MS2Planner

Input

It takes feature table generated from convert_to_table and raw signal .mzTab (curve mode) as input.

Output

It outputs best paths in .txt format, path_finder.log is the log file.

Run

python path_finder.py mode infile outfile intensity_threshold intensity_ratio num_path -infile_raw -intensity_accu -win_len -isolation -restriction -delay -min_scan -max_scan

Unit for retention time is second (sec) and mass-to-charge ratio is dalton (Da).

• mode: apex or baseline or curve
  • apex mode uses apex file and applies path finding algorithm to find best paths.
  • baseline mode simply splits retention time into small windows, length of the window is specified by win_len. At each window, top num_path intensity ions are selected.
  • curve mode uses apex and raw MS1 file. It applies path finding algorithm to find best paths.
• infile: input file, feature tables generated from convert_to_table. It contains five columns: mz, rt, charge, blank intensity, sample intensity. Blank and sample are told from the column index.
• outfile: output file.
• intensity_threshold: threshold for feature filtering. Any sample features below this threshold will be removed.
• intensity_ratio: ratio threshold for feature filtering. Any sample features with sample_intensity / blank_intensity < intensity_ratio wil be removed.
• num_path: number of paths.
• -infile_raw: raw MS1 .mzML file (only include sample, no control), argument for curve mode. File can be .mzML or .mzTab
• -intensity_accu: the amount of intensity user wants to collect on a single feature, argument for apex and curve modes.
• -win_len: (second) rt window length in baseline mode, argument for baseline mode.
• -isolation: (dalton) length of mass to charge isolation window, argument for all three modes.
• -restriction: (second, dalton) the first number is rt restriction and second is mz restriction. Restriction area is calculated as l1 norm. Features out of this area will not be involved, argument for curve mode.
• -delay: (second) the minimum rt requires swithching from one feature to the next, argument for all three modes.
• -min_scan: (second) minimum scan period in acquistion, argument for apex and curve mode.
• -max_scan: (second) maximum scan period in acquistion, argument for apex and curve mode.
• -cluster: clustering algorithm for curve mode. kNN and GMM are provided (only kNN available now).
• -sample: name of the sample (used for mzmine3 full feature table)
• -bg: name of the background sample (used for mzmine3 full feature table)
• -suffix: name of the suffix of the name (Area or Height) (used for mzmine3 full feature table)

Support for MZmine3

-sample, -bg and -suffix are all NECESSARY for parsing MZmine3 full feature table. When these fields are not None, MS2Planner will parse the full feature table and output the new-formatted path, otherwise output format is the original. Since new output format is comma separated, .csv format is recommended for output (instead of .txt).

Example

-sample Sample.mzML -bg Blank.mzML -suffix Area

The corresponding header of MZmine3 full feature table would be

DATAFILE:Sample.mzML:Area   DATAFILE:Blank.mzML:Area

Output Format (old)

path0 mz_center1 mz_isolation1 duration1 rt_start1 rt_end1 intensity1 apex_rt1 charge1 \t mz_center2 mz_center2 mz_window2 duration2 rt_start2 rt_end2 intensity2 apex_rt2 charge2...
path1 mz_center1 mz_isolation1 duration1 rt_start1 rt_end1 intensity1 apex_rt1 charge1...

Each path contains a row in .txt file. Following by

• mz_center: the center of the mz_window
• mz_isolation: the length of the half of mz_window (i.e. true sampling window should be (mz_center - mz_isolation: mz_center + mz_isolation))
• duration: length of collecting (rt_end - rt_start)
• rt_start: start of rt
• rt_end: end of rt
• intensity: intensity of the apex feature
• apex_rt: the retention time of the apex feature
• charge: charge of the apex feature

These numbers are separated by space. Different sampling position are separated by \t. (i.e. rt_end1 and mz_center2 are separated by \t).

Test

• Generate .csv from convert_to_table
• Untar the test data in ./test folder
• Run command line, parameters used for test are as follows

Apex mode

python3 path_finder.py apex test/Blank_to_Sample_mrgd.csv test/path_5_apex.txt 1e5 3 5 -intensity_accu 1e5 -isolation 1 -delay 0.2 -min_scan 0.2 -max_scan 3

Baseline Mode

python3 path_finder.py baseline test/Blank_to_Sample_mrgd.csv test/path_5_baseline.txt 1e5 3 5 -win_len 0.5 -isolation 1 -delay 0.2

Curve Mode

Input raw feature can be in .mzTab format or .mzML format

python3 path_finder.py curve test/Blank_to_Sample_mrgd.csv test/path_5_curve.txt 1e5 3 5 -infile_raw test/Sample.mzTab -intensity_accu 1e5 -restriction 2 0.2 -isolation 1 -delay 0.2 -min_scan 0.2 -max_scan 3 -cluster kNN

or

python3 path_finder.py curve test/Blank_to_Sample_mrgd.csv test/path_5_curve.txt 1e5 3 5 -infile_raw test/Sample.mzML -intensity_accu 1e5 -restriction 2 0.2 -isolation 1 -delay 0.2 -min_scan 0.2 -max_scan 3 -cluster kNN

TOPPAS Run (optional)

.mzML file can directly parsed by the program (curve method).

If you want to use .mzTab:

ExecutePipeline.exe -in MS1mzTab.toppas -out_dir ./total_ion_curr/data/MS1