Inputs

Baleen compares two conditions — a native sample and an IVT (in vitro transcribed, unmodified) control. For each condition you supply three files, plus a single shared reference FASTA.

Required files per condition

File	Flag (native / IVT)	Requirements
BAM	--native-bam / --ivt-bam	Aligned to the reference, coordinate-sorted and indexed (.bai).
FASTQ	--native-fastq / --ivt-fastq	Basecalled reads, bgzip-compressed (.fq.gz).
BLOW5	--native-blow5 / --ivt-blow5	Raw signal in BLOW5/SLOW5 format, indexed (.blow5.idx).

Shared reference

File	Flag	Requirements
Reference FASTA	--ref	The transcriptome the BAM was aligned to, indexed with samtools faidx (.fai).

Indexing commands

# BAM: sort then index
samtools sort -o native.bam native.unsorted.bam
samtools index native.bam

# Reference FASTA
samtools faidx ref.fa

# BLOW5 signal index (slow5tools) — produces nanopore.blow5.idx
slow5tools index native.blow5

No event-alignment index step

The krill engine reads the BLOW5 signal directly via pyslow5, so it only needs the slow5tools index above. There is no separate FASTQ read-index step — FASTQ read IDs are parsed straight from the FASTQ headers.

Read-ID intersection

eventalign silently drops any BAM read whose UUID is absent from the BLOW5 signal file. If your BAM contains reads that have no corresponding raw signal (a common result of separate basecalling/alignment and signal-export steps), those reads survive BAM parsing but vanish during event alignment.

Without correction this biases everything computed before eventalign runs:

• depth statistics count reads that will never produce signal,
• --min-depth filtering keeps or drops contigs against the wrong count,
• subsampling (--subsample-n) selects from a read set larger than the one that actually yields events.

To prevent this, Baleen computes, independently for each condition, the intersection:

reads(BAM) ∩ reads(FASTQ) ∩ reads(BLOW5)

Every downstream stage — contig statistics, the --min-depth filter, subsampling, and the per-contig BAM split — is gated on this intersection, so the read set Baleen reasons about is exactly the one eventalign will align.

How read IDs are enumerated

Source	Method
BAM	Iterate alignments, collect query_name.
FASTQ	First whitespace-delimited token of each @ header. A leftover f5c <fastq>.index.readdb is ignored — krill never creates one, and the f5c single-BLOW5 form (*<TAB>blow5_path) carries no read ids.
BLOW5	pyslow5.Open(path).get_read_ids().

The intersection runs by default. Disable it with --no-read-intersection if you have already guaranteed that all three files share an identical read set (for example, on simulated data).

Empty intersection

If the three-way intersection for a condition is empty, Baleen emits a warning — this almost always means a mismatched file set (wrong BLOW5, or a BAM aligned from a different basecall) rather than a real absence of shared reads.

Next steps

Run the pipeline with these inputs — see the Quick Start and the CLI Reference. For what comes out the other end, see Outputs.