ay-lab
commented
6 months ago Hi,
Can you try to run this step first? https://github.com/ay-lab/dcHiC/tree/master/utility#using-existing-pc-values-with-dchic
Using getcHiCinputfromExistingPCs.r, you can reformat their existing pc values such that they can use dcHiC. The input must be a five-column text file with the following fields:
<pc bedGraph path> <pc type> <replicate name> <sample name>In your case, it should look like the following
/path/to/T1_R1_PC1.bedGraph intra T1_R1_PC1 T1_R1_PC1
/path/to/T1_R2_PC1.bedGraph intra T1_R2_PC1 T1_R2_PC1
....Once you run this code, you can run the dcHiC --pcatype analyze option on the output results.
For that, you will need an input.txt file that will look similar to the following -
T1_R1_PC1_<HiC_Resolution>_matrix.txt <HiC_Resolution>.bed T1_R1_PC1 T1_PC1
T1_R2_PC1_<HiC_Resolution>_matrix.txt <HiC_Resolution>.bed T1_R2_PC1 T1_PC1
T2_R1_PC1_<HiC_Resolution>_matrix.txt <HiC_Resolution>.bed T2_R1_PC1 T2_PC1
T2_R2_PC1_<HiC_Resolution>_matrix.txt <HiC_Resolution>.bed T2_R2_PC1 T2_PC1
.....You don't have to create any of the _matrix.txt; it just serves as a dummy name. However, you must create the
<chrome> <start> <end> <index>Once you have these files, you can run the "analyze" step to get the differential compartments.
Let me know if you face any issues, I will gladly help you.
Hi,
I already have a PC1 data frame of five-time points (T1-T5) at 100kb resolution in a TSV file. Each time point has two biological replicates (R1, R2). The genome assembly is hg38. I generated these PC1s per arm. I would like to identify those differential compartment bins across these five time points. And I also want to use time-series analysis to cluster the compartmentalization score patterns of these differential bins by generating the plots using TCseq package. Could you send me an example code to achieve them?
Thanks a lot, Jiangyuan
chrom start end T1_R1_PC1 T1_R2_PC1 T2_R1_PC1 T2_R2_PC1 T3_R1_PC1 T3_R2_PC1 T4_R1_PC1 T4_R2_PC1 T5_R1_PC1 T5_R2_PC1 chr1 100,000 200,000