Installation and usage

Install

pip install fiberhmm            # Python 3.10 to 3.14
pip install "fiberhmm[all]"     # plus matplotlib (QC figures) and h5py (HDF5 posteriors)

UCSC bedToBigBed is needed for bigBed tracks, and samtools is recommended for speed. See Installation.

Call footprints

fiberhmm-call runs the HMM, nucleosome recall and TF recall in one pass. Pick the chemistry; the bundled model follows:

# Hia5 Fiber-seq, PacBio (sorted, indexed BAM with MM/ML)
fiberhmm-call -i aligned.bam -o calls.bam --enzyme hia5 --seq pacbio -c 8 --region-parallel

# Hia5 Fiber-seq, Nanopore (--seq is detected from the MM tags when omitted)
fiberhmm-call -i aligned.bam -o calls.bam --enzyme hia5 --seq nanopore -c 8 --region-parallel

# DAF-seq, DddB or DddA (aligned with MD tags, e.g. minimap2 --MD)
fiberhmm-call -i aligned.bam -o calls.bam --enzyme dddb -c 8 --region-parallel
fiberhmm-call -i aligned.bam -o calls.bam --enzyme ddda -c 8 --region-parallel

# Unaligned BAM or stdin: stream, optionally into FIRE (fibertools' ft)
fiberhmm-call -i reads.bam -o - --enzyme hia5 --seq pacbio -c 8 | ft fire - fire.bam

For DAF file input, PCR-duplicate marking and SNP screening run automatically. For DddA, CpGs are left out of calling except inside unmethylated CpG islands (--no-use-m5c keeps them). Every file output gets a bounded QC report in qc/ next to the BAM.

Look at the calls

fiberhmm-extract -i calls.bam -o tracks/                 # BED12 / bigBed per feature type
fiberhmm-qc -i calls.bam other.bam -o qc_compare/         # QC and a combined comparison
fiberhmm-consensus --bam calls.bam --region chr1:1000000-1000350 --output classes/

Output tags

Intervals are in the molecule's own orientation, like fibertools. TF calls are in MA/AQ only; --downstream-compat also puts them into ns/nl. See BAM tags.

Train a custom model

fiberhmm-probs -a accessible_control.bam -u inaccessible_control.bam -o probs/sample --mode pacbio-fiber -k 3
fiberhmm-train -i sample.bam -p probs/sample/tables/sample_accessible_A_k3.tsv \
    probs/sample/tables/sample_inaccessible_A_k3.tsv -o model/ -k 3
fiberhmm-call -i sample.bam -o calls.bam -m model/best-model.json --region-parallel

See Training a model and the full command-line reference.