IdPP - Identification Probability and Precision

Code repository for work on characterizing the relationship between identification probability and precision in reference-free metabolomics.

• Dylan Ross ([email protected])
• doi?
• license
• disclaimer

Notebooks and scripts

Jupyter notebooks and Python scripts with code for assembling and cleaning the IdPP database, in addition to performing identification probability analysis and some assorted benchmarks/tests.

NOTE: Many of these scripts and notebooks rely on additional data that is not present in this repository, and thus it unlikely that they will be capable of being sucessfully run after simply downloading from here. The primary purpose of these scripts and notebooks is to serve as documentation of the various analyses performed for this work (though some may be functional or can be made so with varying levels of effort).

Composite MS/MS spectrum tests

• match tests with composite spectra
• plotting for match test results
• analysis of individual and combined spectra by PCA

MS/MS benchmarking

• setup databases for benchmarking
• benchmark script

IdPP database setup and expansion

• perform a fresh build of the IdPP database
• clean the database
• expand database with predicted values

Identification probability analysis

• figure depicting different search tolerances and instrument capabilities
• summarize database coverage and contents
• identification probability analysis on original database
• identification probability analysis on expanded database
• expanded database analysis with more caching of intermediate results

Documentation

Documentation for selected components of the IdPP codebase (i.e. the idpp Python package)

current idpp package version: 0.12.24

IdPP database

• builder
• interface

Utilities

• MS2 spectrum similarity
• identification probability tree data structures

Related Repositories

• molecular property prediction workflow
• retention time prediction model