Dye swap normalization

  • Adanar

    Adanar - 2010-06-22

    I am new in MeV analysis but I am wondering.
    When I do a flip-dye normalization in the end I get a file where I have IA and IB values for each gene.

    But where do these values refer to?
    I have a Control and  Treatment condition in my arrays and I want to know whether IA refers to the Control and IB to the Treatment or vice-versa.

  • John Braisted

    John Braisted - 2010-06-22


    IF your dye swap analysis is done in TM4's Midas then the merged file (the single output file produced) has an orientation of the first of the two files that you select when selecting your flip-dye file pair. This is an important point because it means that the flip dye result, at least the orientation of samples to IA and IB, depends on file selection order.

    One of the two input files probably has a preferred or desired dye orientation.  We (at JCVI) usually make Cy3 correspond to our reference sample and end up in IA in the mev file.  This is just our convention. You might choose another way but TM4 software will take log2(IB/IA) as the ratio so IA as reference is a good choice.

    If you've run a bunch of samples through Midas, I believe that the txt report file and maybe the concatenated output file name will reveal which input file was selected as the first file in the pair.  This will dictate the orientation.  If a result is wrong in terms of orientation you can run that pair again OR you can swap the header names making IA->IB and the original IB->IA.  One additional tip, if you have replicate hyb results, you can often spot this error in the MeV heat map if the ratios seem to go in opposite directions, red becomes green and green becomes red.

    We don't have a sourceforge forum for Midas but midas@jcvi.org is one option for now.

    Best Regards,



Log in to post a comment.