oxbow.core.PySamScanner

oxbow.core.PySamScanner#

class oxbow.core.PySamScanner(src, compressed=False, fields=None, tag_defs=None, coords=None)#

A SAM file scanner.

Parameters:

  • src (str or file-like) – The path to the SAM file or a file-like object.

  • compressed (bool, optional [default: False]) – Whether the source is BGZF-compressed.

  • fields (str or list[str] or None, optional [default: "*"]) – Standard SAM fields to include. "*" for all, None to omit, or a list of field names.

  • tag_defs (list[tuple[str, str]], optional [default: None]) – Tag definitions for the "tags" struct column. None omits the tags column. Use the tag_defs() method to discover definitions.

  • coords (Literal["01", "11"], optional [default: "11"]) – Coordinate system for returning positions and interpreting query ranges. “01” for 0-based half-open, “11” for 1-based closed.

__init__()#

Methods

__init__()

chrom_names()

Return the names of the reference sequences.

chrom_sizes()

Return the names of the reference sequences and their lengths in bp.

field_names()

Return the names of the standard SAM fields.

model()

Return the string representation of the alignment model.

scan([columns, batch_size, limit])

Scan batches of records from the file.

scan_byte_ranges(byte_ranges[, columns, ...])

Scan batches of records from specified byte ranges in the file.

scan_query(region[, index, columns, ...])

Scan batches of records from a genomic range query on a BGZF-encoded file.

scan_unmapped([index, columns, batch_size, ...])

Scan batches of records from the set of unaligned reads.

scan_virtual_ranges(vpos_ranges[, columns, ...])

Scan batches of records from virtual position ranges in a BGZF file.

schema()

Return the Arrow schema.

tag_defs([scan_rows])

Discover tag definitions by sniffing scan_rows records.
chrom_names()#

Return the names of the reference sequences.

chrom_sizes()#

Return the names of the reference sequences and their lengths in bp.

field_names()#

Return the names of the standard SAM fields.

model()#

Return the string representation of the alignment model.

scan(columns=None, batch_size=1024, limit=None)#

Scan batches of records from the file.

Parameters:

  • columns (list[str], optional) – Names of the top-level columns to project.

  • batch_size (int, optional [default: 1024]) – The number of records to include in each batch.

  • limit (int, optional) – The maximum number of records to scan. If None, records are scanned until EOF.

Returns:

An iterator yielding Arrow record batches.

Return type:

arro3 RecordBatchReader (pycapsule)

scan_byte_ranges(byte_ranges, columns=None, batch_size=1024, limit=None)#

Scan batches of records from specified byte ranges in the file.

The byte positions must align with record boundaries.

Parameters:

  • byte_ranges (list[tuple[int, int]]) – List of (start, end) byte position tuples to read from.

  • columns (list[str], optional) – Names of the top-level columns to project.

  • batch_size (int, optional [default: 1024]) – The number of records to include in each batch.

  • limit (int, optional) – The maximum number of records to scan. If None, all records in the specified ranges are scanned.

Returns:

An iterator yielding Arrow record batches.

Return type:

arro3 RecordBatchReader (pycapsule)

scan_query(region, index=None, columns=None, batch_size=1024, limit=None)#

Scan batches of records from a genomic range query on a BGZF-encoded file.

This operation requires an index file.

Parameters:

  • region (str) – Genomic range string in the format “chr:start-end”, “chr:[start,end]” or “chr:[start,end)”.

  • index (path or file-like, optional) – The index file to use for querying the region. If None and the source was provided as a path, we will attempt to load the index from the same path with an additional extension.

  • columns (list[str], optional) – Names of the top-level columns to project.

  • batch_size (int, optional [default: 1024]) – The number of records to include in each batch.

  • limit (int, optional) – The maximum number of records to scan. If None, all records intersecting the query range are scanned.

Returns:

An iterator yielding Arrow record batches.

Return type:

arro3 RecordBatchReader (pycapsule)

scan_unmapped(index=None, columns=None, batch_size=1024, limit=None)#

Scan batches of records from the set of unaligned reads.

This operation requires an index file.

Parameters:

  • index (path or file-like, optional) – The index file to use for querying the region. If None and the source was provided as a path, we will attempt to load the index from the same path with an additional extension.

  • columns (list[str], optional) – Names of the top-level columns to project.

  • batch_size (int, optional [default: 1024]) – The number of records to include in each batch.

  • limit (int, optional) – The maximum number of records to scan. If None, records are scanned until EOF.

Returns:

An iterator yielding Arrow record batches.

Return type:

arro3 RecordBatchReader (pycapsule)

scan_virtual_ranges(vpos_ranges, columns=None, batch_size=1024, limit=None)#

Scan batches of records from virtual position ranges in a BGZF file.

The virtual positions must align with record boundaries. That means that the compressed offset must point to the beginning of a BGZF block and the uncompressed offset must point to the beginning or end of a record decoded within the block.

Parameters:

  • vpos_ranges (list[tuple[vpos, vpos]]) – List of virtual position ranges as pairs. Each virtual position can be given as either a packed virtual position (int), or an unpacked tuple of ints (c, u) specifying the compressed and uncompressed offsets, respectively.

  • columns (list[str], optional) – Names of the top-level columns to project.

  • batch_size (int, optional [default: 1024]) – The number of records to include in each batch.

  • limit (int, optional) – The maximum number of records to scan. If None, all records in the specified ranges are scanned.

Returns:

An iterator yielding Arrow record batches.

Return type:

arro3 RecordBatchReader (pycapsule)

schema()#

Return the Arrow schema.

Return type:

arro3 Schema (pycapsule)

tag_defs(scan_rows=1024)#

Discover tag definitions by sniffing scan_rows records.

The reader stream is reset to its original position after scanning.

Parameters:

scan_rows (int, optional [default: 1024]) – The number of records to scan. If None, all records are scanned.

Returns:

A list of tag definitions, where each definition is a tuple of the tag name and the SAM tag type code.

Return type:

list[tuple[str, str]]

Contents