Can results from Mustache loop calling be used in enrichment

[Naveen-Ahuja]

Hi,

I was wondering if there is a way to use precomputed loops (bedpe file) from mustache in the enrichment analysis?

Thank you

[ay-lab]

Are you asking the question with regard to the usage of dcHiC and combining the result with Mustache?

Otherwise, you should raise the issue to the Mustache thread.

But, in general the loops can be used for enrichment analysis. You just need to find the genes overlapping (if you're doing a geneset enrichment) with the loop anchors using bedtools pairtobed option.

[Naveen-Ahuja]

Yes in regard to the usage of dcHiC and combining the result with Mustache.

For instance I want to run Rscript dchicf.r --file samples_p7.txt --pcatype enrich --genome hg38 --diffdir ctrl_vs_hgps_p7_50Kb --exclA T --region both --pcgroup pcQnm --interaction intra --pcscore F --compare F but with mustache loops and the differential compartments. How would I go about doing this?

[ay-lab]

So, dcHiC by default expects a significant file generated from fithic. It creates a FithicResult.txt file (should be under DifferentialResult/<prefix_resolution>/fithic_run/ directory). This file looks like the following -

ES      NPC
chr10_100000000_chr10_100000000 1       1
chr10_100000000_chr10_100100000 1       1
chr10_100000000_chr10_100400000 1       1
chr10_100000000_chr10_102200000 0       1
chr10_100000000_chr10_102300000 1       1
chr10_100000000_chr10_102400000 0       1

Here columns are sample names (should be same as given in your samples_p7.txt file) and rows are the interacting regions (starting coordinate). Each element represents the interaction category (1 = significant, 0 = not significant) within each sample.

So, you can replace this file with Mustache (Keep the name FithicResult.txt) and run dcHiC. It will do the job.

[Naveen-Ahuja]

Sorry I'm a little confused about the format, so if I have my 2 mustache loop files are as such

ctrl:

> chr1	1030000	1035000	chr1	1305000	1310000
> chr1	1120000	1125000	chr1	1240000	1245000
> chr1	1120000	1125000	chr1	1290000	1295000

hgps:

> chr1	985000	990000	chr1	1290000	1295000
> chr1	1120000	1125000	chr1	1285000	1290000
> chr1	1125000	1130000	chr1	1215000	1220000
> chr1	1745000	1750000	chr1	1900000	1905000

I should attach the starting coordinates of each file? and then the 1 and 0s are if that specific interaction occurs in the specific sample?

I also get this error when running (dchic) naveen@Naveens-MacBook-Pro dchic % ./enrich.sh Finding gene enrichment for ctrl_p7 hgps_p7 samples [1] "ctrl_p7" "hgps_p7" [1] "hg38_50000_goldenpathData" Error in file(file, "rt") : cannot open the connection Calls: geneEnrichment -> read.table -> file In addition: Warning message: In file(file, "rt") : cannot open file 'DifferentialResult/ctrl_vs_hgps_p7_50Kb/fdr_result/differential.intra_compartmentLoops.bedpe': No such file or directory Execution halted

[ay-lab]

For the example, your file should look like the following -

ctrl hgps
chr1_985000_chr1_1290000 0 1
chr1_1030000_chr1_1305000 1 0
chr1_1120000_chr1_1240000 1 0
chr1_1120000_chr1_1285000 0 1
...

Regarding the error, you should first run the --pcatype dloop and then perform the enrichment. This step will create the differential.intra_compartmentLoops.bedpe file.

[ay-lab]

I am closing this issue. Please open another thread if the issue persists.
Thanks.