[BioC] HTqPCR problem
Heidi Dvinge
heidi at ebi.ac.uk
Tue Oct 26 16:42:12 CEST 2010
Dear Steve,
(answer below)
> Dear Heidi,
>
> I have TLDA data in the following format which has 3 biological replicates
> per file (here is the header)
>
> SDS 2.4 RQ Results 1.2
> Filename 7.10.10 8A.2.sdm
> Assay Type RQ Study
> EmbeddedFile 8A 13.9.10 A
> Run DateTime Mon Sep 13 17:08:13 BST 2010
> Operator
> ThermalCycleParams
> EmbeddedFile 8B 14.9.10 A
> Run DateTime Tue Sep 14 15:02:13 BST 2010
> Operator
> ThermalCycleParams
> EmbeddedFile 8C 15.9.10 A
> Run DateTime Wed Sep 15 14:45:18 BST 2010
> Operator
> ThermalCycleParams
>
> # Plate Pos Flag Sample Detector Task Ct
> Delta Ct Avg Delta Ct ?Ct SE Delta
> Delta Ct SD RQ RQ Min RQ Max Omit HMD FOS LME
> NAW EW BPR HRN HNS EAF
> BAF TAF CAF HSD GBO
> 1 8A 13.9.10 A A1 Passed 8A A hsa-let-7a-4373169
> Target 27.691027 10.129 0.000
> 1.000 false
> 2
>
>
> I read the data in as
>
> raw<-readCtData(files = files$File, path=".",SDS=TRUE,n.data=3)
>
>
> and I get:
>
> > summary(raw)
> 8A 13.9.10 A 8B 14.9.10 A 8C 15.9.10 A 1A 1.10.10 A 1B 2.9.10 A
> Min. "17.43" "17.91" "17.11" "15.83" "16.47"
> 1st Qu. "29.51" "29.48" "29.64" "29.79" "29.56"
> Median "35.62" "34.97" "35.20" "35.41" "35.33"
> Mean "34.37" "34.04" "34.27" "34.30" "34.24"
> 3rd Qu. "40.00" "40.00" "40.00" "40.00" "40.00"
> Max. "40.00" "40.00" "40.00" "40.00" "40.00"
> 1C 3.9.10 A 2A 1.8.10 A 2B 2.9.10 A 2C 3.9.10 A 3A 1.8.10 A
> Min. "17.55" "16.05" "16.47" "17.10" "16.57"
> 1st Qu. "29.61" "29.90" "29.31" "29.49" "29.51"
> Median "35.10" "36.27" "34.86" "35.30" "34.72"
> Mean "34.24" "34.50" "33.97" "34.27" "34.07"
> 3rd Qu. "40.00" "40.00" "40.00" "40.00" "40.00"
> Max. "40.00" "40.00" "40.00" "40.00" "40.00"
> 3B 2.9.10 A 3C 3.9.10 A 4A 2.8.10 A 4B 2.9.10 A 4C 13.9.10 A
> Min. "16.78" "17.26" "16.69" "17.06" "17.30"
> 1st Qu. "29.47" "29.57" "29.68" "29.72" "29.76"
> Median "35.52" "35.20" "34.82" "35.67" "35.62"
> Mean "34.24" "34.21" "34.19" "34.27" "34.40"
> 3rd Qu. "40.00" "40.00" "40.00" "40.00" "40.00"
> Max. "40.00" "40.00" "40.00" "40.00" "40.00"
> 5A 13.9.10 A 5B 14.9.10 A 5C 15.9.10 A 6A 13.9.10 A 6B 14.9.10 A
> Min. "16.79" "17.02" "17.36" "17.12" "16.92"
> 1st Qu. "29.06" "29.85" "29.46" "30.34" "29.44"
> Median "34.71" "35.58" "35.78" "38.08" "35.44"
> Mean "33.92" "34.54" "34.30" "34.88" "34.21"
> 3rd Qu. "40.00" "40.00" "40.00" "40.00" "40.00"
> Max. "40.00" "40.00" "40.00" "40.00" "40.00"
> 6C 15.9.10 A 7A 13.9.10 A 7B 14.9.10 A 7C 15.9.10 A
> Min. "17.56" "16.83" "17.85" "17.18"
> 1st Qu. "29.48" "29.30" "30.40" "29.34"
> Median "35.03" "34.94" "38.16" "35.22"
> Mean "34.25" "33.99" "34.87" "34.15"
> 3rd Qu. "40.00" "40.00" "40.00" "40.00"
> Max. "40.00" "40.00" "40.00" "40.00"
>
>
> But if I try and assess the replicates thus:
>
> > plotCtReps(raw, card=1, percent=30)
>
> I get
>
> Error in plot.window(...) : need finite 'xlim' values
> In addition: Warning messages:
> 1: In min(x, na.rm = na.rm) :
> no non-missing arguments to min; returning Inf
> 2: In max(x, na.rm = na.rm) :
> no non-missing arguments to max; returning -Inf
> 3: In min(x) : no non-missing arguments to min; returning Inf
> 4: In max(x) : no non-missing arguments to max; returning -Inf
> 5: In min(x) : no non-missing arguments to min; returning Inf
> 6: In max(x) : no non-missing arguments to max; returning -Inf
>
> Please can you advise what I am doing wrong?
you're not doing anything wrong, HTqPCR just isn't geared towards your
data. plotCtReps() was really just designed for qPCR cards with 2
replicates of each spot, in which case a 2D scatterplot is produced. At
the moment there's no support for 3D scatterplots (I've never had data
with 3 replicates myself). My guess is that the function gets confused
because it suddenly encounters 3 replicates of each spot instead of 2, and
hence doesn't know which pair to plot.
For cards with >2 replicates per feature I guess I should come up with
some other way of visualising the concordance between these. Maybe CV or
standard deviation versus the median Ct value, to assess both the
variation within replicated features and whether these depend on the Ct
"intensity"? Any thoughts?
Best wishes
\Heidi
>
> > sessionInfo()
> R version 2.11.1 (2010-05-31)
> x86_64-unknown-linux-gnu
>
> locale:
> [1] LC_CTYPE=en_GB.ISO-8859-1 LC_NUMERIC=C
> [3] LC_TIME=en_GB.ISO-8859-1 LC_COLLATE=en_GB.ISO-8859-1
> [5] LC_MONETARY=C LC_MESSAGES=en_GB.ISO-8859-1
> [7] LC_PAPER=en_GB.ISO-8859-1 LC_NAME=C
> [9] LC_ADDRESS=C LC_TELEPHONE=C
> [11] LC_MEASUREMENT=en_GB.ISO-8859-1 LC_IDENTIFICATION=C
>
> attached base packages:
> [1] stats graphics grDevices utils datasets methods base
>
> other attached packages:
> [1] HTqPCR_1.2.0 limma_3.4.3 RColorBrewer_1.0-2 Biobase_2.8.0
>
> loaded via a namespace (and not attached):
> [1] affy_1.26.1 affyio_1.16.0 gdata_2.8.0
> [4] gplots_2.7.4 gtools_2.6.2 preprocessCore_1.10.0
> [7] tools_2.11.1
>
>
> Kind regards and thanks,
>
> Steve
>
> P.S. I am not sure if you would rather this go to the BioC group or not so
> apologies if so and feel free to forward to
> that list.
>
> ============================================
> Head of Computational Biology Research Group
> Weatherall Institute of Molecular Medicine
> University of Oxford
> John Radcliffe Hospital
> Headington
> Oxford OX3 9DS
> +44 1865 222640
>
More information about the Bioconductor
mailing list