Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data1
Galaxy Pipeline & Tool Development for
Processing Gas Chromatography – Mass
Spectrometry Data
14.09.2021
Helge Hecht
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data2
Overview
1. Introduction
2. State Of The Art
3. Problem Statement
4. Methods
5. Results
6. Summary
7. Future Work
8. Acknowledgements
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data3
Introduction
„This is presented as a call to the
international environmental health
research community to champion this
effort and work together in this common
goal.“ (10.1016/j.toxrep.2015.11.009)
„... have facilitated the detection of
tens of thousands of ions, metabolite
identification remains one of the
biggest challenges of available
analytical methods.“
(10.1021/acs.chemrestox.6b00179)
„...(iii) the lack of automation of the
annotation/identification process.“
(10.1016/j.envint.2021.106630)
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data4
State of the Art
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data5
State of the Art – GUI-based Tools
Good
̶ easy to use
̶ work well as standalone tools
Bad
̶ not scalable → no distributed
computation
̶ tight coupling of GUI and backend
̶ bad library support → programming
overhead
̶ mostly focus on LC-MS
MZmine2 (10.1186/1471-2105-11-395)
MS-DIAL
(10.1038/n
meth.3393)
AMDIS (10.1016/S1044-
0305(99)00047-1)
MS-FINDER
(10.1021/acs.a
nalchem.6b007
70)
SIRIUS
(10.1093/bi
oinformatic
s/btn603)
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data6
State of the Art – Web-based tools
XCMS Online (10.1021/ac300698c)
GNPS (10.1038/nbt.3597)
MetaboAnalyst
(10.1093/nar/gka
b382)
Good
̶ easy to use
̶ partially scalable
Bad
̶ data storage → sensitive data?
̶ difficult to modify individual steps
̶ little resources for GC-MS
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data7
State of the Art – Coded Workflows
Good
̶ fairly easy to extend & modify
̶ scalable
̶ good library support
Bad
̶ varying (often poor) quality
̶ hard to use
̶ low reproducibility
̶ poorly integrated → group specific
Bioconductor (10.1093/nar/gkab382)
RforMassSpectrometry
Bioconda (10.1038/s41592-018-0046-7)
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data8
State of the Art – Galaxy
Good
̶ easy to use
̶ scalable & modular
̶ data management
Bad
̶ focus on LC-MS
̶ varying tool quality
̶ different application domain
PhenoMeNal (10.1093/gigascience/giy149)
W4M (10.1093/bioinformatics/btu813;
10.1016/j.biocel.2017.07.002)
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data9
Problem Statement
We need data processing pipelines that are
1. easy to use, understand & access,
2. built for large-scale analysis,
3. including various modules & steps,
4. creating reproducible results,
consisting of tools that are
1. well tested & documented,
2. easy to extend & modify based on requirements,
3. specific to the research domain.
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data10
Methods
We implement Galaxy pipelines using new tools
that are
1. tailored for user needs & domain problems,
2. developed open-source,
3. according to professional software standards
& State-of-the-Art packages with modifications to
1. test their behaviour,
2. make them easier to use & understand,
3. make them scalable.
Good
̶ best of all worlds
̶ long-term solution
̶ links to other
infrastructures
Bad
̶ hard to achieve
̶ high complexity
̶ expensive
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data11
Methods
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data12
Methods – Galaxy Tool Development
̶ virtualization → docker & Biocontainers
(10.1093/bioinformatics/btx192)
̶ open-source hosting via GitHub
̶ testing (testthat; pytest) with widely supported
frameworks & code coverage
̶ static code analysis (sonarcloud)
̶ tools according to IUC guidelines
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data13
galaxy workflow for
GC-MS processing
tools that are
complementary to
existing resources
two standalone tools
extracted as extensible
modules
contribution to existing
open-source software
Results
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data14
Results – Galaxy Workflow
Galaxy (cerit-sc.cz)
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data15
RIAssigner
Results - Tools
pyMSPannotator
̶ read & write data in various formats
(csv & msp) using matchms &
pandas
̶ extensible data & computation
modules
̶ published via Bioconda
̶ makes data comparable by aligning
based on high- confidence
annotations
̶ add various metadata fields to mass
spectral libraries
̶ extends functionality of webchem to
python
̶ leverages IDSM (10.1186/s13321-
021-00515-1) service for PubChem
→ query via API
̶ first step of improved high-resolution
filtering workflow
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data16
Results – Capacity Building
̶ participation in Galaxy Metabolomics Community calls & member
in ELIXIR Metabolomics Community
̶ participation in de.nbi network events (metaRbolomics)
̶ Netherlands metabolomics infrastructure (eScienceCenter →
matchms)
̶ contributions to Galaxy Training Network
̶ member in US Thermo GC Orbitrap working group, BP4NTA,
mQACC
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data17
Summary
̶ strong need for automation & harmonization
in data processing
̶ state-of-the-art tools are scattered
̶ lack of high-quality resources
̶ Galaxy as platform for harmonization of
large scale analysis
̶ rapid progress (compared to others!)
̶ high quality developments take time
̶ potential for publication of workflow & tools
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data18
Future Work
̶ integration of complementary tools → ADAP-GC4.0
(10.1021/acs.analchem.9b01424), NormAE
(10.1021/acs.analchem.9b05460) etc.
̶ additional steps → reporting (10.1021/acs.analchem.8b04310,
biotransformation, prediction of RI (10.1016/j.aca.2020.12.043)
̶ applying machine learning techniques
̶ experimenting with similarity scores (10.1371/journal.pcbi.1008724;
10.1101/2021.04.18.440324) & molecular networking
̶ workflow for improved high-resolution filtering
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data19
Acknowledgements
Martin Čech
Elliott James Price
Aleš Křenek
Jiří Novotný
Maksym Skoryk
Matej Troják
Ondřej Melichar
Gabriela Karásková
Vojtěch Bartoň
Jana Klánová
Muhammad Usman
Zdenka Dudová
Karolína Trachtová
RECETOX research infrastructure
supported by the Ministry of
Education, Youth and Sports of the
Czech Republic (LM2018121) and
funding from the CETOCOEN
EXCELLENCE Teaming 2 project
supported by Horizon
2020 (857560) and the Ministry of
Education, Youth and Sports of the
Czech Republic
(02.1.01/0.0/0.0/18_046/0015975).
Computational resources provided
by the e-INFRA CZ (LM2018140).
Helge Hecht - Galaxy Pipeline & Tool Development for Processing Gas Chromatography - Mass Spectrometry Data21
Thank you for your attention!
Questions?