IPO

Version: 1.4.1

Website

https://github.com/glibiseller/IPO

Description

Isotopologue parameter optimisation (IPO) is designed to increase the reliability of peak picking (centWave and matchedFilter), retention time alignment (orbiwarp) and grouping (density) of XCMS, by optimising the parameters. A design of experiments approach (DoE) is used to optimised parameters. Box-Behnken designs (BBD) are used to create DoE. Feature detection parameters are optimised first, followed by simultaneous optimisation of retention time correction and group parameters. Peak picking scores are calculated based on the ratio of reliable peaks (isotopologues). Relative retention time deviations are minimised to produce the retention time correction score (RCS) using the inverse of the average retention time score. The ratio of the squared reliable groups to non-reliable groups is the grouping scores (GS). The norms of the RSC and GS are then combined as the retention time correction and grouping target value (RGTV). These scores are the evaluated using response-surface models. The centre of the next DoE is the combination of parameters that produced the best score in the response-surface models.

Functionality

• Other Tools/Software Optimisation

Instrument Data Type

• MS/LC-MS

Approaches

• Untargeted

Computer Skills

Medium

Software Type

R Package

Interface

Command line interface

Operating System (OS)

Unix/Linux, Mac OS, Windows

Language

R

Dependencies

xcms, rsm, CAMERA, grDevices, graphics, stats, utils

Input Formats - Open

mzData, mzML, mzXML, netCDF

Input Formats - Proprietary

N/A

Published

2015

Last Updated

2017

License

GPL (≥ 2) + file LICENSE

Paper

http://www.ncbi.nlm.nih.gov/pubmed/25888443

PMID

25888443