3D/HiCDCPlus.R

# if (!requireNamespace("BiocManager", quietly = TRUE))
#   install.packages("BiocManager")
# 
# BiocManager::install("HiCDCPlus")
# BiocManager::install("BSgenome.Hsapiens.UCSC.hg19")


library(HiCDCPlus)
library(BSgenome.Hsapiens.UCSC.hg19)

# outdir <- "data/3D/K562_HiC/"
# hicfile_path <- paste0(outdir, "GSE63525_K562_combined_30.hic")
outdir <- "data/3D/GM12878_HiC/"
hicfile_path <- paste0(outdir, "GSE63525_GM12878_insitu_primary_30.hic")


process_chromosome <- function(chr) {
  cat("Processing chromosome:", chr, "\n")
  
  feature_path <- paste0(outdir, "hg19_5kb_GATC_", chr)
  
  construct_features(output_path=feature_path,
                     gen="Hsapiens",gen_ver="hg19",
                     sig="GATC",
                     binsize=5000,
                     chrs=c(chr),
                     )
  # generated file under hg19_5kb_GATC_bintolen.txt.gz
  
  # generate gi_list instance
  feature_output <- paste0(feature_path, "_bintolen.txt.gz")
  
  gi_list<-generate_bintolen_gi_list(bintolen_path=feature_output)
  # add .hic counts
  gi_list<-add_hic_counts(gi_list,hic_path = hicfile_path)
  # expand features for modeling
  gi_list<-expand_1D_features(gi_list)
  
  # run HiC-DC+
  set.seed(1010) #HiC-DC downsamples rows for modeling
  gi_list<-HiCDCPlus_parallel(gi_list,ncore=2)
  
  cat("gi list length:", length(gi_list), "\n")
  
  
  # select FDR < 0.1
  alpha <- 0.1
  
  significant_interactions <- list()
  
  # 遍历 gi_list 中的每个 GInteractions 对象
  gi_list_names <- names(gi_list)
  for (i in seq_along(gi_list)) {
    name <- gi_list_names[i]
    # 提取当前 GInteractions 对象
    gi <- gi_list[[i]]
    metadata_columns <- mcols(gi)
    qvalues <- metadata_columns$qvalue
    significant_interaction <- gi[qvalues <= alpha, ]
    
    significant_interactions[[i]] <- significant_interaction
    names(significant_interactions) <- name
  }
  
  
  # save file
  # save_name = paste0(outdir,'GSE63525_K562_HiC30_FDR_1_', chr)
  save_name = paste0(outdir,'GSE63525_GM12878_HiC30_FDR_1_', chr)
  hicdc2hic(significant_interactions,
            hicfile=paste0(save_name, '.hic'),
            mode='normcounts',
            gen_ver='hg19'
            )
  gi_list_write(significant_interactions,fname=paste0(save_name, '.txt.gz'))
  
}


chromosomes <- c("chr1", "chr2", "chr3", "chr4", "chr5", "chr6", "chr7", "chr8", "chr9", "chr10",
                 "chr11", "chr12", "chr13", "chr14", "chr15", "chr16", "chr17", "chr18", "chr19", "chr20",
                 "chr21", "chr22", "chrX")

# Process each chromosome
for (chr in chromosomes) {
  process_chromosome(chr)
}



# # test
# library(InteractionSet)
# 
# # 提取第一个 GInteractions 对象
# gi <- gi_list[[1]]
# 
# # 提取元数据列
# metadata_columns <- mcols(gi)
# 
# # 查看 qvalue 和 pvalue 列
# qvalues <- metadata_columns$qvalue
# pvalues <- metadata_columns$pvalue
# 
# # 检查 qvalue 和 pvalue 列的唯一值
# unique_qvalues <- unique(qvalues)
# unique_pvalues <- unique(pvalues)
# 
# num_qvalues_less_than_0.1 <- sum(qvalues < 0.1, na.rm = TRUE)