Skip to contents

Convert LFQData + ProteinAnnotation to AnnData with round-trip metadata

Source: R/preprocess_DIANN_anndata.R
preprocess_anndata_from_lfq.Rd

Builds an AnnData object from LFQData and ProteinAnnotation, with the prolfquapp uns namespace for lossless round-trip via LFQData_from_anndata.

Usage

preprocess_anndata_from_lfq(lfqdata, protAnnot, source_software = "unknown")

Arguments

lfqdata

LFQData object

protAnnot

ProteinAnnotation object

source_software

character, name of the source software (e.g. "DIANN")

Value

anndataR::AnnData object

Details

Handles both protein-level (hierarchyDepth=1) and peptide-level (hierarchyDepth=2) data. For peptide-level data, var contains one row per feature (peptide) with protein annotation joined in.

Examples

if (requireNamespace("anndataR", quietly = TRUE)) {
  # Protein-level
  res <- sim_data_protAnnot(Nprot = 10, PROTEIN = TRUE)
  adata <- preprocess_anndata_from_lfq(res$lfqdata, res$pannot, "simulated")
  adata  # 12 obs x 10 var

  # Peptide-level (multiple peptides per protein)
  res2 <- sim_data_protAnnot(Nprot = 10, PROTEIN = FALSE)
  adata2 <- preprocess_anndata_from_lfq(res2$lfqdata, res2$pannot, "simulated")
  adata2  # 12 obs x ~28 var (peptides)

  # Round-trip back to LFQData
  back <- LFQData_from_anndata(adata2)
  back$lfqdata$hierarchy_counts()
}
#> creating sampleName from fileName column
#> completing cases
#> completing cases done
#> setup done
#> Warning: no exp_nr_children column specified, computing using nr_obs_experiment function
#> creating sampleName from fileName column
#> completing cases
#> completing cases done
#> setup done
#> Warning: no exp_nr_children column specified, computing using nr_obs_experiment function
#> # A tibble: 1 × 3
#>   isotopeLabel protein_Id peptide_Id
#>   <chr>             <int>      <int>
#> 1 light                10         28