A Python Package for Mass Spectrometry

Developed by Akshay Jay.
Use case: create an mzml_repo object. After providing a Zenodo database ID, the object stores the name, URL, and size of all files in the database.

Requesting a scan from the file will provide the scan's mz and intensity values along with other metadata. If not already in memory, calling get_scan() will store the byte offsets of that scan and all scans that come after it.

How to Use the Class

• all_files: dictionary storing the name (key) and size (value) of each file in the database.
• all_scans: dictionary of tuples where each key is a file name and the value is a tuple containing:
  1. scan_dict: a dictionary with keys as scan numbers and values as byte offsets
  2. first_scan: the first scan number in the file
  3. max_scan: the last scan number in the file
  4. file_url: the file's URL
  5. file_size: the file's size in bytes

E.g. to access the scan_dict for a file, use all_scans[file_name][0].
The populate_all_scans() method populates the scan_dict for a given file.
Toggle between partial and full indexing via the partial_indexing attribute.

get_scan()

The get_scan() method retrieves a specific scan from the file and returns it as a dictionary.
Data structure returned by get_scan() is a dictionary with keys:

• 'mz': m/z values (numpy array)
• 'intensities': normalized intensity values (numpy array)
• 'rt time': retention time (float or 'N/A')
• 'charge': charge state (int or 'N/A')
• 'collision energy': collision energy (float or 'N/A')
• 'ms level': MS level (int or 'N/A')
• 'precursor mz': Precursor m/z value (float or 'N/A')

get_scan() depends on populate_all_scans() since it retrieves the desired scan's byte offset from all_scans.

Valid Formats

The class is able to parse the following idRef formats:

• idRef="SPECTRUM_XXXX"
• idRef="controllerType=X controllerNumber=X scan=XXXX"

Example Usage

The following example pulls from this Zenodo database: https://zenodo.org/records/7824517
Uses this file: subset_dq_00086_11cell_90min_hrMS2_A9.mzML
And this scan number: 16799

import zenodo_mzml_repo

database = 7824517
test_repo = zenodo_mzml_repo.mzml_repo(database)
test_repo.partial_indexing = True
file_name = 'subset_dq_00086_11cell_90min_hrMS2_A9.mzML'
scan1 = test_repo.get_scan(file_name, '16799') # Scan number can be str or int

print("Scan 1's retention time: " + str(scan1['rt time']))
print("Scan 1's charge: " + str(scan1['charge']))
print("Scan 1's collision energy: " + str(scan1['collision energy']))
print("Scan 1's MS level: " + str(scan1['ms level']))
print("Scan 1's precursor m/z: " + str(scan1['precursor_mz']))
print("First 10 values of scan 1's m/z array: " + str(scan1['m/z array'][:10]))
print("First 10 values of scan 1's intensity array: " + str(scan1['intensity array'][:10]))

This code should produce the following:

Available files:
subset_dq_00086_11cell_90min_hrMS2_A9.mzML (1.90 MB)
subset_dq_00084_11cell_90min_hrMS2_A5.mzML (1.47 MB)
subset_dq_00087_11cell_90min_hrMS2_A11.mzML (1.77 MB)
Scan 16799 not found in all_scans for subset_dq_00086_11cell_90min_hrMS2_A9.mzML.
Populating all scans...
First scan: 3381, Max scan: 38682
Downloaded scan 16799
Time taken to retrieve scan 16799: 2.68 seconds
Scan 1's retention time: 2484.495791584002
Scan 1's charge: 4
Scan 1's collision energy: 35.0
Scan 1's MS level: 2
Scan 1's precursor m/z: 508.044647216797
First 10 values of scan 1's m/z array: [100.10274506 100.51384735 101.07068634 101.09996796 102.05473328
 102.21772003 104.05259705 110.06845856 110.07099915 112.08670044]
First 10 values of scan 1's intensity array: [  876.2010498    811.77313232  4443.26025391  2841.00219727
  1590.89794922   841.73175049  1415.2277832   1670.71008301
 29214.15820312  1161.09606934]

Runtime

Based on 5 tests from 4 Zenodo databases.
Using a wired connection with ~100Mbps.
Average file size: 55 MB

First Scan in a File

• Average runtime: 12.08 seconds
• Median runtime: 10.29 seconds

All Successive Scans in the File

• Average Runtime: 0.78 seconds
• Median Runtime: 0.83 seconds
  (Data table coming soon)