phiperio

The phiperio package provides utilities to import, validate, and manage PhIP-Seq datasets, including standardized conversion pipelines, data checks, and access to cached peptide metadata.

Installation

You can install the development version of phiperio from GitHub with either pak or devtools:

# install.packages("pak")
# pak::pak("Polymerase3/phiperio")
install.packages("phiperio", repos = c('https://polymerase3.r-universe.dev', 'https://cloud.r-project.org'))

# or, using devtools:
# install.packages("devtools")
devtools::install_github("Polymerase3/phiperio")

Usage

For guided walk-throughs, see the pkgdown vignettes:

• Importing long tidy data (convert_standard) — cross-sectional and longitudinal tidy inputs.
• Importing multiple files at once — batch ingest of many CSV/Parquet files with sample_id_from_filenames.
• Importing legacy PhIP-Seq data (convert_legacy) — classic wide matrices (exist/fold_change/raw counts) plus sample/timepoint metadata.

Aim and key features

phiperio focuses on reliable ingest and validation of PhIP-Seq data, so downstream analyses start from a clean, standardized base. Key features include:

• DuckDB backend + Parquet first: uses DuckDB under the hood and writes/reads Parquet by default as the transaction layer between the phiper data source and phiperio, giving fast I/O and great interoperability.
• Scales to millions of rows: lazy database pipelines and Parquet storage let you work efficiently with very large PhIP-Seq datasets.
• Import helpers for common PhIP-Seq inputs and peptide metadata (peptide library cached and maintained in the companion phiper repo).
• Strong validation and consistency checks to catch data issues early.
• Lightweight, reproducible pipelines to standardize raw inputs into <phip_data> objects.

Issues

Spotted a bug or want to request a feature? Please open an issue: https://github.com/Polymerase3/phiperio/issues