Read 'Shimadzu' IT-TOF stream — read_sz

Read the TTFL Raw Data mass spectrometry stream written by the 'Shimadzu' LCMS-IT-TOF to .lcd files.

Usage

read_sz_ittof(
  path,
  format_out = c("matrix", "data.frame", "data.table"),
  data_format = "long",
  levels = c("MS1", "MS2"),
  read_metadata = TRUE,
  metadata_format = "shimadzu_lcd",
  scale = TRUE
)

Arguments

path: Path to .lcd file.
format_out: Class of output. Either data.table or data.frame. Spectra are long and mixed-type, so matrix resolves to data.table.
data_format: Ignored: spectra are always returned in long format.
levels: Which MS levels to return, spelled MS1 and MS2. Both are decoded either way.
read_metadata: Logical. Whether to attach metadata.
metadata_format: Format to output metadata.
scale: Logical. Whether to scale the stored values to the units of the total ion current. FALSE returns them as stored.

Value

A named list holding whichever of the requested levels the file has, each a long table with columns scan, rt (minutes), mz and intensity, and a precursor_mz column before mz for MS2. A per-scan summary (retention time, event, MS level, polarity, precursor m/z, point count) is attached to each as a scan_info attribute.

Details

Data Index holds one 16-byte record per scan: the scan's byte offset into MS Raw Data (uint64), the cycle number and a scan counter. Each scan in MS Raw Data is a 256-byte header followed by a run-length-encoded profile. The header fields read here are:

Offset	Type	Field
0–3	`uint32`	Scan index (0-based)
4–7	`uint32`	Cycle number
8–11	`uint32`	Start of the cycle (milliseconds), as in `Retention Time`
12–15	`uint32`	Acquisition time of the scan (milliseconds)
16–19	`uint32`	Acquisition event (0-based)
20–23	`uint32`	MS level in the upper 16 bits, polarity in the lower 16 (0 = positive, 1 = negative)
144–147	`uint32`	Last flight-time bin of the scan range
148–151	`uint32`	Flight time of bin 0
154–155	`uint16`	Intensity multiplier
164–167	`uint32`	Length of the profile in 16-bit words
168–171	`uint32`	Total ion current
176–179	`uint32`	Precursor m/z x 10000 (MS2 and above)

Profile. The profile is a sequence of 16-bit words. A word with bit 15 clear is a gap: the number of empty flight-time bins to skip. A word with bit 15 set opens a run, and its lower 15 bits give the number of intensities that follow, one per consecutive bin; the intensities themselves use all 16 bits. A run of length 0 ends the scan.

Intensities. The header's total ion current, and the TIC Data streams, equal the sum of the stored values times the scan's intensity multiplier, so scale = TRUE multiplies them by it. The multiplier is usually 128, but reaches 4096 in some intense scans of PXD025121. 'LabSolutions' displays and exports the profile after a 5-point moving average, which is not applied here.

Mass axis. Bin b of a scan has flight time t = b + t0, where t0 is read from the header. TTFL Tuning/Tuning Result 00 stores, for each polarity, the reference masses of the calibrant (sodium trifluoroacetate clusters) with their measured flight times. Between two reference ions, sqrt(mz) is the quadratic in t through that pair and the next ion up; below the first, it is the line through the first two. The last interval and anything above it reuse the quadratic through the last three ions.

Checked against the vendor's own TTFLDataExport profile export of an MTBLS688 file, this reproduces every intensity of all 3002 scans exactly, once the 5-point moving average is applied, and every m/z to the four decimal places the export keeps. The centroids of an MTBLS432 file (vendor ANDI-MS exports) and of a PXD020792 file (Mascot Distiller peak lists) agree to within a fraction of a flight-time bin. Above the last reference ion, the median error against the Mascot Distiller centroids of PXD025121 files stays within 2 ppm up to m/z 1500.

Author

Ethan Bass