The fourth round of the Critical Assessment of Small Molecule Identification (CASMI) Contest ( www.casmi-contest.org ) was held in 2016, with two new categories for automated methods. This article covers the 208 challenges in Categories 2 and 3, without and with metadata, from organization, participation, results and post-contest evaluation of CASMI 2016 through to perspectives for future contests and small molecule annotation/identification. The Input Output Kernel Regression (CSI:IOKR) machine learning approach performed best in “Category 2: Best Automatic Structural Identification—In Silico Fragmentation Only”, won by Team Brouard with 41% challenge wins. The winner of “Category 3: Best Automatic Structural Identification—Full Information” was Team Kind (MS-FINDER), with 76% challenge wins. The best methods were able to achieve over 30% Top 1 ranks in Category 2, with all methods ranking the correct candidate in the Top 10 in around 50% of challenges. This success rate rose to 70% Top 1 ranks in Category 3, with candidates in the Top 10 in over 80% of the challenges. The machine learning and chemistry-based approaches are shown to perform in complementary ways. The improvement in (semi-)automated fragmentation methods for small molecule identification has been substantial. The achieved high rates of correct candidates in the Top 1 and Top 10, despite large candidate numbers, open up great possibilities for high-throughput annotation of untargeted analysis for “known unknowns”. As more high quality training data becomes available, the improvements in machine learning methods will likely continue, but the alternative approaches still provide valuable complementary information. Improved integration of experimental context will also improve identification success further for “real life” annotations. The true “unknown unknowns” remain to be evaluated in future CASMI contests.

A new lipid class named ‘fatty acid esters of hydroxyl fatty acids’ (FAHFA) was recently discovered in mammalian adipose tissue and in blood plasma and some FAHFAs were found to be associated with type 2 diabetes. To facilitate the automatic annotation of FAHFAs in biological specimens, a tandem mass spectra (MS/MS) library is needed. Due to the limitation of the commercial available standard compounds, we proposed building an in silico MS/MS library to extend the coverage of molecules. We developed a computer-generated library with 3267 tandem mass spectra (MS/MS) for 1089 FAHFA species. FAHFA spectra were generated based on authentic standards with negative mode electrospray ionization and 10, 20, and 40 V collision induced dissociation at 4 spectra/s as used in in ultra-high performance liquid chromatography-QTOF mass spectrometry studies. However, positional information of the hydroxyl group is only obtained either at lower QTOF spectra acquisition rates of 1 spectrum/s or at the MS3 level in ion trap instruments. Therefore, an additional set of 4290 fragment-rich MS/MS spectra was created to enable distinguishing positional FAHFA isomers. The library was generated based on ion fragmentations and ion intensities of FAHFA external reference standards, developing a heuristic model for fragmentation rules and extending these rules to large swaths of computer-generated structures of FAHFAs with varying chain lengths, degrees of unsaturation and hydroxyl group positions. Subsequently, we validated the new in silico library by discovering several new FAHFA species in egg yolk, showing that this library enables high-throughput screening of FAHFA lipids in various biological matrices. The developed library and templates are freely available for commercial or noncommercial use at http://fiehnlab.ucdavis.edu/staff/yanma/fahfa-lipid-library . This in silico MS/MS library allows users to annotate FAHFAs from accurate mass tandem mass spectra in an easy and fast manner with NIST MS Search or PepSearch software. The developing template is provided for advanced users to modify the parameters and export customized libraries according to their instrument features.

Mouse knockouts facilitate the study ofgene functions. Often, multiple abnormal phenotypes are induced when a gene is inactivated. The International Mouse Phenotyping Consortium (IMPC) has generated thousands of mouse knockouts and catalogued their phenotype data. We have acquired metabolomics data from 220 plasma samples from 30 unique mouse gene knockouts and corresponding wildtype mice from the IMPC. To acquire comprehensive metabolomics data, we have used liquid chromatography (LC) combined with mass spectrometry (MS) for detecting polar and lipophilic compounds in an untargeted approach. We have also used targeted methods to measure bile acids, steroids and oxylipins. In addition, we have used gas chromatography GC-TOFMS for measuring primary metabolites. The metabolomics dataset reports 832 unique structurally identified metabolites from 124 chemical classes as determined by ChemRICH software. The GCMS and LCMS raw data files, intermediate and finalized data matrices, R-Scripts, annotation databases, and extracted ion chromatograms are provided in this data descriptor. The dataset can be used for subsequent studies to link genetic variants with molecular mechanisms and phenotypes.