Intervals and layouts

Each interval combines a role prefix with a shape.

Interval roles

Prefix	Role	Typical use
b	barcode	Cell, round, or other identifying barcode.
s	sample barcode	Barcode used to distinguish samples.
u	UMI	Unique molecular identifier.
r	read sequence	Biological sequence retained for downstream analysis.
x	discard	Sequence that is consumed but not needed in output.
f	fixed sequence	Literal adapter, linker, or anchor sequence.

The role records intent and lets the compiler build named sequence components; it does not by itself correct or filter an interval.

Interval shapes

Shape	Meaning	Example
[n]	Exactly n bases.	u[10]
[a-b]	Between a and b bases, inclusive.	b[9-10]
:	The remaining unbounded sequence.	r:
[ACGT]	A literal nucleotide sequence; valid for f.	f[CAGAGC]

Variable and unbounded intervals need a surrounding layout that makes their boundaries determinable. An anchor is a common way to delimit a variable prefix.

Read layouts

Braces group intervals belonging to an input read:

1{b[16]u[10]}
2{r:}

Paired files are supplied as --file1 and --file2. The output transformation uses the same numbered layout notation:

1{b<bc>[16]u<umi>[10]}
2{r<bio>:}
-> 1{<bc><umi>} 2{<bio>}

An output can contain only extracted labels; it need not reproduce consumed anchors or discarded sequence.

Definitions and references

Prefer definitions when an interval has a protocol-level name or operation:

bc = filter_within_dist(b[8], "barcodes.txt", 1)
umi = u[10]
1{<bc><umi>r:}
-> 1{<bc><umi>}

Definitions also provide the attachment point for operation annotations such as #[edit(1)] and the property annotation #[ambig_policy = no_match].