pretreatFTICR.R script

Raw FT-ICR DMS files are very large and are painful to analyse directly. Two reduction methods can be used and combined:

1. filtering the data to match as well as possible a predefined regular grid

2. keeping only the data that match a list of targets

Combining both methods might enable to reduce the MS size by nearly 90%, in the usual applications.

The script is controlled by the following parameters:

#====================================================
# User configuration params =========================
#====================================================

## Data and results directories
origMsDir = 'Test_FTICR_2/FTICR'
compMsDir = 'Test_FTICR_2/FTICR_compressed'
ms_type   = 'fticr'

## Compress mode
compMode  = c('grid','targets','grid+targets')[3]

## Grid specifications
mzMin     = 70
mzMax     = 250
dmz       = 0.001

## Targets specifications
tgTable   = 'Test_FTICR_2/targets_list.csv'
dmzTarget = 0.5 # Delta m/z to keep around target

## Short run to check ?
test      = FALSE

where

• origMsDir: (string) is the path to a directory containing the files to treat (all files within the directory will be treated).

• compMsDir: (string) is the path to a directory where the compressed files will be stored.

• ms_type: (string) is the type of MS to treat. Only fticr at the moment.

• compMode: (string) compression modality, which can be grid, targets or their combination grid+targets.

• mzMin, mzMax and dmz: (numbers) define the regular grid used for grid selection.

• tgTable: (string) is the path to the targets file used for targets selection.

• dmzTarget: (number) is the half-width of the m/z interval to keep around a target. The interval is centered on ms_ref vales in tgTable. The width should be large enough to not miss the peaks, but small enough for efficient compression…

• test: (logical) test run only, to check if data are OK.

Note: it is best to store the spectra as compressed .gz files. The disp space is well reduced and all scripts can handle them transparently.