Primer Length The length of the primers (# bases).
    %GC The %GC content of the primer oligos.
    Tm Primer melting temperature. We use the following standard formula for computing melting temperature:

        Tm = ΔH / [ΔS+A+R ln (C/4)] - 273.15 + 16.6 log10[Na+]


    A is a helix initiation correction (-10.8 cal/Mol), C is the default oligo concentration (50 [nM]), [Na+] is the default salt concentration (50 [mM]), R is the molar gas constant (1.987 [cal K-1M-1]), and the hermodynamic parameters based on Breslauer et al., 1986.

    Future versions of muPlex may allow users to specify various parameters, including oligo and salt concentrations. Current defaults are identical to those employed by primer3 (Rozen and Skaletsky,2000)


    Breslauer K, Frank R, Blocker H, and Marky L (1986). Predicting DNA duplex stability from the base sequence. PNAS 83:3746-3750.

    Rozen S, Skaletsky H (2000). Primer3 on the WWW for general users and for biologist programmers. In: Krawetz S, Misener S (eds) Bioinformatics Methods and Protocols: Methods in Molecular Biology. Humana Press, Totowa, NJ, pp 365-386.

    Product Product amplicon size.
    Product Diff The minimum or target amplicon product size difference from one amplicon to the next.

    This feature is useful when trying to distinguish amplicons on a gel. The specified minimum value is the absolute minimum required product size difference between any two pair of amplicons in a single reaction tube. The minimum is treated as an absolute hard constraint. Tubes containing amplicons less than this minimum are never allowed in any solution. The specified 'maximum' is really a target product size difference. MuPlex tries to minimize the number of number of amplicon pairs across all tubes whose product size difference falls below this target threshold. This target is a soft constraint that the optimizer may or may not achieve. For example, if you specify a minimum product size difference of 10 and a target product size difference of 40, no solution will include a tube containing products with a size difference of less than 10 bp and the system will try to make all product size differences greater than or equal to 40 bp.

    3'/5' Dist This parameter defines how close (on the 3' end) or how far

    (on the 5' end) the primer oligo can be to the SNP. If your sequences contain extended brackets rather than single SNPs, the distances refer the left-most bracketed sequence for designing forward primers, and the right-most bracketed sequence for designing reverse primers. Note: If your sequences contain

    no SNPs or bracketed regions, this constraint is ignored.
    Base Repeat The maximum number of repeated bases within a valid primer.
    3' Tail ΔG The worst case ΔG alignment between the 3' tail of

    one primer anywhere along the sequence of another. This parameter is used both to screen primer pairs within a particular locus as well as to identify cross-interactions among primers within

    a single multiplex assay.
    Max Tm Diff The maximum Tm range for all primers within a single tube.
    Align Score The maximum Smith-Waterman local alignment score between any

    two primers in a single tube, or between the forward and reverse primers of any primer pair candidate for any SNP. Greater scores

    constitute a potential cross-hybridization and are disallowed.
    Match For the alignment, the score attributed to complementary bases.
    Mismatch For the alignment, the penalty attributed to non-complementary bases.
    Gap Penalty For the alignment, the penalty attributed to gaps in the alignment.
    BLAT Filter
    Align primer pairs against the human genome using BLAT (Kent 2002; Genome Res. 12:656-664)

    in order to improve primer pair specificity and reduce or eliminate the appearance of artifacts. To turn this feature

    on, check the box and specify the maximum product size to screen for.
    Primer Pairs
    (New in 2.2)
    The maximum number of primers per DNA locus sequence (1-500). The system generates all possible candidate primer pairs and then prunes the list randomly to conform to the specified limit. This is a system performance enhancement to prevent certain kinds of inputs from overwhelming our servers.


    Solutions are emailed to the user as a plain text document. The solution report provides a compartive summary of the best solutions discovered by muPlex. In the solution summary, each row is a particular solution and each column is the evaluation with respect to a particular criteria. Note: If columns in the report do not line up, try configuring your email client to display text documents using a fixed font such as courier. Alternatively, copy and paste the report into a text editor or spreadsheet. Each solution is evaluated with respect to the following objectives:

    • SolnId: Solution Id (1,2,3...)
    • #Tube: The total number of tubes / assays required
    • #Full: The number of full tubes, i.e., those achieving the target multiplexing level
    • MinSz: Minimum size tube (multiplexing level)
    • AvgSz: Average tube size
    • MaxSz: Maximum tube size
    • #Sizes: The number of unique tube sizes
    • CovAll: Coverage-All (% of SNPs whose associated primers are assigned to at least one tube)
    • CovMax: Coverage-Max (% of SNPs whose associated primers are assigned to maximum-sized tubes.) Note: The maximum sized tube may be less than or equal to the target multiplexing level.
    • CloseSzPairs: # of amplicons having product size less than or equal to target product size difference.

    Detailed solutions corresponding to each row in the solution summary are then provided. Each solution specifies the # of tubes (assay) of each size. Then the solution lists the contents of each tube, including the associated SNP or sequence name, amplicon size (bases), the 5' position, length Tm, and GC content of the forward and reverse primers along with the primer sequence itself. The forward primer is listed first followed immediately by the reverse primer.


    Tube#SEQID   ProdSz   5'   Len   Tm   %GC   Primer


    What's new

    Version 2.2

    - Added user-specified limit on the number of primer pair candidates per sequence. This avoids long running times or solution failures that some users experienced for certain types of input.

    - Added an input field allowing users to optionally specify the name of the assay for reference purposes.

    - Enhanced solution report now includes all of your input parameters along with a summary of the number of primer pair candidates found for each sequence.

    Version 2.1

    Added support for a new data format where brackets [ ] can occur anywhere in the sequence. The forward/reverse primers will be designed to flank the bracketed sequence.

    Version 2.0

    - Multiplex assays can now be designed for sequences not containing SNPs.

    - Added support for product-size base pair differences. Users can specified both an absolute minimum that will never be allowed in any single tube as well as a target difference that muPlex will allow but endeavor to avoid if possible.

    - Fixed a bug that caused the system to produce no solutions whenever the number of submitted sequences was less than the target multiplexing level.

    What's planned

    - will add cookie support to remember individually assigned input parameters from one session to the next. If you use a standard set of parameters, you will no longer need to re-enter them each time you visit the muPlex server. Use of the cookies will be entirely optional. If cookies are turned off in your browser, the interface will silently revert to its default parameters.

    - will introduce a muPlex forum (wiki?) to provide support for our expanding list of users!

    - Do you have ideas for improving muPlex? Our on-going development efforts are primarily in response to suggestions made by our users. (See Questions and Feedback below.)

    About us

    muPlex was designed and implented by John Rachlin in collaboration with Chunming Ding, Charles Cantor, and Simon Kasif. This work was supported in part by NSF grants DBI-0239435 and ITR-048715 as well as by NHGRI grant #1R33HG002850-01A1.

    Center for Advanced Genomic Technology
    Center for Advanced Biotechnology
    Bioinformatics at Boston University

    Questions and Feedback

    For questions or feedback, kindly email us.

    Copyright (c) 2005-2006. Boston University. All rights reserved.

    Protein Engineering