[R] Stepwise rQTL-unknown warning message and odd QTL curve

[Claire O'Quin]

Sorry, I'll try to provide more detail about what I have done so far with
code and any relevant output results.

>library(qtl)
>sawfly.cross <- read.cross(format="csv",
file="~/Desktop/Sawfly_data/QTL/Sawfly_QTL.csv", na.strings="NA",
genotypes=c("A", "B"), alleles=c("A", "B"), estimate.map=F)
--Read the following data:
 430  individuals
 506  markers
 19  phenotypes
 --Cross type: bc

>print(sawfly.cross)
--This is an object of class "cross".
  It is too complex to print, so we provide just this summary.
    Backcross

    No. individuals:    430

    No. phenotypes:     19
    Percent phenotyped: 99.8 99.8 99.3 99.1 99.1 99.1 99.1 99.1 99.5 99.8
99.8 99.5 98.8 99.8 99.8 99.8 99.8 98.4 99.5

    No. chromosomes:    7
        Autosomes:      1 2 3 4 5 6 7

    Total markers:      506
    No. markers:        103 89 75 74 65 51 49
    Percent genotyped:  96.2
    Genotypes (%):      AA:49.7  AB:50.3
    Backcross

    No. individuals:    430

    No. phenotypes:     19
    Percent phenotyped: 99.8 99.8 99.3 99.1 99.1 99.1 99.1 99.1 99.5 99.8
99.8 99.5 98.8 99.8 99.8 99.8 99.8 98.4 99.5

    No. chromosomes:    7
        Autosomes:      1 2 3 4 5 6 7

    Total markers:      506
    No. markers:        103 89 75 74 65 51 49
    Percent genotyped:  96.2
    Genotypes (%):      AA:49.7  AB:50.3

>sawfly.cross <- calc.genoprob(sawfly.cross, step=2.5, error.prob=0.1,
map.function="kosambi", stepwidth="fixed")

**I am using head size as a covariant.**

>head.covar <- pull.pheno(sawfly.cross, pheno.col=19)
>sawfly.cross.stepwise.peryellow <- scantwo(sawfly.cross, pheno.col=2,
model="normal", method="hk", addcovar=head.covar, use="all.obs",
clean.output=F, verbose=T, n.perm=1000, batchsize=100);
save.image("~/Desktop/Sawfly_data/QTL/SawflyQTL.RData")
--Warning messages:
1: In checkcovar(cross, pheno.col, addcovar, intcovar, perm.strata,  :
  Dropping 1 individuals with missing phenotypes.

2: In checkcovar(cross, pheno.col, addcovar, intcovar, perm.strata,  :
  Dropping 1 individuals with missing covariates.

> sawfly.cross.stepwise.peryellow.pen <- calc.penalties(alpha=0.05,
perms=sawfly.cross.stepwise.peryellow)

>> sawfly.cross.stepwise.peryellow.stepqtl <- stepwiseqtl(sawfly.cross,
pheno.col=2, method="hk", max.qtl=10,
penalties=sawfly.cross.stepwise.peryellow.pen , verbose=T,
keeplodprofile=T, covar=head.covar, scan.pairs=F, keeptrace=T)
--Error in covar[!hasmissing, , drop = FALSE] : incorrect number of
dimensions

**I corrected this with the next piece of code

>sawfly.cross.stepwise.peryellow.stepqtl <- stepwiseqtl(sawfly.cross,
pheno.col=2, method="hk", max.qtl=10,
penalties=sawfly.cross.stepwise.peryellow.pen , verbose=T,
keeplodprofile=T, covar=as.data.frame(sawfly.cross$pheno$Head.Area),
scan.pairs=F, keeptrace=T)

The stepwise than ran and I got to the point where I got the warning
message I posted

about:Warning message:
In lastout[[i]] - (max(lastout[[i]]) - dropresult[rn == qn[i], 3]) :
  longer object length is not a multiple of shorter object length

I proceeded to examine the output

>sawfly.cross.stepwise.peryellow.stepqtl
  QTL object containing genotype probabilities.

      name chr    pos n.gen
Q1 1 at 106.1   1 106.11     2
Q2 2 at 180.0   2 179.97     2
Q3 3 at 181.9   3 181.91     2
Q4 3 at 181.9   3 181.91     2
Q5 5 at 142.5   5 142.50     2

  Formula: y ~ sawfly.cross$pheno$Head.Area + Q1 + Q2 + Q3 + Q4 + Q5 +
Q4:Q5

  pLOD:  166.23


In my late night of googling, I did see that the warning can indicate that
dimensions of the arguments do not match, but I do not know how to
translate that to my data or output.

Thank you.

On Sat, May 23, 2015 at 3:36 AM, Uwe Ligges <ligges at statistik.tu-dortmund.de
> wrote:

>
>
> On 23.05.2015 01:07, Claire O'Quin wrote:
> > Hi There,
> >
> > I am running a stepwise QTL for a backcross and got the following warning
> > message:
> >
> > Warning message:
> > In lastout[[i]] - (max(lastout[[i]]) - dropresult[rn == qn[i], 3]) :
> >    longer object length is not a multiple of shorter object length
>
> So dimensions of the arguments may not match?
> >
> > I can not discern what this means. When I created my plot, the QTL curve
> on
> > chromosome 3 is very odd (tried attaching it), so I suspect that the
> > warning is connected to that odd curve plot.
> >
> > I tried running the fitqtl just to see what would happen and got an error
> > (Error in solve.default(t(Z) %*% Z, t(Z) %*% X) : system is
> computationally
> > singular: reciprocal condition number = 1.49755e-24).
> >
> > Any thoughts about what is going on?
>
> No, without knoing what the arguments and the actual code was.
>
> Best,
> Uwe Ligges
>
> >
> > Thank you,
> > Claire
> >
> >
> >
> > -----------------------------------
> > Claire O'Quin, PhD
> > Postdoctoral Research Scholar
> > University of Kentucky
> > http://www.linnenlab.com/home.html
> >
> >
> >
> > ______________________________________________
> > R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> > https://stat.ethz.ch/mailman/listinfo/r-help
> > PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> > and provide commented, minimal, self-contained, reproducible code.
> >
>



-- 
-----------------------------------
Claire O'Quin, PhD
Postdoctoral Research Scholar
University of Kentucky
http://www.linnenlab.com/home.html

	[[alternative HTML version deleted]]