1.1.0 - Venusian Pyrite

[1.1.0] - Venusian Pyrite - 2026-09-26

Added

• Support for early/late Repli-seq: per-sample log2(early/late) replication-timing tracks, optional replication-timing index, RT domain segmentation and gene classification (test_repliseq profile).
• Support for high-resolution (16-fraction) Repli-seq, following Zhao, Sasaki & Gilbert (2020): initiation zones, timing transition regions, breakages, termination sites and constant-timing regions (test_hr_repliseq profile).
• Support for the eSPAN experiment type (#28), with a test_espan profile.
• peak_callers, to run several peak callers in one run (e.g. --peak_callers macs3,genrich), and skip_peak_calling.
• minibwa as an additional aligner, and experiment-type-specific alignment settings for CUT&RUN, CUT&Tag, TIP-seq and eSPAN.
• MultiQC sections for per-stage samtools stats, Picard, phantompeakqualtools, MACS3, SEACR, UMI deduplication and deepTools QC.
• FRiP scores, HOMER annotation and QC for SEACR peaks, as for MACS3.
• Stubs for all local modules, enabling end-to-end dry-runs with -stub (#61).
• Run provenance / RO-Crate generation with nf-prov.
• A pipeline website: crepas.grothlab.workers.dev.

Changed

• Updated to the nf-core tools 4.1.0 template, nf-schema 2.7.3 and nf-prov 1.7.0, and updated the nf-core modules.
• Local modules use purpose-built Wave / BioContainers images pulled over https://, with conda environments pinned to the container versions.
• Version reporting moved to the versions topic channel.
• Renamed min_reps_consensus to consensus_min_replicates_per_sample, and the samtools stats summaries to *.all.tsv and *.totals.tsv.
• Renamed the partition plot outputs (*_plot_raw → *_plot_smoothed, *_plot_smoothed → *_plot_smoothed.gam).
• Consensus peak featureCounts and DESeq2 QC now run separately for single- and paired-end libraries (.se/.pe suffix).
• --gtf is no longer required; --gff alone is enough.
• The grothlab institutional profiles moved to grothlab/configs.

Fixed

• Parameter validation guards that never triggered; enabling blacklist filtering without a blacklist silently emptied every downstream channel.
• CISRPM normalization and division-by-zero crashes when a sample had no spike-in reads.
• Duplicated --maxins in the bowtie2 arguments, which overrode the CUT&RUN, CUT&Tag and TIP-seq insert-size limit.
• Sample names containing dots (e.g. H3.3) being truncated in plots and the samtools stats summary.
• Peak QC, consensus featureCounts and DESeq2 QC results never reaching MultiQC, or not running at all.
• HISAT2 alignment failing under Singularity and Apptainer.
• Chromap segmentation faults and memory issues.
• denopa is marked as unsupported for now; the pipeline stops early if it is requested.

Full Changelog: 1.0.0...1.1.0