[BioC] ChIPpeakAnno - VennDiagram P-value - NaN

[Zhu, Julie]

Hi Khademul,

I just found a block of code that was added in the new version that caused
it run significantly slower. I will update the code this weekend that should
help with the speed issue. Thanks!

Kind regards,

Julie


On 8/9/10 10:30 AM, "Julie Zhu" <julie.zhu at umassmed.edu> wrote:

> Hi Khademul,
> 
> Regarding the speed concern with very big bed files, splitting the files
> chromosome by chromosome would help if you can run R in a cluster.
> 
> Regarding p-value for overlap, set totalTest appropriately is critical. Please
> refer to the posts.
> 
> http://permalink.gmane.org/gmane.science.biology.informatics.conductor/29476
> http://permalink.gmane.org/gmane.science.biology.informatics.conductor/30115
> 
> Kind regards,
> 
> Julie
> 
> 
> On 8/9/10 6:02 AM, "Khademul Islam" <khademul.islam at gmail.com> wrote:
> 
> Hi Julie,
> 
> Nice to see published ChIPpeakAnno paper.
> 
> I was trying to do overlapping between my ChIPseq peak bed file (peak1: ~3100
> peaks) and Exon_Intron Boundary bed file ( peak2: ~400000 ).
> 
> There are two concerns:
> 
> 1. Its takes too long time (overnight) to calculate all these even with
> powerful machine........., specially when one bed file is too big.
> 
> 2. It produced "NaN" for p-value
> 
> 
> $p.value
> [1] NaN
> 
> $vennCounts
>      peak1 peak2  Counts
> [1,]     0     0 -398821
> [2,]     0     1  395828
> [3,]     1     0    2253
> [4,]     1     1     840
> attr(,"class")
> [1] "VennCounts"
> 
> 
> So it counts "-" (minus) -398821 ???
> 
> There were 22 warnings but no other error. Warning says that it produces "NaN"
> 
> I pasted warnings below. I wanted to attach the bed files if you need to
> check, but its too large. Well, in case if you need it, let me know, I will
> upload it in FTP site. I got the venn diagram pic without any problem
> (attached).
> 
> Command line was:
> 
> 
> ol <- findOverlappingPeaks(peaks1, peaks2, maxgap=1, multiple=F,
> NameOfPeaks1="Peak1", NameOfPeaks2="Peak2")
> 
> vdg <- makeVennDiagram(RangedDataList(peaks1, peaks2), NameOfPeaks=c("peak1",
> "peak2"), maxgap=0, totalTest=100, cex = 1, counts.col = "red")
> 
>> dev.copy2eps()
> 
>> vdg
> 
> 
> So, my question is, how can I get proper p-value from the overlap, what I have
> to do to fix it?
> 
> Thank you,
> 
> Khademul
> 
> 
> 
>> sessionInfo()
> R version 2.11.1 (2010-05-31)
> x86_64-unknown-linux-gnu
> 
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods   base
> 
> other attached packages:
>  [1] ChIPpeakAnno_1.4.1
> .....
> 
>> warnings()
> Warning messages:
> 1: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 2: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 3: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 4: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 5: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 6: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 7: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 8: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 9: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 10: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 11: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 12: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 13: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 14: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 15: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 16: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 17: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 18: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 19: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 20: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 21: In .local(query, subject, maxgap, minoverlap, type, select,  ... :
>   argument 'multiple' is deprecated; use 'select'.
> 22: In phyper(q, m, n, k, lower.tail, log.p) : NaNs produced
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> [[alternative HTML version deleted]]
> 
> _______________________________________________
> 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
>