Skip to content

Data formats

AGP

C-Studio expects tab-delimited AGP with nine columns. Component rows provide object coordinates, component type, component ID, source coordinates, and orientation. N and U rows provide gap length, type, linkage, and evidence.

  • AGP coordinates are 1-based closed.
  • Internal C-Studio intervals are 0-based half-open.
  • Component IDs are the source lookup keys for other evidence.
  • +, -, ?, 0, and na orientations are retained as legal input/export values.
  • Gap metadata is retained while the gap exists. Delete gap / join blocks explicitly removes it.

On export, object coordinates and part numbers are regenerated from the current placement order. Source component coordinates remain tied to the edited source intervals.

Operation history sidecar

Saving an edited AGP also writes <AGP-prefix>.history.json. The sidecar stores the applied and undone operation timeline plus the canonical AGP used to validate compatibility. C-Studio restores it only beside the matching AGP; malformed, stale, or mismatched sidecars are ignored.

The sidecar supplements the AGP and does not replace it. AGP remains the authoritative assembly layout and the file used by other tools.

COOL and MCOOL

.cool and .mcool are Cooler-compatible HDF5 contact containers. C-Studio reads chromosome/bin identifiers, sparse pixels, stored resolutions, and available weights needed by the selected normalization.

  • .cool contains one matrix resolution.
  • .mcool can contain a resolution pyramid and is preferred for large data.
  • These files must not be gzip wrapped.
  • Matrix identifiers must match AGP component IDs.

PAF

C-Studio reads standard PAF's first 12 fields. Query name, query length and interval, strand, target name and interval, residue matches, alignment block length, and mapping quality are used to construct the synteny view. Query names must match AGP component IDs.

Coverage

Coverage input is whitespace-delimited and bedGraph-like:

chrom  start  end  value

Intervals are 0-based half-open. start must be non-negative, end must be greater than start, and value must be numeric. Lines beginning with #, track, or browser are ignored.

GFA

The current parser uses these GFA records:

  • S: segment name, sequence or LN length, and optional rd depth tag;
  • L: oriented segment link and overlap CIGAR/string;
  • A: per-segment count and optional HG haplotype tag.

A records and an rd tag are segment metadata; they are not promoted to GFA edge support. Only explicit valid L records create topology links. Links to missing S segments, invalid rows, and duplicate links are warned about or filtered.

Gzip text input

AGP, GFA, PAF, depth, and bedGraph text may use a final .gz suffix. C-Studio uses multi-member gzip decoding. Auto-save never overwrites compressed AGP; save to a plain .agp target instead.