RawRead

RawRead.cs contains the C# code for extraction of all info in raw file. This is an updated version of RawRead.cs using the .NET 8 assemblies distributed in the ThermoFisher RawFileReader repository. The repository and its fork, contains the ThermoFisher CommonCore 8.0.37 packages under Libs/NetCore/Net8/. The project keeps the original RawRead functionality: RAW metadata, mass-precision output, base-peak chromatogram, centroid/profile MGF output, MS1 filtered peak output, and FFT output.

Quick start

wget https://dot.net/v1/dotnet-install.sh -O dotnet-install.sh
bash dotnet-install.sh --channel 8.0 --install-dir "$PWD"
git clone https://github.com/animesh/RawFileReader
git clone https://github.com/animesh/RawRead
cd RawRead
../dotnet build -c Release
../dotnet bin/Release/net8.0/RawRead.dll 171010_Ip_Hela_ugi.raw

Key outputs

<raw>.MZ.txt — mass precision estimates (Mass, mmu, ppm) <raw>.chromatogram.txt — BasePeak chromatogram (RT, intensity) <raw>.centroid.MGF, <raw>.profile.MGF — MS2 blocks for centroid/profile scans <raw>.profile.intensity{insThr}.charge{chgThr}.MS.txt — per-scan profile/intensity listing <raw>.intensity{insThr}.charge{chgThr}.FFT.txt — FFT of max-intensity over scan index (see below) <raw>.mzFFT.txt — mass-domain FFT of averaged MS1 spectrum (see below)

FFT output interpretation

The FFT file columns are bin, frequency_per_min, period_min, magnitude.

The signal being transformed is the per-scan maximum centroid intensity, sampled at the acquisition rate. Each FFT bin k corresponds to a periodic pattern that repeats every period_min = 1 / frequency_per_min minutes across the run. Only the first N/2 bins are written (the second half is the conjugate mirror of a real-valued input).

Key bins in a typical DDA run (example: 10.7-min Orbitrap Elite, 1221 scans):

bin	period	interpretation
0	∞	DC — mean intensity across all scans
1	= run length	Full LC gradient envelope
2–5	half/quarter run	Broad chromatographic cluster envelopes
~11	~1 min	DDA TopN precursor cycle repetition period
~31	~20 sec	MS2 burst rate within a TopN cycle
72–98	6–9 sec	Individual survey + fragmentation scan timing

High-magnitude bins at short periods (< 0.2 min) indicate the instrument's raw scan repeat rate. A strong bin near 1 minute is characteristic of DDA TopN methods. Peaks in the 2–5 min range reflect the chromatographic peak cluster structure of the gradient.

Mass-domain FFT (<raw>.mzFFT.txt)

Columns: bin, spacing_Da, magnitude.

The averaged MS1 spectrum is placed on a 0.01 Da grid and FFT'd. A magnitude peak at a given spacing_Da means ions appear at regular intervals of that many Da across the averaged spectrum. This reveals:

spacing_Da	interpretation
1.0	z=1 isotope envelope
0.5	z=2 isotope envelope (most peptides)
0.33	z=3
0.25, 0.2, …	z=4, z=5, … — multiply charged peptides
14	CH₂ repeat — lipid/fatty acid series
22	CO₂ loss ladder or specific polymer
44	CO₂ (44 Da) repeat — polyether/lipid
162	hexose sugar unit — glycan series

Strong peaks at sub-Da spacings (0.1–0.5 Da) directly report the dominant charge states in the run. A run dominated by doubly and triply charged tryptic peptides will show the highest magnitude at spacing_Da ≈ 0.33–0.50.

Notable behaviors & caveats

Uses GetReaction(0).PrecursorMass for PEPMASS in MGF; some scans may lack reactions — code assumes they exist. Extracts charge from trailer labels by matching "Charge State:" — trailer labeling may vary by instrument/firmware. Heuristic branches (e.g., title.Contains(" ms ")) determine some output formats; these heuristics may not be universal. Large files: the code keeps arrays sized by the number of scans and may use significant memory for very long runs.

The ThermoFisher packages are version 8.0.37. The project uses MathNet.Numerics 5.0.0 for the FFT, which is compatible with .NET 8.

Build and Run with system-wide dotnet installation

dotnet build -c Release
dotnet run -c Release -- /path/to/file.raw
dotnet run -c Release -- /path/to/file.raw 1000
dotnet run -c Release -- /path/to/file.raw 1000 2

After building:

dotnet bin/Release/net8.0/RawRead.dll /path/to/file.raw

The RAW reader itself is the ThermoFisher CommonCore implementation. No old Planet Orbitrap / legacy DLL references remain.

Important migration changes

• Target framework changed from the old Mono/.NET Framework compilation model to net8.0.
• ThermoFisher CommonCore packages changed to 8.0.37, matching the uploaded repository.
• The old command-line mcs/csc reference instructions were replaced with a .csproj.
• The scan array is indexed relative to FirstSpectrum, rather than assuming the first scan is 1.
• FFT phase uses Atan2 rather than Atan(imaginary / real).
• Numeric command-line parsing uses invariant culture.
• RAW acquisition/error handling was made explicit.

The actual Thermo API calls used by RawRead remain the same APIs exposed by the 8.0.37 assemblies, including RawFileReaderAdapter.FileFactory, GetSegmentedScanFromScanNumber, GetCentroidStream, chromatogram access, and mass-precision estimation.