bioconductor-hicdcplus

name

bioconductor-hicdcplus

description

Systematic 3D interaction calls and differential analysis for Hi-C and HiChIP. The HiC-DC+ (Hi-C/HiChIP direct caller plus) package enables principled statistical analysis of Hi-C and HiChIP data sets – including calling significant interactions within a single experiment and performing differential analysis between conditions given replicate experiments – to facilitate global integrative studies. HiC-DC+ estimates significant interactions in a Hi-C or HiChIP experiment directly from the raw con

HiCDCPlus

Workflows

Standard Workflow

Systematic 3D interaction calls and differential analysis for Hi-C and HiChIP. The HiC-DC+ (Hi-C/HiChIP direct caller plus) package enables principled statistical analysis of Hi-C and HiChIP data sets – including calling significant interactions within a single experiment and performing differential analysis between conditions given replicate experiments – to facilitate global integrative studies. HiC-DC+ estimates significant interactions in a Hi-C or HiChIP experiment directly from the raw con

library(HiCDCPlus)
outdir <- tempdir()

# Generate genomic features (GC, mappability, effective length)
construct_features(
  output_path = paste0(outdir, "/hg19_50kb_GATC"),
  gen = "Hsapiens",
  gen_ver = "hg19",
  sig = "GATC",
  bin_type = "Bins-uniform",
  binsize = 50000,
  chrs = "chr21"
)

# Initialize gi_list from the bintolen file
gi_list <- generate_bintolen_gi_list(
  bintolen_path = paste0(outdir, "/hg19_50kb_GATC_bintolen.txt.gz")
)

# Expand 1D features to 2D for modeling
gi_list <- expand_1D_features(gi_list)

Input: Genome version, restriction enzyme pattern, and bin size; Output: A gi_list object containing 2D genomic features ready for interaction calling.

When to Use

• Chromatin Interaction Calling: To call significant 3D chromatin interactions from Hi-C or HiChIP data using HiCDCPlus or HiCDCPlus_parallel.
• Differential Interaction Analysis: To perform differential interaction analysis across conditions with replicates using hicdcdiff.
• Feature Generation: To generate genomic features (GC content, mappability, effective length) using construct_features.
• TAD Calling: To find Topologically Associating Domains (TADs) using gi_list_topdom.
• Compartment Analysis: To extract A/B compartments (eigenvectors) from .hic files using extract_hic_eigenvectors.

When NOT to Use

• 1D Peak Calling: For standard 1D ChIP-seq peak calling, use MACS2 or epigraHMM.
• RNA-seq Differential Expression: For RNA-seq differential expression, use standard DESeq2 or edgeR directly.

Data Requirements

• Hi-C Formats: Hi-C data in .hic, .matrix, or .allValidPairs formats.
• Enzyme Patterns: Restriction enzyme cut site patterns (e.g., "GATC").
• Reference Genome: Reference genome (e.g., "hg19", "hg38") and chromosome names.

Key Parameters

• bin_type ("Bins-uniform"): Type of binning, either "Bins-uniform" or "Bins-RE-sites".
• binsize (50000): Size of genomic windows or number of restriction fragments to merge.
• sig ("GATC"): Restriction enzyme recognition sequence.
• ssize (0.1): Downsampling rate of rows for modeling in HiCDCPlus.
• mode ("normcounts"): Output mode for hicdc2hic (e.g., 'pvalue', 'qvalue', 'normcounts', 'zvalue', 'raw').
• fitType ("mean"): Fit type for dispersion in hicdcdiff.

Best Practices

• Feature Generation: Generate genomic features using construct_features before initializing the gi_list to ensure GC and length biases are modeled.
• Feature Expansion: Expand 1D features to 2D using expand_1D_features prior to running the negative binomial regression.
• Parallelization: Use HiCDCPlus_parallel for efficient genome-wide interaction calling across multiple chromosomes.
• TAD Normalization: For TAD calling, perform ICE normalization first using hic2icenorm_gi_list before running gi_list_topdom.

Common Pitfalls

• Missing Feature Expansion: Forgetting to expand 1D features to 2D before running HiCDCPlus; always call expand_1D_features first.
• Chromosome Name Mismatch: Mismatched chromosome names between the feature files and count files; ensure consistent naming (e.g., "chr21" vs "21").
• Insufficient Replicates: Insufficient replicates for differential analysis; use hicdcdiff with biological replicates to ensure robust statistical power.

Alternatives

• diffHic: For differential interaction analysis using sliding windows.
• HiTC: For basic Hi-C data manipulation and ICE normalization.
• FitHiC: For calling significant interactions.

Citations

• Sahin, M., Wong, W., Zhan, Y., Van Deyze, K., Koche, R., and Leslie, C. S. (2021) HiC-DC+: systematic 3D interaction calls and differential analysis for Hi-C and HiChIP. Nature Communications, 12(3366).

References

• Homepage: bioconductor.org/packages/HiCDCPlus
• Vignette: https://bioconductor.org/packages/release/bioc/vignettes/HiCDCPlus/inst/doc/HiCDCPlus.html