v1.0.0 - Crimson Goregutter

v1.0.0 - Crimson Goregutter - [2026-08-17]

This is the first stable release of sanger-tol/genomeassembly. This pipeline is now equivalent with the workflows currently used in the Tree of Life Assembly process, and should produce an entirely equivalent assembly given the same input data. Note that at time of release this pipeline is not yet used in production in Tree of Life.

The input schema has been re-designed to allow for a flexible but repeatable specification of a genome assembly from raw data, and the pipeline is tooled to reduce redundant computation where differently specified assemblies share steps, allowing for comparison of multiple genome assemblies where desired.

Future releases will add explicit support for polyploid genome assembly, including both assembly with >2 haplotypes, and polyploid aware phased scaffolding, as well as homology-based scaffolding of haplotypes against the primary haplotype with RagTag. These features will be added as the processes are developed within the Tree of Life programme.

The full changelog is as follows:

Added

• #109 Split the pipeline inputs into two separate samplesheets (by @prototaxites).
  • --genomic_data describes the specific datasets available to produce an assembly, including the sample name, sequencing platform, path to the sequencing data, and optional paths to FastK databases built from the reads.
  • --assembly_specs describes the assemblies to build as a function of the provided datasets, including parameters for assembly at different stages, and the choice of genome assembler (currently hifiasm for nuclear genome assembly, or oatk and mitohifi for organellar genome assembly) .
  • For nuclear assemblies, each assembly is split into stages (overlap graph construction, raw assembly, purging, polishing, and scaffolding). The inputs for each stage (including raw data, previous assembly, and params) are hashed and deduplicated, so that when assemblies share the same stage the computation is only executed once. Thus, two identical assembly specifications in the input sheet would only result in one set of genome assemblies being processed, but published twice, and two assemblies that differ only in scaffolding parameters would result in a single computation running until branching into two at the scaffolding stage.
  • For organellar assemblies, available assemblers are oatk (mitochondrial and plastid genomes) and mitohifi (mitochondrial genomes) and use the same staging logic, although they presently only have a single stage. These are also de-duplicated prior to running.
  • #124 Update the documentation to reflect the new input types.
• #119 Add index fixes to the outputs for each assembly, an overall index describing the data, params and tool versions used, and one for each stage within an assembly, describing the same elements but only those used thus far (by @prototaxites)
• #132 Add comprehensive local testing with nf-test with new test datasets. Three new test datasets are available (by @prototaxites)
  • icAdaBipu1, the two-spotted ladybird, for testing pri/alt assembly and phased assembly
  • aBomVar4/1/2, the yellow-bellied toad, for testing trio assembly
  • dhQueRobu3, the pedunculate oak, for testing ONT assembly as well as Pacbio+UL ONT assembly, and plastid assembly
• #139 Add a pipeline logo and metro map! (by @prototaxites)
• #140 Use the new secrets functionality of the nf-core/setupnextflow GitHub action in CI (by @prototaxites)
• #141 Allow specification of user-provided reference files for MitoHiFi (by @prototaxites)
  • adds the following specification YAML entries: mitohifi_mito_reference_fa, mitohifi_mito_reference_gb, mitohifi_plastid_reference_fa, mitohifi_plastid_reference_gb
• #141 Fix issue with mitohifi_arguments not being correctly applied (by @prototaxites)
• #160 Add --save_genomescope2_outputs parameter flag to save the model outputs from GenomeScope2. It is disabled by default (by @prototaxites)
• #164 Add the parameters --build_pretext_map, --build_cooler_map, --build_juicer_map to control which contact maps are built (by @prototaxites)

Fixed

• #110 Fixed issue with minimap2 mapping settings not being correctly applied at the index building stage (reported by @muffato, fixed by @prototaxites)
• #115 Pipeline now conforms to Nextflow strict syntax (by @prototaxites)
• #130 Fixed issue with BUSCO where bbtools sometimes ran out of memory, causing BUSCO failure (by @prototaxites)
• #132 Fix issues with trio assembly causing pipeline failures due to lack of testing (by @prototaxites)
• #132 Fix issue where MITOHIFI_FINDMITOREFERENCE didn't pull down chloroplast references (by @prototaxites)
• #132 --busco_lineage_directory correctly converted to a path (by @prototaxites)
• #147 Correctly apply params.minimap2_ont_map_mode parameter (reported by @amytims, fixed by @prototaxites)
• #162 Update nf-core template to 4.1.0 (by @prototaxites).
  • Note that this update removes the existing Teams and Slack notification functionality. If you were using this functionality, please configure the nf-slack or nf-teams Nextflow plugins.
• #159 - Add --opt-out-run-stats to BUSCO call when running with a local BUSCO mirror to ensure complete offline capacity (bug reported by @cjfields, fix by @prototaxites)
• #159 - Fix issue in hashing process where the data type name was hashed instead of the dataset name, resulting in clashing hashes (reported by @amakunin and @TannerMyers, fix by @prototaxites)
• #163 - FastK databases no longer built for Hi-C datasets (by @prototaxites)

Dependencies

Module	Tool	Old version	New version
bcftools/concat	bcftools	1.21	<removed, unused>
bcftools/consensus	bcftools	1.21	1.23.1
bcftools/index	bcftools	1.21	1.23.1
bcftools/norm	bcftools	1.21	1.23.1
bcftools/sort	bcftools	1.21	1.23.1
bcftools/view	bcftools	1.21	1.23.1
bedtools/bamtobedsort	samtools	1.22.1	1.23
busco/busco	busco	6.0.0	6.1.0
fastk/fastk	fastk	1.1.0	1.2
fastk/histex	fastk	1.1.0	1.2
freebayes	freebayes	1.3.6	1.3.10
gawk	gawk	5.3.0	5.3.1
gatk4/mergevcfs	gatk4	4.6.1.0	4.6.2.0
genomescope2	genomescope2	2.0	2.1.0
gfastats	gfastats	1.3.10	1.3.11
htslib/bgziptabix	bgzip	-	1.24
merquryfk/hapmaker	merquryfk	1.1.1	1.2
merquryfk/merquryfk	merquryfk	1.1.1	1.2
pretextmap	samtools	1.17	1.23
pretextmap	pretextmap	0.1.9	0.2.3
pretextsnapshot	pretextsnapshot	0.0.4	0.0.5
fastxalign/pyfastxindex	pyfastx	2.2.0	2.3.0
fastxalign/fastxalign	pyfastx	2.2.0	2.3.0
fastxalign/fastxalign	samtools	1.22.1	1.23
samtools/faidx	samtools	1.22.1	1.24
samtools/flagstat	samtools	1.22.1	1.24
samtools/idxstats	samtools	1.22.1	1.24
samtools/index	samtools	1.22.1	1.24
samtools/merge	samtools	1.22.1	1.24
samtools/mergedup	samtools	1.22.1	1.23.1
samtools/splitheader	samtools	1.22.1	1.24
samtools/stats	samtools	1.22.1	1.24
seqkit/grep	seqkit	2.9.0	2.13.0
tabix/bgzip	bgzip	1.21.0	-
yahs/makepairsfile	samtools	1.22.1	1.23

Deprecated

A lot of parameters have been deprecated, and replaced with fields in the assembly specification samplesheet.

Parameter	Replacement
--input	--genomic_data, --assembly_specs
--enable_hic_phasing	phased_assembly in assembly spec
--enable_trio_binning	trio_assembly in assembly spec
--hifiasm_error_correction_options	hifiasm_bin_arguments in assembly spec
--hifiasm_assembly_options	hifiasm_arguments in assembly spec
--purging_assemblytypes	purge: true in assembly spec
--purging_purge_middle	purge_middle: true in assembly spec
--purging_cutoffs	purging_cutoffs in assembly spec
--enable_organelle_assembly	assembler: <oatk,mitohifi> in assembly spec
--mitohifi_reads_args	mitohifi_arguments in assembly spec
--mitohifi_contigs_args	mitohifi_arguments in assembly spec
--oatk_kmer_size	oatk_kmer_size in assembly spec
--oatk_coverage	oatk_coverage_cutoff in assembly spec
--enable_polishing	polish: true in assembly spec
--polishing_assemblytypes	polish: true in assembly spec
--enable_scaffolding	scaffold: true in assembly spec
--yahs_break_contigs	yahs_arguments in assembly spec
--yahs_resolutions	yahs_arguments in assembly spec
--yahs_min_contig_length	yahs_arguments in assembly spec