Skip to contents

Bin a BAM/CRAM file — NIPTeR style

nipter_bin_bam.Rd

Replicates NIPTeR's bin_bam_sample(): counts reads in fixed-width bins across chromosomes 1-22, X, and Y. The original NIPTeR counts all mapped reads with no MAPQ filter; real-world NIPT pipelines typically pre-filter with samtools view --min-MQ 40 -F 1024 before binning. Both modes are supported through the mapq, exclude_flags, and require_flags parameters.

Usage

nipter_bin_bam(
  bam,
  binsize = 50000L,
  mapq = 0L,
  require_flags = 0L,
  exclude_flags = 0L,
  rmdup = c("none", "flag"),
  separate_strands = FALSE,
  con = NULL,
  reference = NULL
)

Arguments

bam

Path to an indexed BAM or CRAM file.

binsize

Bin size in base pairs. Default 50000 (NIPTeR's fixed bin size). Must match the binsize used when building any control group.

mapq

Minimum mapping quality. Default 0L (NIPTeR original: all mapped reads). Set to 40L to match common NIPT pipeline pre-filtering (samtools view --min-MQ 40).

require_flags

Integer bitmask; only reads with all bits set are kept (samtools -f). Default 0L (no requirement).

exclude_flags

Integer bitmask; reads with any bit set are dropped (samtools -F). Default 0L. Set to 1024L (0x400) to exclude reads marked as duplicates (samtools view -F 1024).

rmdup

Duplicate removal strategy. "none" (default, NIPTeR standard) or "flag" (equivalent to exclude_flags = 1024L; for BAMs already processed by Picard / sambamba). NIPTeR does not perform streaming deduplication.

separate_strands

Logical; when TRUE, produces a SeparatedStrands object with independent forward/reverse count matrices. Default FALSE (CombinedStrands).

con

Optional open DBI connection with duckhts already loaded.

reference

Optional FASTA reference path for CRAM inputs.

Value

An object of class c("NIPTeRSample", <strand_type>):

CombinedStrands (default): autosomal_chromosome_reads is a list of one 22-row integer matrix (rows "1"–"22"); sex_chromosome_reads is a list of one 2-row matrix (rows "X", "Y").

SeparatedStrands (separate_strands = TRUE): autosomal_chromosome_reads is a list of two matrices — element 1 is forward (rows "1F"–"22F"), element 2 is reverse (rows "1R"–"22R"); sex_chromosome_reads similarly contains forward ("XF", "YF") and reverse ("XR", "YR") matrices.

Details

The result is reshaped into a NIPTeRSample object whose structure parallels NIPTeR's NIPTSample: autosomal reads as a chromosome-by-bin matrix and sex chromosome reads as a separate two-row matrix.

When separate_strands = TRUE, forward (+) and reverse (-) reads are counted independently, producing two matrices per chromosome set (class "SeparatedStrands"). This doubles the predictor pool for nipter_regression() — see NIPTeR documentation for details.

See also

bam_convert(), nipter_bin_bam_bed(), nipter_gc_correct(), nipter_regression()

Examples

if (FALSE) { # \dontrun{
# NIPTeR original defaults: all mapped reads, no dedup
sample <- nipter_bin_bam("sample.bam", binsize = 50000L)

# Common NIPT pipeline: MAPQ >= 40, exclude duplicate-flagged reads
sample <- nipter_bin_bam("sample.dm.bam", binsize = 50000L,
                         mapq = 40L, exclude_flags = 1024L)

# SeparatedStrands for regression with doubled predictor pool
sample_ss <- nipter_bin_bam("sample.bam", separate_strands = TRUE)

sample$autosomal_chromosome_reads[[1]]["21", ]   # chr21 bin counts
} # }