FingerID

Version: 1.4

Website

https://github.com/icdishb/fingerid

Description

FingerID can be used for metabolite identification of MS/MS data using kernel methods. It used SVM to first predict a large set of molecular properties for matching against large databases (e.g. PubChem, MassBank, METLIN) for metabolite identification. This is an alternative technique to the traditional method of directly mapping the metabolite to the spectrum. Two types of kernel are used: integral mass accuracy kernel and high mass accuracy kernel. The molecular properties used are: topology, physio-chemical and electrical. A probabilistic score from the predicted molecular fingerprints is used to retrieve candidate molecules from metabolite databases.

Functionality

• Annotation/MS/Level 2a - Library Spectrum Match

Instrument Data Type

• MS/LC-MS

Approaches

Computer Skills

Advanced

Software Type

Web App

Interface

Web user interface

Operating System (OS)

• Unix/Linux
• Mac OS
• Windows

Language

MATLAB, Python

Dependencies

Python ≥ 2.7, Numpy ≥ 1.4.0, LibSVM 3.17 python interface

Input Formats - Open

Peak list with m/z and RT

Input Formats - Proprietary

N/A

Published

2012

Last Updated

2016

License

Apache 2.0

Paper

http://www.ncbi.nlm.nih.gov/pubmed/22815355

PMID

22815355