Hi, After the segmentation of CGH data in some papers the results are frequently shifted by the density mode. To be more precise the mode of the highest peak is used. However this procedure depends on the condition that there is clearly one prominent peak dominating the density function. Currently, in some of my samples, I do have the problem of two prominent peaks flanking the y-axis which make the decision about the correct shift direction a difficult one. Moreover in some of the cases a shift in one direction seems to be obvious, in some other cases a shift in the other direction seems more preferable and in a third group the preference is not quite clear. But in all groups a segmentation profile in chromosomes 1-3 is nearly identical which suggests that I do observe the same gain or loss (depending on the shift direction) in all these samples. Does anyone have an idea how to assess this problem and how to solve it? Is there another frequently used procedure aside the density mode shifting used for such data? I do have pictures of some samples displaying the problem but they are too big for the mailing list. Is there an official repository I can upload them to? Thanks in advance, best regards, Benjamin ====================================== Benjamin Otto University Hospital Hamburg-Eppendorf Institute For Clinical Chemistry Martinistr. 52 D-20246 Hamburg Tel.: +49 40 42803 1908 Fax.: +49 40 42803 4971 ====================================== -- Pflichtangaben gem?? Gesetz ?ber elektronische Handelsregister und Genossenschaftsregister sowie das Unternehmensregister (EHUG): Universit?tsklinikum Hamburg-Eppendorf K?rperschaft des ?ffentlichen Rechts Gerichtsstand: Hamburg Vorstandsmitglieder: Prof. Dr. J?rg F. Debatin (Vorsitzender) Dr. Alexander Kirstein Ricarda Klein Prof. Dr. Dr. Uwe Koch-Gromus

On Mon, Jun 30, 2008 at 7:02 AM, Benjamin Otto <b.otto at="" uke.uni-hamburg.de=""> wrote: > Hi, > > After the segmentation of CGH data in some papers the results are frequently > shifted by the density mode. To be more precise the mode of the highest peak > is used. However this procedure depends on the condition that there is > clearly one prominent peak dominating the density function. > > Currently, in some of my samples, I do have the problem of two prominent > peaks flanking the y-axis which make the decision about the correct shift > direction a difficult one. Moreover in some of the cases a shift in one > direction seems to be obvious, in some other cases a shift in the other > direction seems more preferable and in a third group the preference is not > quite clear. But in all groups a segmentation profile in chromosomes 1-3 is > nearly identical which suggests that I do observe the same gain or loss > (depending on the shift direction) in all these samples. > > Does anyone have an idea how to assess this problem and how to solve it? Is > there another frequently used procedure aside the density mode shifting used > for such data? > > I do have pictures of some samples displaying the problem but they are too > big for the mailing list. Is there an official repository I can upload them > to? We use the lower of the two modes when there is a "tie". Since CGH data is naturally censored at the low end, this makes some sense, biologically, to do. However, this is not a perfect solution nor can there usually be one, as heterogeneity appears to be almost universally present. So, there is little chance of determining the absolute copy number which translates to an inability to absolutely determine the correct mode. Sean

On Mon, Jun 30, 2008 at 8:08 AM, Benjamin Otto <b.otto at="" uke.uni-hamburg.de=""> wrote: > I was hoping there would be a solution, which incorporates some global information about all samples. Suppose I would only have one samples with not so obvious shifting but a segmentation following all others clearly, then I could use this information for the correction. Unfortunately it's not so easy in my case. Yet it naturally makes a difference if my samples group 50 percent to 50 percent according to gain and loss calls versus when they agree on one of the solutions. > You are free to come up with a solution that in some way minimizes discordance between samples, but I do not believe that such a solution is a generally biologically tenable hypothesis unless there is a very special sample set with some well-understood biology. The underlying goal of mode-centering is to minimize the number of probes that are called gains or losses for an individual sample; if you do not accept that premise for your samples, then you will likely need to come up with another method. > How do I know in that case from a pure bioinformatical point of view that my processing procedures provide reliable results and not some sort of gambling result? > I have a feeling that the proportion of ambiguous cases using something like mode-centering is small, so I doubt that such cases will globally affect results. Of course, as with any statistical method or heuristic, it is important to assess the extent to which global results might be biased. If one chooses the "lower mode" as we usually do, for example, then the data will be skewed away from homozygous deletions. We do scan the data by eye for these types of features for just this reason. Perhaps other folks on the list will have a better answer for you. Sean > -----Urspr?ngliche Nachricht----- > Von: seandavi at gmail.com [mailto:seandavi at gmail.com] Im Auftrag von Sean Davis > Gesendet: Monday, June 30, 2008 1:55 PM > An: Benjamin Otto > Cc: bioconductor at stat.math.ethz.ch > Betreff: Re: [BioC] gains and losses via mode shifting > > On Mon, Jun 30, 2008 at 7:02 AM, Benjamin Otto > <b.otto at="" uke.uni-hamburg.de=""> wrote: >> Hi, >> >> After the segmentation of CGH data in some papers the results are frequently >> shifted by the density mode. To be more precise the mode of the highest peak >> is used. However this procedure depends on the condition that there is >> clearly one prominent peak dominating the density function. >> >> Currently, in some of my samples, I do have the problem of two prominent >> peaks flanking the y-axis which make the decision about the correct shift >> direction a difficult one. Moreover in some of the cases a shift in one >> direction seems to be obvious, in some other cases a shift in the other >> direction seems more preferable and in a third group the preference is not >> quite clear. But in all groups a segmentation profile in chromosomes 1-3 is >> nearly identical which suggests that I do observe the same gain or loss >> (depending on the shift direction) in all these samples. >> >> Does anyone have an idea how to assess this problem and how to solve it? Is >> there another frequently used procedure aside the density mode shifting used >> for such data? >> >> I do have pictures of some samples displaying the problem but they are too >> big for the mailing list. Is there an official repository I can upload them >> to? > > We use the lower of the two modes when there is a "tie". Since CGH > data is naturally censored at the low end, this makes some sense, > biologically, to do. However, this is not a perfect solution nor can > there usually be one, as heterogeneity appears to be almost > universally present. So, there is little chance of determining the > absolute copy number which translates to an inability to absolutely > determine the correct mode. > > Sean > > > > -- > Pflichtangaben gem?? Gesetz ?ber elektronische Handelsregister und Genossenschaftsregister sowie das Unternehmensregister (EHUG): > > Universit?tsklinikum Hamburg-Eppendorf > K?rperschaft des ?ffentlichen Rechts > Gerichtsstand: Hamburg > > Vorstandsmitglieder: > Prof. Dr. J?rg F. Debatin (Vorsitzender) > Dr. Alexander Kirstein > Ricarda Klein > Prof. Dr. Dr. Uwe Koch-Gromus > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor