msmu.read_diann

Reads DIA-NN output and returns a MuData object.

The block-diagonal precursor quantification is stored as a SciPy sparse .X: the precursor pivot is (n_precursor_obs x n_run) with ~one non-null per row, so only the observed cells are built (no dense pivot). Downstream tools handle the sparse .X transparently (absent cells restored as NaN).

Parameters:

Name	Type	Description	Default
identification_file	str | Path | list	Path to the DIA-NN output file or directory.	required
level	Literal['precursor', 'protein_group']	Level of the output to read ('precursor' or 'protein_group'). Note: 'protein_group' is not yet implemented.	'precursor'

Returns:

Type	Description
MuData	A MuData object containing the DIA-NN data.