Step Field CCS Automation Package

This python package aims to find reliable features and compute the CCS values for chemical standards. It is assumed that prior to running this software you have:

• preprocessed the raw data files and each of the seven fields is split into a separate file (https://omics.pnl.gov/software/pnnl-preprocessor)
• obtained the correspondent features from Agilent's MP software (a CEF file from each field)

How to install python packages for this app.

pip install -r requirements.txt

How to use the app. (python ccs_auto_app.py –help)

usage: ccs_auto_app.py [-h] [--target_list_file TARGET_LIST_FILE]
                       [--config_file CONFIG_FILE] [--data_folder DATA_FOLDER]
                       [--output OUTPUT]
 
optional arguments:
  -h, --help            show this help message and exit
  --target_list_file TARGET_LIST_FILE
                        Target list file (Tab-delimited text format)
  --config_file CONFIG_FILE
                        Configuration file
  --data_folder DATA_FOLDER
                        Data folder containing all the cef and meta data files
  --output OUTPUT       Output file to save a output table

Demo

You can test this package with the CEF files here included. This is a subset of the standards from the PNNL CCS DB found at https://panomics.pnnl.gov/metabolites/

How to run

python ccs_auto_app.py --target_list_file demo/TargetList.tsv --config_file demo/config.xml --data_folder demo/data

Output files

Please refer to the demo/results folder.

1. ccs_table.tsv

Running this package provides a summary of all about computing CCS values of compounds listed on a target list file (e.g., TargetList.tsv). Each row mean a CCS computation results of a single replicate. And each column means the following items.

Column Names	Descriptions
Compound_id	Compound ID. It will be the same to CompID in the target list file
name	Compound Name. It will be the same to NeutralName in the target list file
Ionization	Ionization. It will also be the same to Ionization in the target list file
adduct	adduct
mass	mass for the adduct form
ccs_avg	an average of CCS values of same adducts for all the replicates
ccs_rsd	a relative stdv of CCS values of same adducts for all the replicates
ccs	a CCS value of this adduct for this replicate
intensity_org_#	original intensity of a corresponding feature in #th field
intensity_z_#	z-score of an intensity of a corresponding feature in #th field
mass_error_#	mass error of a corresponding feature in #th field (ppm)
num_features	number of features found in all step fields
intercept	intercept of a CCS regression line
slope	slope of a CCS regression line
k0	K0 computed from a CCS regression line
p_value	p value of a CCS regression line
r_value	R value of a CCS regression line
replicate	original file for this CCS computation (.d file)

2. <compound_id>.pdf (e.g., S0000001_neg_[M-H].pdf)

3. <compound_id>__intensity_dist.pdf (e.g., S0000001_neg_intensity_dist.pdf)

• x-axis: log-scale of original intensities

• y-axis: kde estimation (normalized distribution)

• dots: features within 15ppm (according to mz_tolerance setting in config.xml) of (e.g., [M-H] in the below)

• Vertical lines

  • Most left line: median of feature intensities
  • Center line: 10 * median
  • Most right line: 2 * std

4. <compound_id>__meta.pdf (e.g., S0000001_neg_meta.pdf)

In the figures, dots represent values in the frames of the selected ranges (frame_offset in config.xml) to compute CCS values.