The BAM format. More...

#include <seqan3/io/sam_file/format_bam.hpp>

Inheritance diagram for seqan3::format_bam:

Classes
struct	alignment_record_core
	Stores all fixed length variables which can be read/written directly by reinterpreting the binary stream. More...

Public Member Functions
Constructors, destructor and assignment
	format_bam ()=default
	Defaulted.

	format_bam (format_bam const &)=default
	Defaulted.

format_bam &	operator= (format_bam const &)=default
	Defaulted.

	format_bam (format_bam &&)=default
	Defaulted.

format_bam &	operator= (format_bam &&)=default
	Defaulted.

	~format_bam ()=default
	Defaulted.

Static Public Attributes
static std::vector< std::string >	file_extensions {{"bam"}}
	The valid file extensions for this format; note that you can modify this value.

Protected Member Functions
template<typename stream_type , typename seq_legal_alph_type , typename ref_seqs_type , typename ref_ids_type , typename stream_pos_type , typename seq_type , typename id_type , typename ref_seq_type , typename ref_id_type , typename ref_offset_type , typename cigar_type , typename flag_type , typename mapq_type , typename qual_type , typename mate_type , typename tag_dict_type , typename e_value_type , typename bit_score_type >
void	read_alignment_record (stream_type &stream, sam_file_input_options< seq_legal_alph_type > const &options, ref_seqs_type &ref_seqs, sam_file_header< ref_ids_type > &header, stream_pos_type &position_buffer, seq_type &seq, qual_type &qual, id_type &id, ref_seq_type &ref_seq, ref_id_type &ref_id, ref_offset_type &ref_offset, cigar_type &cigar_vector, flag_type &flag, mapq_type &mapq, mate_type &mate, tag_dict_type &tag_dict, e_value_type &e_value, bit_score_type &bit_score)
	Read from the specified stream and back-insert into the given field buffers.

template<typename stream_type , typename header_type , typename seq_type , typename id_type , typename ref_seq_type , typename ref_id_type , typename cigar_type , typename qual_type , typename mate_type , typename tag_dict_type >
void	write_alignment_record (stream_type &stream, sam_file_output_options const &options, header_type &&header, seq_type &&seq, qual_type &&qual, id_type &&id, ref_seq_type &&ref_seq, ref_id_type &&ref_id, std::optional< int32_t > ref_offset, cigar_type &&cigar_vector, sam_flag const flag, uint8_t const mapq, mate_type &&mate, tag_dict_type &&tag_dict, double e_value, double bit_score)
	Write the given fields to the specified stream.

Private Member Functions
std::vector< cigar >	parse_binary_cigar (std::string_view const cigar_str) const
	Parses a cigar string into a vector of operation-count pairs (e.g. (M, 3)).

template<typename stream_view_type >
void	read_float_byte_field (stream_view_type &&stream_view, float &target)
	Reads a float field from binary stream by directly reinterpreting the bits.

template<std::integral number_type>
void	read_integral_byte_field (std::string_view const str, number_type &target)
	This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.

template<typename stream_view_type , std::integral number_type>
void	read_integral_byte_field (stream_view_type &&stream_view, number_type &target)
	Reads a arithmetic field from binary stream by directly reinterpreting the bits.

void	read_sam_dict (std::string_view const tag_str, sam_tag_dictionary &target)
	Reads the optional tag fields into the seqan3::sam_tag_dictionary.

template<typename value_type >
int32_t	read_sam_dict_vector (seqan3::detail::sam_tag_variant &variant, std::string_view const str, value_type const &value)
	Reads a list of values separated by comma as it is the case for SAM tag arrays.

template<typename stream_t , typename header_type >
void	write_header (stream_t &stream, sam_file_output_options const &options, header_type &header)
	Writes the SAM header.

Private Member Functions inherited from seqan3::detail::format_sam_base
template<typename ref_id_type , typename ref_id_tmp_type , typename header_type , typename ref_seqs_type >
void	check_and_assign_ref_id (ref_id_type &ref_id, ref_id_tmp_type &ref_id_tmp, header_type &header, ref_seqs_type &)
	Checks for known reference ids or adds a new reference is and assigns a reference id to `ref_id`.

template<arithmetic arithmetic_target_type>
void	read_arithmetic_field (std::string_view const &str, arithmetic_target_type &arithmetic_target)
	Reads arithmetic fields using std::from_chars.

template<std::ranges::forward_range target_range_type>
void	read_forward_range_field (std::string_view const str, target_range_type &target)
	Reads from `str` to `target`, converting values with seqan3::views::char_to.

template<typename stream_view_type , std::ranges::forward_range target_range_type>
void	read_forward_range_field (stream_view_type &&stream_view, target_range_type &target)
	Reads a range by copying from stream_view to target, converting values with seqan3::views::char_to.

template<typename stream_view_type , typename ref_ids_type , typename ref_seqs_type , typename seq_legal_alph_type >
void	read_header (stream_view_type &&stream_view, sam_file_header< ref_ids_type > &hdr, ref_seqs_type &, sam_file_input_options< seq_legal_alph_type > const &options)
	Reads the SAM header.

int32_t	soft_clipping_at_front (std::vector< cigar > const &cigar_vector) const
	Returns the soft clipping value at the front of the `cigar_vector` or 0 if none present.

template<typename stream_t , typename header_type >
void	write_header (stream_t &stream, sam_file_output_options const &options, header_type &header)
	Writes the SAM header.

	format_sam_base ()=default
	Defaulted.

	format_sam_base (format_sam_base const &)=default
	Defaulted.

format_sam_base &	operator= (format_sam_base const &)=default
	Defaulted.

	format_sam_base (format_sam_base &&)=default
	Defaulted.

format_sam_base &	operator= (format_sam_base &&)=default
	Defaulted.

	~format_sam_base ()=default
	Defaulted.

Static Private Member Functions
static std::string	get_tag_dict_str (sam_tag_dictionary const &tag_dict)
	Writes the optional fields of the seqan3::sam_tag_dictionary.

static uint16_t	reg2bin (int32_t beg, int32_t end) noexcept
	Computes the bin number for a given region [beg, end), copied from the official SAM specifications.

Private Attributes
bool	header_was_read {false}
	A variable that tracks whether the content of header has been read or not.

std::string	string_buffer {}
	Local buffer to read into while avoiding reallocation.

Private Attributes inherited from seqan3::detail::format_sam_base
std::array< char, 316 >	arithmetic_buffer {}
	A buffer used when parsing arithmetic values with std::from_chars.

bool	header_was_written {false}
	A variable that tracks whether the content of header has been written or not.

bool	ref_info_present_in_header {false}
	Tracks whether reference information (@SQ tag) were found in the SAM header.

Static Private Attributes
static constexpr std::array< uint8_t, 256 >	char_to_sam_rank
	Converts a cigar op character to the rank according to the official BAM specifications.

Static Private Attributes inherited from seqan3::detail::format_sam_base
static constexpr std::array	format_version {'1', '.', '6'}
	The format version string.

Detailed Description

The BAM format.

The BAM format is the binary version of the SAM format:

Introduction

SAM is often used for storing alignments of several read sequences against one or more reference sequences. See the article on wikipedia for an introduction of the format or look into the official SAM format specifications. SeqAn implements version 1.6 of the SAM specification.

Take a look at our tutorial SAM Input and Output in SeqAn for a walk through of how to read SAM/BAM files.

fields_specialisation

The SAM format provides the following fields: seqan3::field::seq, seqan3::field::qual, seqan3::field::id, seqan3::field::ref_seq, seqan3::field::ref_id seqan3::field::ref_offset, seqan3::field::offset, seqan3::field::flag, seqan3::field::mapq and seqan3::field::mate. In addition there is the seqan3::field::header_ptr, which is usually only used internally to provide the range-based functionality of the file.

None of the fields are required when writing. If they are not given, a default value of '0' for numeric fields and '*' for other fields is used.

SAM format columns -> fields

Since many users will be accustomed to the columns of the SAM format, here is a mapping of the common SAM format columns to the SeqAn record fields:

#	SAM Column ID	FIELD name
1	QNAME	seqan3::field::id
2	FLAG	seqan3::field::flag
3	RNAME	seqan3::field::ref_id
4	POS	seqan3::field::ref_offset
5	MAPQ	seqan3::field::mapq
6	CIGAR	seqan3::field::cigar
7	RNEXT	seqan3::field::mate (tuple pos 0)
8	PNEXT	seqan3::field::mate (tuple pos 1)
9	TLEN	seqan3::field::mate (tuple pos 2)
10	SEQ	seqan3::field::seq
11	QUAL	seqan3::field::qual

Format Check

The format checks are implemented according to the official SAM format specifications in order to ensure correct SAM file output.

If a non-recoverable format violation is encountered on reading, or you specify invalid values/combinations when writing, seqan3::format_error is thrown.

Header implementation

The SAM header (if present) is read/written once in the beginning before the first record is read/written.

Remarks: For a complete overview, take a look at SAM File; For a complete overview, take a look at SAM File

Member Function Documentation

◆ get_tag_dict_str()

std::string seqan3::format_bam::get_tag_dict_str ( sam_tag_dictionary const & tag_dict )

inlinestaticprivate

Writes the optional fields of the seqan3::sam_tag_dictionary.

Parameters

[in] tag_dict The tag dictionary to print.

◆ parse_binary_cigar()

std::vector< cigar > seqan3::format_bam::parse_binary_cigar ( std::string_view const cigar_str ) const

inlineprivate

Parses a cigar string into a vector of operation-count pairs (e.g. (M, 3)).

Parameters

[in] cigar_str A std::string_view that points to the information of the CIGAR string in the BAM file.

Returns: A std::vector over seqan3::cigar, that describes the alignment.

◆ read_alignment_record()

template<typename stream_type , typename seq_legal_alph_type , typename ref_seqs_type , typename ref_ids_type , typename stream_pos_type , typename seq_type , typename id_type , typename ref_seq_type , typename ref_id_type , typename ref_offset_type , typename cigar_type , typename flag_type , typename mapq_type , typename qual_type , typename mate_type , typename tag_dict_type , typename e_value_type , typename bit_score_type >

void seqan3::format_bam::read_alignment_record	(	stream_type &	stream,
		sam_file_input_options< seq_legal_alph_type > const &	options,
		ref_seqs_type &	ref_seqs,
		sam_file_header< ref_ids_type > &	header,
		stream_pos_type &	position_buffer,
		seq_type &	seq,
		qual_type &	qual,
		id_type &	id,
		ref_seq_type &	ref_seq,
		ref_id_type &	ref_id,
		ref_offset_type &	ref_offset,
		cigar_type &	cigar_vector,
		flag_type &	flag,
		mapq_type &	mapq,
		mate_type &	mate,
		tag_dict_type &	tag_dict,
		e_value_type &	e_value,
		bit_score_type &	bit_score
	)

inlineprotected

Read from the specified stream and back-insert into the given field buffers.

Template Parameters

stream_type	The input stream type; Must be derived from std::ostream.
stream_pos_type	Type of the position buffer, aka, the std::streampos of the current record.
ref_seqs_type	e.g. std::deque<ref_sequence_type> or decltype(std::ignore).
seq_type	Type of the seqan3::field::seq input (see seqan3::sam_file_input_traits).
qual_type	Type of the seqan3::field::qual input (see seqan3::sam_file_input_traits).
id_type	Type of the seqan3::field::id input (see seqan3::sam_file_input_traits).
ref_seq_type	Type of the seqan3::field::ref_seq input (see seqan3::sam_file_input_traits).
ref_id_type	Type of the seqan3::field::ref_id input (see seqan3::sam_file_input_traits).
ref_offset_type	Type of the seqan3::field::ref_offset input (see seqan3::sam_file_input_traits).
cigar_type	Type of the seqan3::field::cigar input (a std::vector<cigar> or std::ignore).
flag_type	Type of the seqan3::field::flag input (see seqan3::sam_file_input_traits).
mapq_type	Type of the seqan3::field::mapq input (see seqan3::sam_file_input_traits).
mate_type	std::tuple<ref_id_type, ref_offset_type, int32_t> or decltype(std::ignore).
tag_dict_type	seqan3::sam_tag_dictionary or decltype(std::ignore).
e_value_type	Type of the seqan3::field::evalue input (see seqan3::sam_file_input_traits).
bit_score_type	Type of the seqan3::field::bit_score input (see seqan3::sam_file_input_traits).

Parameters

[in,out]	stream	The input stream to read from.
[in,out]	position_buffer	The buffer to store the current record's position.
[in]	options	File specific options passed to the format.
[out]	ref_seqs	The reference sequences to the corresponding alignments.
[out]	header	A pointer to the seqan3::sam_file_header object.
[out]	seq	The buffer for seqan3::field::seq input.
[out]	qual	The buffer for seqan3::field::qual input.
[out]	id	The buffer for seqan3::field::id input.
[out]	ref_seq	The buffer for seqan3::field::ref_seq input.
[out]	ref_id	The buffer for seqan3::field::ref_id input.
[out]	ref_offset	The buffer for seqan3::field::ref_offset input.
[out]	cigar_vector	The buffer for seqan3::field::cigar input.
[out]	flag	The buffer for seqan3::field::flag input.
[out]	mapq	The buffer for seqan3::field::mapq input.
[out]	mate	The buffer for seqan3::field::mate input.
[out]	tag_dict	The buffer for seqan3::field::tags input.
[out]	e_value	The buffer for seqan3::field::evalue input.
[out]	bit_score	The buffer for seqan3::field::bit_score input.

Additional requirements

The function must also accept std::ignore as parameter for any of the fields, except stream, options and header. [This is enforced by the concept checker!]
In this case the data read for that field shall be discarded by the format.

◆ read_float_byte_field()

template<typename stream_view_type >

void seqan3::format_bam::read_float_byte_field	(	stream_view_type &&	stream_view,
		float &	target
	)

inlineprivate

Reads a float field from binary stream by directly reinterpreting the bits.

Template Parameters

stream_view_type The type of the stream as a view.

Parameters

[in,out]	stream_view	The stream view to read from.
[out]	target	An float value to store the parsed value in.

◆ read_integral_byte_field()

template<typename stream_view_type , std::integral number_type>

void seqan3::format_bam::read_integral_byte_field	(	stream_view_type &&	stream_view,
		number_type &	target
	)

inlineprivate

Reads a arithmetic field from binary stream by directly reinterpreting the bits.

Template Parameters

stream_view_type	The type of the stream as a view.
number_type	The type of number to parse; must model std::integral.

Parameters

[in,out]	stream_view	The stream view to read from.
[out]	target	An integral value to store the parsed value in.

◆ read_sam_dict()

void seqan3::format_bam::read_sam_dict	(	std::string_view const	tag_str,
		sam_tag_dictionary &	target
	)

inlineprivate

Reads the optional tag fields into the seqan3::sam_tag_dictionary.

Parameters

[in,out]	tag_str	The string_view to parse.
[out]	target	The seqan3::sam_tag_dictionary to store the tag information.

Exceptions

seqan3::format_error if any unexpected character or format is encountered.

Reading the tags is done according to the official SAM format specifications.

The function throws a seqan3::format_error if any unknown tag type was encountered. It will also fail if the format is not in a correct state (e.g. required fields are not given), but throwing might occur downstream of the actual error.

◆ read_sam_dict_vector()

template<typename value_type >

int32_t seqan3::format_bam::read_sam_dict_vector	(	seqan3::detail::sam_tag_variant &	variant,
		std::string_view const	str,
		value_type const &	value
	)

inlineprivate

Reads a list of values separated by comma as it is the case for SAM tag arrays.

Template Parameters

value_type The type of values to be stored in the tag array.

Parameters

[in,out]	variant	A std::variant object to store the tag arrays.
[in,out]	str	The string_view to parse.
[in]	value	A temporary value that determines the underlying type of the tag array.

Returns: The length of the vector processed.

Reading the tags is done according to the official SAM format specifications.

The function throws a seqan3::format_error if any unknown tag type was encountered. It will also fail if the format is not in a correct state (e.g. required fields are not given), but throwing might occur downstream of the actual error.

◆ write_alignment_record()

template<typename stream_type , typename header_type , typename seq_type , typename id_type , typename ref_seq_type , typename ref_id_type , typename cigar_type , typename qual_type , typename mate_type , typename tag_dict_type >

void seqan3::format_bam::write_alignment_record	(	stream_type &	stream,
		sam_file_output_options const &	options,
		header_type &&	header,
		seq_type &&	seq,
		qual_type &&	qual,
		id_type &&	id,
		ref_seq_type &&	ref_seq,
		ref_id_type &&	ref_id,
		std::optional< int32_t >	ref_offset,
		cigar_type &&	cigar_vector,
		sam_flag const	flag,
		uint8_t const	mapq,
		mate_type &&	mate,
		tag_dict_type &&	tag_dict,
		double	e_value,
		double	bit_score
	)

inlineprotected

Write the given fields to the specified stream.

Template Parameters

stream_type	Output stream, must model seqan3::output_stream_over with `char`.
seq_type	Type of the seqan3
id_type	Type of the seqan3
ref_seq_type	Type of the seqan3
ref_id_type	Type of the seqan3
ref_offset_type	Type of the seqan3
flag_type	Type of the seqan3
mapq_type	Type of the seqan3
qual_type	Type of the seqan3
mate_type	Type of the seqan3
tag_dict_type	Type of the seqan3
e_value_type	Type of the seqan3
bit_score_type	Type of the seqan3

Parameters

[in,out]	stream	The output stream to write into.
[in]	options	File specific options passed to the format.
[in]	header	A pointer to the header object of the file.
[in]	seq	The data for seqan3::field::seq, i.e. the query sequence.
[in]	qual	The data for seqan3::field::qual, e.g. the query quality sequence.
[in]	id	The data for seqan3::field::id, e.g. the read id.
[in]	ref_seq	The data for seqan3::field::ref_offset, i.e. the reference sequence.
[in]	ref_id	The data for seqan3::field::ref_id, e.g. the reference id..
[in]	ref_offset	The data for seqan3::field::ref_offset, i.e. the start position of the alignment in `ref_seq`.
[in]	cigar_vector	The data for seqan3::field::cigar, e.g. representing the alignment between query and ref.
[in]	flag	The data for seqan3::field::flag, e.g. the SAM mapping flag value.
[in]	mapq	The data for seqan3::field::mapq, e.g. the mapping quality value.
[in]	mate	The data for seqan3::field::mate, e.g. the mate information of paired reads.
[in]	tag_dict	The data for seqan3::field::tags, e.g. the optional SAM field tag dictionary.
[in]	e_value	The data for seqan3::field::e_value, e.g. the e-value of the alignment (BLAST).
[in]	bit_score	The data for seqan3::field::, e.g. the bit score of the alignment (BLAST).

◆ write_header()

template<typename stream_t , typename header_type >

void seqan3::format_bam::write_header	(	stream_t &	stream,
		sam_file_output_options const &	options,
		header_type &	header
	)

inlineprivate

Writes the SAM header.

Template Parameters

stream_t The stream type.

Parameters

[in,out]	stream	The stream to print to.
[in]	options	The options to alter printing.
[in]	header	The header to print.

Exceptions

seqan3::format_error if the header object contains the wrong information or the contents are ill-formed.

Before writing the header, the contents are checked for correctness according to the rules of the official SAM format specifications.

Member Data Documentation

◆ char_to_sam_rank

constexpr std::array<uint8_t, 256> seqan3::format_bam::char_to_sam_rank

staticconstexprprivate

Initial value:

{
        []() constexpr {
            std::array<uint8_t, 256> ret{};
 
            using index_t = std::make_unsigned_t<char>;
 
            
            ret[static_cast<index_t>('I')] = 1;
            ret[static_cast<index_t>('D')] = 2;
            ret[static_cast<index_t>('N')] = 3;
            ret[static_cast<index_t>('S')] = 4;
            ret[static_cast<index_t>('H')] = 5;
            ret[static_cast<index_t>('P')] = 6;
            ret[static_cast<index_t>('=')] = 7;
            ret[static_cast<index_t>('X')] = 8;
 
            return ret;
        }()
    }

Converts a cigar op character to the rank according to the official BAM specifications.

The documentation for this class was generated from the following file:

format_bam.hpp

Classes

Public Member Functions

Static Public Attributes

Protected Member Functions

Private Member Functions

Static Private Member Functions

Private Attributes

Static Private Attributes

Detailed Description

Introduction

fields_specialisation

SAM format columns -> fields

Format Check

Header implementation

Member Function Documentation

◆ get_tag_dict_str()

◆ parse_binary_cigar()

◆ read_alignment_record()

Additional requirements

◆ read_float_byte_field()

◆ read_integral_byte_field()

◆ read_sam_dict()

◆ read_sam_dict_vector()

◆ write_alignment_record()

◆ write_header()

Member Data Documentation

◆ char_to_sam_rank