Choosing PAF Or MUMmer Coords

Use this guide when you are deciding which whole-genome alignment format should be the primary evidence for a ChromoSort run.

For most new ChromoSort runs, minimap2 PAF is the recommended starting point. It is fast, compact, supports MAPQ filtering, and works across chromo sort, chromo fix, chromo manual, and chromo plot. MUMmer coords remains a good choice when a project already has a tuned nucmer workflow, when you want a second aligner view, or when minimap2 gives a surprising result.

The Core Idea

ChromoSort normalizes MUMmer coords and minimap2 PAF rows into the same internal alignment model before sorting, plotting, and fixing. The remaining differences usually come from the aligners, not from separate ChromoSort decision logic.

Expect differences in:

row fragmentation,
chaining behavior,
primary versus secondary alignment handling,
MAPQ availability,
identity fields,
repeat and paralog placement.

Those differences are useful evidence. They should not be treated as automatic proof that one file is wrong.

What ChromoSort Reads

Format	ChromoSort uses	Most useful when
MUMmer coords	Reference/query names, reference/query coordinates, lengths, row lengths, percent identity, and orientation.	You already have filtered nucmer outputs, want a second aligner perspective, or are comparing against older MUMmer-based workflows.
minimap2 PAF	Query name/length/start/end, target name/length/start/end, strand, matching bases, block length, and MAPQ.	You want a fast default input with MAPQ filtering and compact whole-genome alignments.

Provide exactly one of --coords or --paf to commands that accept alignment evidence.

Starting Recommendation

For same-species or close-reference production runs, start with:

minimap2 \
  -x asm5 \
  -c \
  -t 16 \
  --secondary=no \
  reference.fa \
  assembly.fa \
  > paf/sample.ref_vs_asm.paf

Then use that PAF anywhere you would otherwise use MUMmer coords:

chromo sort \
  --ref-fasta reference.fa \
  --assembly-fasta assembly.fa \
  --paf paf/sample.ref_vs_asm.paf \
  --output-prefix results/sample

The -c flag matters. It asks minimap2 to perform base-level alignment and write CIGAR-bearing PAF rows. ChromoSort does not parse the CIGAR string directly, but the base-level alignment changes the PAF match and block-length columns used for identity summaries and identity filtering.

Preset Gallery

Choose the strictest minimap2 assembly preset that recovers the expected chromosome-scale alignments.

Preset	Consider it when	Watch for
`asm5`	Same species, same breeding pool, cultivar or line comparisons, or a new assembly against a close reference.	If expected contigs are missing or fragmented, the preset may be too stringent.
`asm10`	More divergent same-species material, wild or exotic accessions, pangenome references, or close relatives where `asm5` misses syntenic blocks.	More repeat, paralog, or ambiguous signal may appear; inspect plots and best-reference shares.
`asm20`	Related species or highly divergent reference choices where `asm10` still misses expected chromosome-scale alignments.	This is the noisiest choice for ChromoSort and may require stricter MAPQ or segment filters.

The presets are starting points, not hard biological cutoffs. If a preset changes the conclusion for a marginal contig, review the plot and reports rather than choosing by habit.

When MUMmer Coords Is The Better First Choice

Use MUMmer coords first when:

the project already has a trusted nucmer plus delta-filter workflow,
the reference and assembly are close enough that your MUMmer settings recover the expected chromosome-scale alignments cleanly,
you need continuity with previous reports or published workflows,
PAF results look surprising and you want a second aligner perspective.

A typical coords export is:

nucmer \
  -t 16 \
  -c 500 \
  -p mummer/sample \
  reference.fa \
  assembly.fa

delta-filter \
  -i 95 \
  -l 10000 \
  -1 \
  mummer/sample.delta \
  > mummer/sample.filter

show-coords \
  -r \
  -c \
  -l \
  mummer/sample.filter \
  > mummer/sample.coords

Interpreting Disagreements

Small PAF-vs-coords disagreements are expected. In soybean chromo fix testing, split counts differed by about 5-10%, while marginal split-contig sets differed by about 20-30% depending on mode. Use those numbers as practical expectations, not universal guarantees.

When the two formats disagree, ask:

Did both files align the same exact FASTA pair?
Was minimap2 run with -c --secondary=no?
Were secondary PAF rows included or skipped?
Are low-MAPQ PAF rows driving the call?
Did MUMmer filtering remove short or lower-identity blocks that PAF retained?
Are the discordant blocks long, coherent, and visible on a per-reference plot?
Does graph, GAF, or long-read PAF evidence support either interpretation?

Use chromo eval and chromo manual when the decision affects sequence.

Filter Cheat Sheet

Signal	Useful filter or setting
Too many small repeat-like PAF rows	Increase `--min-segment-bp` or keep `--secondary=no`.
Low-confidence PAF placements	Increase `--min-mapq`.
PAF identity looks unexpectedly low	Confirm minimap2 was run with `-c`.
Expected syntenic blocks are missing	Try `asm10` after `asm5`, then inspect plots before using `asm20`.
MUMmer output is very fragmented	Review `delta-filter -i`, `-l`, and `-1` settings.
Coords and PAF disagree on a marginal split	Use dot plots, `chromo eval`, graph evidence, and long-read evidence before editing sequence.

Common Traps

Do not run both formats and then silently mix decisions. Choose one primary alignment for a workflow stage, then use the other as review evidence when needed.

Do not assume asm20 is better because it is more permissive. More alignments can mean more repeat noise and more ambiguous placements.

Do not include secondary PAF rows by default. ChromoSort skips rows marked tp:A:S unless the relevant --include-secondary-* option is set, and --secondary=no keeps the file cleaner from the start.

Do not compare old coords against new PAF from a different FASTA stage. Format choice and FASTA-stage choice are separate questions.

Do not set identity filters before looking at the identity distribution from your chosen aligner and preset.

What To Look At Next In ChromoSort

Use Alignment Evidence And The Exact FASTA Rule before reusing any alignment file.
Use Input Files for the current minimap2 recipe.
Use Troubleshooting when the two formats disagree.
Use How to Interpret Dot Plots to decide whether a disagreement is biologically meaningful.