# Tag Info

4

There is no such thing as a hypergeometric test, at least in statistical textbooks. It's a fisher test based on hypergeometric distribution. If it is chip-seq for the same target, i.e biological replicates, significance of overlap is not quite meaningful. You get more information by for example checking the correlation of the coverage between your ...

4

I would ignore peak calling for this and instead compute enrichment of ChIP/input for the genome (e.g., with deepTools or presumably homer) and then plot it for the genes of interest individually (e.g., using IGV or pyGenomeTracks) or as a group (e.g., with computeMatrix). If the peaks are obvious and you trust your peak calling then sure you can just use ...

3

I'd say mostly this is a question of understanding the underlying biology and the relevant literature. If it is not known in the literature whether a mark is peaky or broad, evidence might come from FISH studies or Low throughput qPCR. Another way to look would be to examine the signal expressed as fold enrichment over input on a genome browser and look if ...

3

The sure why any software should be confined to only broad peaks, but there is two packages that should do what I thing you have in mind. Just to clarify, for smallRNAs peak detection is sometimes also referred to as loci or cluster detection: ShortStack which is developed for alignment, annotation, and quantification of small RNAs. The de-novo cluster ...

2

You could try setting a high p-value threshold when you're calling peaks to retain the "non-significant" values. Something like: macs2 callpeak -t <ChIP>.bam -c <Control>.bam -f BED -g hs -n test -B -p 0.7 You probably don't want to set the threshold to 1, since you'll likely want to avoid some noise. But you can then use these "non-...

2

Automated gating methods are gaining popularity over manual and time consuming analysis in say FlowJo, you should take a look at the openCyto bioconductor package this is a framework which builds on top of existing Bioconductor infrastructure for flow cytometry (PLOS compbio paper) with the aim of making various analysis routes more accessible to the ...

2

There are quite a few algorithms developed for the automatic classification of multidimensional flow cytometry data, you can see a (not so recent?) review here: http://www.nature.com/nmeth/journal/v10/n3/full/nmeth.2365.html. In your case, you are interested in the unsupervised methods, since you do not have data from single populations (which would provide ...

2

Depends how the files were made. In the simplest case yes, the height represents the pipeup of reads from the BAM file that was used. For a direct (visual) comparison you have to normalize the files though as otherwise sequencing depth confounds the height of the peaks, like if file A has ten times more reads and the peak is 10 times higher than in B then ...

2

bedtools intersect with multiple files after -b performs pairwise intersections between -a and each file in -b. It does not intersect all files simultaneously. See here, from the Bedtools documentation. The green section only contains intervals from -a that also intersect at least one interval in any -b file. What I think you're looking for is bedtools ...

1

You just need to strip the / from $d, for which there are a number of options: for d in */ ; do dname=basename$d findPeaks $d -style factor -o${dname}.txt done or for d in */ ; do dname=${d%%/} findPeaks$d -style factor -o ${dname}.txt done There are likely other ways one could go about this. Having said that, since you're not ... 1 that was solved by changing seq level: newnames <- paste0(c("1","2","3","4", "5","6","7", "8","9","10","11","12","13","14","15","16","17","18","19","20",&... 1 You can use rtracklayer from R bioconductor to import it, filter and do much more stuff, i use the code from this blog for importing the narrow bed file. Below I use an example file since you did not provide yours: library(rtracklayer) fl = "zas1_default_peaks.narrowPeak" extraCols_narrowPeak <- c(signalValue = "numeric", pValue = &... 1 You can do it by awk. awk -v OFS="\t" '{ if ( ($3-$2) >= 500 && ($3-\$2) <= 1000) { print } }' in.bed > out.bed Assuming there is no header.

1

You can use windowBed from bedtools with -c if you just want counts. Or use -wa -wb and then use groupBy for any other operations.

1

For sharing, you can test the proportion of overlapping peaks using bedtools intersect or the find.overlap function in GenomicRanges packages in R. What makes more sense is perhaps to merge all the peaks together using bedtools merge, and then count the number of reads/fragments in each peak for each sample using featureCounts. The output of this is a matrix,...

1

Just to make sure, I took a quick look at chippeakanno, and it uses a hypergeometric test for gene enrichment, not necessarily overlap of two chip-seq results. As you mentioned, you can imagine in your set of genes, you have those that overlap with your peaks, and those that do not. Hypergeometric distribution calculates the probability of getting a certain ...

1

Venn diagrams can be helpful, but for peak calling they can be somewhat misleading. This topic is addressed in the DiffBind documentation, in the "Comparison of occupancy and affinity based analyses" section. Suffice to say, asking "which peaks are shared?" is not the same as asking "is there evidence for differential peaks, and where are they?". For this ...

1

Check out the tool Intervene. It's exactly designed for this. Github link.

Only top voted, non community-wiki answers of a minimum length are eligible