Mercurial > repos > prog > lcmsmatching
view README.md @ 5:fb9c0409d85c draft
planemo upload for repository https://github.com/workflow4metabolomics/lcmsmatching.git commit 608d9e59a0d2dcf85a037968ddb2c61137fb9bce
author | prog |
---|---|
date | Wed, 19 Apr 2017 10:00:05 -0400 |
parents | |
children | f86fec07f392 |
line wrap: on
line source
LC/MS matching ============== [![Build Status](https://travis-ci.org/workflow4metabolomics/lcmsmatching.svg?branch=master)](https://travis-ci.org/workflow4metabolomics/lcmsmatching) An LC/MS matching tool for [Galaxy](https://galaxyproject.org/), part of the [Workflow4Metabolomics](http://workflow4metabolomics.org/) project, and developed during the [MetaboHUB](http://www.metabohub.fr/en) project. The two matching algorithms used in this tool have been imported from developments made at [CEA](http://www.cea.fr/english) Saclay, inside the *DSV/IBITEC-S/SPI*. They have been translated from C# to R. For more information, see the galaxy tool page, help section, available inside `galaxy/lcmsmatching.xml`. ## search-mz This is the script, included in this repository, that allows run on command line an MZ matching on one of the available database types. Please run `search-mz -h` for a help page listing all options and presenting some examples. ## Dependencies * `libssl-dev`. * `libcurl4-openssl-dev`. * `libxml2-dev`. * `R` version `3.2.2`. * `R` packages: - `getopt` >= `1.20.0`. - `stringr` >= `1.0.0`. - `plyr` >= `1.8.3`. - `XML` >= `3.98`. - `bitops` >= `1.0_6`. - `RCurl` >= `1.95`. - `jsonlite` >= `1.1`. ## Updates ### 3.3.1 * Correct a bug while trying to connect to Peakforest for getting the list of chromatographic columns. ### 3.3.0 * The file database (in-house) field names are now presented in individual choice lists instead of a single text box where you had to insert a very long keys/values string. * The tool now tries to guess the names of the file database fields, the values of the MS mode column, and the names of the input file columns. * Allows to select the unit (minutes or seconds) of retention time values inside the input file, but also inside the file database (in-house).