Solving CASMI 2013 with MetFrag, MetFusion and MOLGEN-MS/MS

Area of Science:

• Analytical Chemistry
• Computational Chemistry
• Biochemistry

Background:

• The Critical Assessment of Small Molecule Identification (CASMI) contest evaluates computational methods for identifying chemical structures.
• Automated workflows are increasingly important for high-throughput analysis in chemistry and biology.

Purpose of the Study:

• To report the performance of an automated workflow for small molecule identification in the CASMI 2013 contest.
• To assess the utility of computational tools like MOLGEN-MS/MS, MetFrag, and MetFusion for molecular formula and structure determination.
• To inform the development of decision-making criteria for automated, high-throughput unknown compound identification.

Main Methods:

• Utilized MOLGEN-MS/MS for molecular formula calculation (Category 1).
• Employed MetFrag and MetFusion for structure identification (Category 2), querying KEGG, PubChem, and ChemSpider databases.
• Integrated Category 1 results to guide Category 2 searches (formula vs. exact mass).

Main Results:

• Category 2 structure identification yielded impressive results given the automated approach and database size.
• Category 1 molecular formula calculation was impacted by large m/z and ppm values in the challenge data.
• Automated methods, while effective, did not outperform the manual approach of the contest winner.

Conclusions:

• The automated workflow provided valuable insights for developing decision-making criteria in high-throughput unknown identification.
• The CASMI 2013 experience highlighted areas for improvement in automated small molecule identification strategies for future contests and applications.