bam-file-coordinate-sorting-verification

BAM file coordinate sorting verification

Summary

Verify that input BAM files are coordinate-sorted before preprocessing with SnapATAC2's pp.make_fragment_file, ensuring compatibility with fragment file generation and correct output of chromatin accessibility data.

When to use

Before invoking pp.make_fragment_file on a BAM file from alignment or external sources, especially when the BAM's sort order is unknown or when integrating BAM files from multiple sequencing platforms (10X, standard genomics pipelines, or custom aligners) into a unified SnapATAC2 analysis.

When NOT to use

• Input is already a fragment file (BED.gz, .zst, or TSV) — coordinate sorting verification is only needed for BAM inputs.
• Using preprocessing pipelines that auto-sort BAM files internally (e.g., precellar) — verify documentation first.
• Input BAM is from a tool that streams coordinate-sorted output by design — may skip explicit verification if tool documentation guarantees sort order.

Inputs

• BAM file (aligned single-cell ATAC-seq reads with optional cell barcodes in tags)
• SAM/BAM header or file metadata indicating current sort order

Outputs

• Confirmation of coordinate sort order (e.g., samtools view header output or sort validation report)
• Re-sorted BAM file (if input was unsorted or query-name sorted)

How to apply

Check BAM file headers using standard tools (samtools view -H) to confirm the @HD line contains SO:coordinate, or verify coordinate sort order by spot-checking that genomic positions increase monotonically across reads. If the BAM is unsorted or sorted by read name (SO:queryname), re-sort it using samtools sort -o output.bam input.bam before passing to pp.make_fragment_file. The pp.make_fragment_file function expects coordinate-sorted input and will produce compressed fragment files (BED.gz or .zst format) containing fragment coordinates, cell barcodes, and quality metrics only when the input is correctly ordered; misalignment of input sort order will compromise fragment boundary detection and barcode grouping.