Shared featurization for PeptDeep RT, CCS, and MS2 ONNX predictors. More...

Static Public Member Functions
static PeptDeepInputBatch	buildUnmodifiedPeptideBatch (const std::vector< std::string > &peptides, const PeptDeepInputConfig &config=PeptDeepInputConfig())
	Tokenizes a batch of unmodified peptides into a padded aa_indices/mod_x tensor pair.

static PeptDeepInputBatch	buildUnmodifiedChargedBatch (const std::vector< std::string > &peptides, const std::vector< float > &charges, const PeptDeepInputConfig &config=PeptDeepInputConfig())
	Like buildUnmodifiedPeptideBatch(), additionally populating the (scaled) charges field.

static PeptDeepInputBatch	buildUnmodifiedInstrumentBatch (const std::vector< std::string > &peptides, const std::vector< float > &charges, const std::vector< float > &nces, const std::vector< int64_t > &instrument_indices, const PeptDeepInputConfig &config=PeptDeepInputConfig())
	Like buildUnmodifiedChargedBatch(), additionally populating the (scaled) nces and instrument_indices fields.

Detailed Description

Shared featurization for PeptDeep RT, CCS, and MS2 ONNX predictors.

Modified peptides are intentionally rejected for now. The generated mod_x tensor is zero-filled and therefore represents unmodified peptides only.

Member Function Documentation

static PeptDeepInputBatch buildUnmodifiedChargedBatch	(	const std::vector< std::string > &	peptides,
		const std::vector< float > &	charges,
		const PeptDeepInputConfig &	config = `PeptDeepInputConfig()`
	)

static

Like buildUnmodifiedPeptideBatch(), additionally populating the (scaled) charges field.

Parameters

peptides	A vector of raw uppercase peptide strings.
charges	Precursor charge states, one per peptide. Must match peptides size.
config	Padding policy (terminal tokens, fixed sequence length).

Returns: A PeptDeepInputBatch with aa_indices, mod_x, and charges populated.

Exceptions

Exception::IllegalArgument if charges.size() != peptides.size(), or via buildUnmodifiedPeptideBatch().

static PeptDeepInputBatch buildUnmodifiedInstrumentBatch	(	const std::vector< std::string > &	peptides,
		const std::vector< float > &	charges,
		const std::vector< float > &	nces,
		const std::vector< int64_t > &	instrument_indices,
		const PeptDeepInputConfig &	config = `PeptDeepInputConfig()`
	)

static

Like buildUnmodifiedChargedBatch(), additionally populating the (scaled) nces and instrument_indices fields.

Parameters

peptides	A vector of raw uppercase peptide strings.
charges	Precursor charge states, one per peptide. Must match peptides size.
nces	Normalized collision energies, one per peptide. Must match peptides size.
instrument_indices	Categorical instrument identifiers, one per peptide. Must match peptides size.
config	Padding policy (terminal tokens, fixed sequence length).

Exceptions

Exception::IllegalArgument if any of charges/nces/instrument_indices does not match peptides.size(), or via buildUnmodifiedChargedBatch().

static PeptDeepInputBatch buildUnmodifiedPeptideBatch	(	const std::vector< std::string > &	peptides,
		const PeptDeepInputConfig &	config = `PeptDeepInputConfig()`
	)

static

Tokenizes a batch of unmodified peptides into a padded aa_indices/mod_x tensor pair.

Parameters

peptides	A vector of raw uppercase peptide strings. Must be non-empty.
config	Padding policy (terminal tokens, fixed sequence length).

Returns: A PeptDeepInputBatch with aa_indices and a zero-filled mod_x populated; charges/nces/instrument_indices left empty.

Exceptions

Exception::IllegalArgument	if peptides is empty or contains an invalid sequence (see validatePeptide()).
Exception::InvalidValue	if config.fixed_sequence_length is shorter than the longest encoded peptide.