[BioC] Problems with aCGH package

[grischa tödt]

Dear all,

I am interested in identifying different chromosomal states (gain,loss..) in
array CGH data.
Therefore I want to use the aCGH package, but I run into some problems.
(I have tested Version 1.04 (stable) and the developmental Version 1.0.6)
Maybe I can get some hints on how to proceed....


a) when I try to impute missing values (impute.lowess) I get the following
errors:

>Error in if (kb >= chrominfo$centromere[chr]) median(vec[clones.info$Chrom
==  : 
>        missing value where TRUE/FALSE needed

and additional in warnings() I find:

>1: calling par(new=) with no plot
>2: Collapsing to unique x values in: approx(lowess(kbl[ind], vecl[ind], f =
smooth), xout = kbl[-ind])
>3: Collapsing to unique x values in: approx(lowess(kbl[ind], vecl[ind], f =
smooth), xout = kbl[-ind])
...

it just works fine if I remove all the NAs from my data...but for sure
that’s not the way the function is supposed to work ;o)


b) when I try to find the hmm.states (find.hmm.states) I get this errors:

>hmm(ex.acgh) <- find.hmm.states(ex.acgh)
>Error in solve.default(z$hessian) : system is computationally singular:
reciprocal condition number = >9.65772e-017 In addition: There were 50 or
more warnings (use warnings() to see the first 50) Error in >nlm(likel, p,
hessian = TRUE, print.level = print.level, typsiz = typsiz,  : 
>        non-finite value supplied by nlm In addition: There were 50 or more
warnings (use warnings() to >see the first 50) Error in nlm(likel, p,
hessian = TRUE, print.level = print.level, typsiz = typsiz,  : 
>        non-finite value supplied by nlm In addition: There were 49
warnings (use warnings() to see >them) Error in solve.default(z$hessian) :
system is computationally singular: reciprocal condition >number =
1.74131e-017 In addition: There were 50 or more warnings (use warnings() to
see the first 50) >Error in nlm(likel, p, hessian = TRUE, print.level =
print.level, typsiz = typsiz,  : 
>        non-finite value supplied by nlm In addition: There were 50 or more
warnings (use warnings() to >see the first 50) Error in nlm(likel, p,
hessian = TRUE, print.level = print.level, typsiz = typsiz,  : 
>        non-finite value supplied by nlm In addition: Warning messages: 
>1: NA/Inf replaced by maximum positive value
>2: NA/Inf replaced by maximum positive value
>3: NA/Inf replaced by maximum positive value Warning messages: 
>1: NA/Inf replaced by maximum positive value
>2: NaNs produced in: sqrt(diag(cov))
>3: Coercing LHS to a list
>4: Coercing LHS to a list
>5: NA/Inf replaced by maximum positive value
>6: NA/Inf replaced by maximum positive value
>7: NaNs produced in: sqrt(diag(cov))
>8: NA/Inf replaced by maximum positive value
>9: NA/Inf replaced by maximum positive value
>10: NaNs produced in: sqrt(diag(cov))

Here I'm totally lost...


c) when I try to run the demo I get these errors
.

>demo(“aCGH”)
>        demo(aCGH)
>        ---- ~~~~
>Type  <Return>   to start : 
> if (dev.cur() <= 1) get(getOption("device"))()
> opar <- par(ask = interactive() && (.Device %in% c("X11",
>    "GTK", "windows", "quartz")))
> datadir <- system.file("examples", package = "aCGH")
> latest.mapping.file <- file.path(datadir, "human.clones.info.Jul03.csv")
> ex.acgh <- aCGH.read.Sprocs(system(paste("ls -1", 
>    file.path(datadir, "*.txt")), intern = TRUE), latest.mapping.file, 
>    chrom.remove.threshold = 23)
>Error in system(paste("ls -1", file.path(datadir, "*.txt")), intern = TRUE)
: 
>        ls not found


I suggest to use the built-in dir() function from R. It should be system
independent
..



kind regards,
Grischa
 

--------------- Deutsches Krebsforschungszentrum -------------

Grischa Tödt

Molecular Genetics (B060)                       room TP3 1.105
German Cancer Research Center       phone ++49 (0)6221 42-4582
Im Neuenheimer Feld 580               fax ++49 (0)6221 42-4639
D-69120 Heidelberg                       email g.toedt at dkfz.de
Germany