Structural Annotation Research Articles

A chromosome-level genome assembly of the soybean pod borer: insights into larval transcriptional response to transgenic soybean expressing the pesticidal Cry1Ac protein

BackgroundGenetically modified (GM) crop plants with transgenic expression of Bacillus thuringiensis (Bt) pesticidal proteins are used to manage feeding damage by pest insects. The durability of this technology is threatened by the selection for resistance in pest populations. The molecular mechanism(s) involved in insect physiological response or evolution of resistance to Bt is not fully understood.ResultsTo investigate the response of a susceptible target insect to Bt, the soybean pod borer, Leguminivora glycinivorella (Lepidoptera: Tortricidae), was exposed to soybean, Glycine max, expressing Cry1Ac pesticidal protein or the non-transgenic parental cultivar. Assessment of larval changes in gene expression was facilitated by a third-generation sequenced and scaffolded chromosome-level assembly of the L. glycinivorella genome (657.4 Mb; 27 autosomes + Z chromosome), and subsequent structural annotation of 18,197 RefSeq gene models encoding 23,735 putative mRNA transcripts. Exposure of L. glycinivorella larvae to transgenic Cry1Ac G. max resulted in prediction of significant differential gene expression for 204 gene models (64 up- and 140 down-regulated) and differential splicing among isoforms for 10 genes compared to unexposed cohorts. Differentially expressed genes (DEGs) included putative peritrophic membrane constituents, orthologs of Bt receptor-encoding genes previously linked or associated with Bt resistance, and those involved in stress responses. Putative functional Gene Ontology (GO) annotations assigned to DEGs were significantly enriched for 36 categories at GO level 2, respectively. Most significantly enriched cellular component (CC), biological process (BP), and molecular function (MF) categories corresponded to vacuolar and microbody, transport and metabolic processes, and binding and reductase activities. The DEGs in enriched GO categories were biased for those that were down-regulated (≥ 0.783), with only MF categories GTPase and iron binding activities were bias for up-regulation genes.ConclusionsThis study provides insights into pathways and processes involved larval response to Bt intoxication, which may inform future unbiased investigations into mechanisms of resistance that show no evidence of alteration in midgut receptors.

Mirror, mirror on the wall, which phytochemicals in Clinacanthus nutans inhibits advanced glycation end products of them all?

In our quest to discover advanced glycation end products (AGEs) inhibitors from Clinacanthus nutans (Burm.f.) Lindau leaves, we conducted a bioactivity-based molecular networking. This approach integrates LC-MS2 profiling and in vitro antiglycation data to predict bioactive compounds. We began by screening three extracts: 100% ethanol, 70% ethanol and 100% water alongside the in vitro antioxidant activity, total phenolics content (TPC) and schaftoside content. Among these extracts, 100% ethanol extract exhibited the highest total AGEs inhibition effects (IC50 = 80.18 ± 11.6 μg/mL), DPPH scavenging activity (IC50 = 747.40 ± 10.30 μg/mL) and TPC (26.54 ± 2.09 μg GAE /mg extract). Intriguingly, 100% ethanol extract contained the lowest amount of schaftoside, suggesting the involvement of other phytochemicals in the antiglycation effects. The molecular networking and in silico structural annotations of 401 LC-MS features detected in the fractions from 100% ethanol extract predicted 21 bioactive compounds (p < 0.05, r > 0.90), including several C40 carotenoids, alkaloids containing tetrapyrrole structures and fatty acids. On the contrary, all phenolics showed weak correlations with antiglycation effects. These predictions were further validated in vitro, where carotenoid lutein showed half maximal inhibitory concentration, IC50 = 96 ± 8 μM and selected flavonoid-C-glycosides exhibited weaker inhibitions (IC50 between 568 and 1922 μM). Notably, lutein content was higher in freeze-dried leaves (12.42 ± 0.82 mg/100 g) than oven-dried, although the former was associated with elevated mercury levels. In summary, C. nutans exhibited potential antiglycation and antioxidant activity, and lutein was identified as the main bioactive principle.

Improved selection of canonical proteins for reference proteomes.

The 'canonical' protein sets distributed by UniProt are widely used for similarity searching, and functional and structural annotation. For many investigators, canonical sequences are the only version of a protein examined. However, higher eukaryotes often encode multiple isoforms of a protein from a single gene. For unreviewed (UniProtKB/TrEMBL) protein sequences, the longest sequence in a Gene-Centric group is chosen as canonical. This choice can create inconsistencies, selecting >95% identical orthologs with dramatically different lengths, which is biologically unlikely. We describe the ortho2tree pipeline, which examines Reference Proteome canonical and isoform sequences from sets of orthologous proteins, builds multiple alignments, constructs gap-distance trees, and identifies low-cost clades of isoforms with similar lengths. After examining140 000 proteins from eight mammals in UniProtKB release 2022_05, ortho2tree proposed 7804 canonical changes for release 2023_01, while confirming 53 434 canonicals. Gap distributions for isoforms selected by ortho2tree are similar to those in bacterial and yeast alignments, organisms unaffected by isoform selection, suggesting ortho2tree canonicals more accurately reflect genuine biological variation. 82% of ortho2tree proposed-changes agreed with MANE; for confirmed canonicals, 92% agreed with MANE. Ortho2tree can improve canonical assignment among orthologous sequences that are >60% identical, a group that includes vertebrates and higher plants.

Genomics and Phylogeny of Rhodotorula glutinis and Rhodotorula kratochvilovae Isolated from the Northern Peruvian Andes

Genomes of oleaginous yeast strains Rhodotorula glutinis CON-5 and Rhodotorula kratochvilovae POR-3, isolated from areas in the northern Peruvian Andes using SPAdes, were sequenced and assembled applying Illumina and de novo. Genomes of 20,515,696 and 20,738,185 bp, respectively, were determined. From the structural and functional annotations, the Basidiomycota phylum showed a similarity of 76.8% and 86.5% with 6,976 and 8,124 pairs of proteins in both yeasts respectively, with homologues in the UniProt data bank. Using OrthoVenn, a relationship between both yeasts was obtained from 450 orthologous groups. Likewise, the above-mentioned yeasts and R. toruloides (oleaginous Basidiomycota) showed 1,574 orthologous groups, indicating a good relationship. Construction of phylogenetic trees of genes encoding metabolic enzymes was also carried out, based on the ITS sequences which showed that CON-5 and POR-3 have a greater relationship with R. graminis. Their phylogenetic relationship was ascertained and determined that the enzymes involved in the metabolism of CON-5 and POR-3 are related to each other. It was also found that the protein sequences of the Basidiomycota phylum differ from Ascomycota. The study showed functional evidence regarding the lipid accumulation phenotype, an important aspect in the context of obtaining lipids or oleochemicals.

Predicting Protein Functions Based on Heterogeneous Graph Attention Technique.

In bioinformatics, protein function prediction stands as a fundamental area of research and plays a crucial role in addressing various biological challenges, such as the identification of potential targets for drug discovery and the elucidation of disease mechanisms. However, known functional annotation databases usually provide positive experimental annotations that proteins carry out a given function, and rarely record negative experimental annotations that proteins do not carry out a given function. Therefore, existing computational methods based on deep learning models focus on these positive annotations for prediction and ignore these scarce but informative negative annotations, leading to an underestimation of precision. To address this issue, we introduce a deep learning method that utilizes a heterogeneous graph attention technique. The method first constructs a heterogeneous graph that covers the protein-protein interaction network, ontology structure, and positive and negative annotation information. Then, it learns embedding representations of proteins and ontology terms by using the heterogeneous graph attention technique. Finally, it leverages these learned representations to reconstruct the positive protein-term associations and score unobserved functional annotations. It can enhance the predictive performance by incorporating these known limited negative annotations into the constructed heterogeneous graph. Experimental results on three species (i.e., Human, Mouse, and Arabidopsis) demonstrate that our method can achieve better performance in predicting new protein annotations than state-of-the-art methods.

PyCoM: a python library for large-scale analysis of residue-residue coevolution data.

Computational methods to detect correlated amino acid positions in proteins have become a valuable tool to predict intra- and inter-residue protein contacts, protein structures, and effects of mutation on protein stability and function. While there are many tools and webservers to compute coevolution scoring matrices, there is no central repository of alignments and coevolution matrices for large-scale studies and pattern detection leveraging on biological and structural annotations already available in UniProt. We present a Python library, PyCoM, which enables users to query and analyze coevolution matrices and sequence alignments of 457622 proteins, selected from UniProtKB/Swiss-Prot database (length ≤ 500 residues), from a precompiled coevolution matrix database (PyCoMdb). PyCoM facilitates the development of statistical analyses of residue coevolution patterns using filters on biological and structural annotations from UniProtKB/Swiss-Prot, with simple access to PyCoMdb for both novice and advanced users, supporting Jupyter Notebooks, Python scripts, and a web API access. The resource is open source and will help in generating data-driven computational models and methods to study and understand protein structures, stability, function, and design. PyCoM code is freely available from https://github.com/scdantu/pycom and PyCoMdb and the Jupyter Notebook tutorials are freely available from https://pycom.brunel.ac.uk.

Prot2Text: Multimodal Protein’s Function Generation with GNNs and Transformers

In recent years, significant progress has been made in the field of protein function prediction with the development of various machine-learning approaches. However, most existing methods formulate the task as a multi-classification problem, i.e. assigning predefined labels to proteins. In this work, we propose a novel approach, Prot2Text, which predicts a protein's function in a free text style, moving beyond the conventional binary or categorical classifications. By combining Graph Neural Networks(GNNs) and Large Language Models(LLMs), in an encoder-decoder framework, our model effectively integrates diverse data types including protein sequence, structure, and textual annotation and description. This multimodal approach allows for a holistic representation of proteins' functions, enabling the generation of detailed and accurate functional descriptions. To evaluate our model, we extracted a multimodal protein dataset from SwissProt, and demonstrate empirically the effectiveness of Prot2Text. These results highlight the transformative impact of multimodal models, specifically the fusion of GNNs and LLMs, empowering researchers with powerful tools for more accurate function prediction of existing as well as first-to-see proteins.

Classifying protein kinase conformations with machine learning.

Protein kinases are key actors of signaling networks and important drug targets. They cycle between active and inactive conformations, distinguished by a few elements within the catalytic domain. One is the activation loop, whose conserved DFG motif can occupy DFG-in, DFG-out, and some rarer conformations. Annotation and classification of the structural kinome are important, as different conformations can be targeted by different inhibitors and activators. Valuable resources exist; however, large-scale applications will benefit from increased automation and interpretability of structural annotation. Interpretable machine learning models are described for this purpose, based on ensembles of decision trees. To train them, a set of catalytic domain sequences and structures was collected, somewhat larger and more diverse than existing resources. The structures were clustered based on the DFG conformation and manually annotated. They were then used as training input. Two main models were constructed, which distinguished active/inactive and in/out/other DFG conformations. They considered initially 1692 structural variables, spanning the whole catalytic domain, then identified ("learned") a small subset that sufficed for accurate classification. The first model correctly labeled all but 3 of 3289 structures as active or inactive, while the second assigned the correct DFG label to all but 17 of 8826 structures. The most potent classifying variables were all related to well-known structural elements in or near the activation loop and their ranking gives insights into the conformational preferences. The models were used to automatically annotate 3850 kinase structures predicted recently with the Alphafold2 tool, showing that Alphafold2 reproduced the active/inactive but not the DFG-in proportions seen in the Protein Data Bank. We expect the models will be useful for understanding and engineering kinases.

Introduction of a human- and keyboard-friendly N-glycan nomenclature.

In the beginning was the word. But there were no words for N-glycans, at least, no simple words. Next to chemical formulas, the IUPAC code can be regarded as the best, most reliable and yet immediately comprehensible annotation of oligosaccharide structures of any type from any source. When it comes to N-glycans, the venerable IUPAC code has, however, been widely supplanted by highly simplified terms for N-glycans that count the number of antennae or certain components such as galactoses, sialic acids and fucoses and give only limited room for exact structure description. The highly illustrative - and fortunately now standardized - cartoon depictions gained much ground during the last years. By their very nature, cartoons can neither be written nor spoken. The underlying machine codes (e.g., GlycoCT, WURCS) are definitely not intended for direct use in human communication. So, one might feel the need for a simple, yet intelligible and precise system for alphanumeric descriptions of the hundreds and thousands of N-glycan structures. Here, we present a system that describes N-glycans by defining their terminal elements. To minimize redundancy and length of terms, the common elements of N-glycans are taken as granted. The preset reading order facilitates definition of positional isomers. The combination with elements of the condensed IUPAC code allows to describe even rather complex structural elements. Thus, this "proglycan" coding could be the missing link between drawn structures and software-oriented representations of N-glycan structures. On top, it may greatly facilitate keyboard-based mining for glycan substructures in glycan repositories.

De novo assembly and annotation of Popillia japonica’s genome with initial clues to its potential as an invasive pest

BackgroundThe spread of Popillia japonica in non-native areas (USA, Canada, the Azores islands, Italy and Switzerland) poses a significant threat to agriculture and horticulture, as well as to endemic floral biodiversity, entailing that appropriate control measures must be taken to reduce its density and limit its further spread. In this context, the availability of a high quality genomic sequence for the species is liable to foster basic research on the ecology and evolution of the species, as well as on possible biotechnologically-oriented and genetically-informed control measures.ResultsThe genomic sequence presented and described here is an improvement with respect to the available draft sequence in terms of completeness and contiguity, and includes structural and functional annotations. A comparative analysis of gene families of interest, related to the species ecology and potential for polyphagy and adaptability, revealed a contraction of gustatory receptor genes and a paralogous expansion of some subgroups/subfamilies of odorant receptors, ionotropic receptors and cytochrome P450s.ConclusionsThe new genomic sequence as well as the comparative analyses data may provide a clue to explain the staggering invasive potential of the species and may serve to identify targets for potential biotechnological applications aimed at its control.

Direct infusion–multiple reaction monitoring cubed (DI–MRM3) enables widely targeted bi-omics of Colla Corii Asini (Chinese name: Ejiao)

Food authenticity is extremely important and widely targeted bi-omics is a promising pipeline attributing to incorporating metabolomics and peptidomics. Colla Corii Asini (CCA, Ejiao) is one of the most popular tonic edible materials, with counterfeit and adulterated products being widespread. An attempt was devoted to develop a high-throughput and reliable DI–MRM3 program facilitating widely targeted bi-omics of CCA. Firstly, predictive MRM program captured metabolites and peptides in trypsin-digestive gelatins. After data alignment and structure annotation, primary parameters such as Q1 → Q3 → QLIT, CE, and EE were optimized for all 17 metabolites and 34 peptides by online ER-MS. Though a single run merely consumed 6.5 min, great selectivity was reached for each analyte. Statistical results showed that nine peptides contributed to distinguish CCA from other gelatins. After cross-validation with LC–MRM, DI–MRM3 was justified to be reproducible and high-throughput for widely targeted bi-omics of CCA, suggesting a meaningful tool for food authenticity.

Machine learning-based prediction of proteins’ architecture using sequences of amino acids and structural alphabets

In addition to the growth of protein structures generated through wet laboratory experiments and deposited in the PDB repository, AlphaFold predictions have significantly contributed to the creation of a much larger database of protein structures. Annotating such a vast number of structures has become an increasingly challenging task. CATH is widely recognized as one the most common platforms for addressing this challenge, as it classifies proteins based on their structural and evolutionary relationships, offering the scientific community an invaluable resource for uncovering various properties, including functional annotations. While CATH annotation involves – to some extent – human intervention, keeping up with the classification of the rapidly expanding repositories of protein structures has become exceedingly difficult. Therefore, there is a pressing need for a fully automated approach. On the other hand, the abundance of protein sequences stemming from next generation sequencing technologies, lacking structural annotations, presents an additional challenge to the scientific community. Consequently, 'pre-annotating’ protein sequences with structural features, ensuring a high level of precision, could prove highly advantageous. In this paper, after a thorough investigation, we introduce a novel machine-learning model capable of classifying any protein domain, whether it has a known structure or not, into one of the 40 main CATH Architectures. We achieve an F1 Score of 0.92 using only the amino acid sequence and a score of 0.94 using both the sequence of amino acids and the sequence of structural alphabets. Communicated by Ramaswamy H. Sarma

Metabolomics and Molecular Networking-Guided Screening of Bacillus-Derived Bioactive Compounds Against a Highly Lethal Vibrio Species.

Microorganisms are important sources of bioactive natural products. However, the complexity of microbial metabolites and the low abundance of active compounds render the isolation and purification process laborious and inefficient. During our search for active substances capable of inhibiting the newly discovered highly lethal Vibrio strain vp-HL, we found that the fermentation broth of multiple Bacillus strains exhibited antibacterial activity. However, the substances responsible for the activity remained unclear. Metabolomics, molecular networking (MN), and the Structural similarity Network Annotation Platform for Mass Spectrometry (SNAP-MS) were employed in conjunction with bioactivity screening to predict the antibacterial compounds from Bacillus strains. The analysis of fractions, and their isolation, NMR-based annotation, and bioactivity evaluation of an amicoumacin compound partially confirmed the prediction from these statistical analyses. This work presents the potential of marine Bacillus in producing active substances against Vibrio species. Additionally, it highlighted the significance and feasibility of metabolomics and MN in the dereplication of compounds and the determination of isolation targets.

Bioinformatic Sweeties: a unified portal for characterizing human proteins and their variants

Next-generation sequencing techniques provide an unprecedented characterisation of human Variants of Unknown Significance (VUS). Single-residue variations are collected in public databases and associated to diseases and phenotypes. However, for detailing at molecular level mechanisms involved in the onset of diseases, variants need structural and functional annotation. Here we propose a new portal called Bioinformatic Sweeties, collecting resources ranging from databases for human protein annotation to computational methods for predicting impact of variants. The tools, included in the portal, allow computing different protein properties, ranging from solvent accessible surface to stability and interactions and do not require login or installation. The portal, speeding up the variant characterisation process, is available at: https://bioinformaticsweeties.biocomp.unibo.it

The method and implementation of a Taiwan building recognition model based on YOLOX-S and illustration enhancement

Earthquakes pose significant risks in Taiwan, necessitating effective risk assessment and preventive measures to reduce damage. Obtaining complete building structure data is crucial for the accurate evaluation of earthquake-induced losses. However, manual annotation of building structures is time-consuming and inefficient, resulting in incomplete data. To address this, we propose YOLOX-CS, an object detection model, combined with the Convolutional Block Attention Module (CBAM), to enhance recognition capabilities for small structures and reduce background interference. Additionally, we introduce the Illustration Enhancement data augmentation method to improve the recognition of obscured buildings. We collected diverse building images and manually annotated them, resulting in a dataset for training the model. YOLOX-CS with CBAM significantly improves recognition accuracy, particularly for small objects, and Illustration Enhancement enhances the recognition of occluded buildings. Our proposed approach advances building structure recognition, contributing to more effective earthquake risk assessment systems in Taiwan and beyond.

False-positive MDA findings in HRMS-based screening of putrefied postmortem blood samples-Identification of the interference as N-acetyltyramine.

An unidentified compound in putrefied postmortem blood samples showed identical accurate mass and chromatographic behavior as 3,4-methylenedioxyamphetamine (MDA) and led to false-positive preliminary screening results. The aim of the study was to identify this unknown interference. Postmortem blood samples were analyzed after protein precipitation on a QExactive Focus high-resolution mass spectrometer (Thermo Fisher, Germany) coupled to a RP C18 column (Macherey-Nagel, Germany). Based on the analysis of mass spectrometry (MS) adducts and isotope ratios using fullscan (m/z 134-330) information, the empiric formula of the protonated molecule [M + H]+ of the unknown compound was found to be C10H14O2N (+ 0.6 ppm). Product ion spectra recorded using normalized collision energy 22% showed a base peak of C8H9O1 (+ 1.5 ppm) and a low-abundant water loss to C7H9 (+ 1.9 ppm), neutral losses of C2H2O and NH3 were found. Based on fullscan and MS-MS information and under consideration of the observed order of neutral losses, the compound was presumptively identified as N-acetyltyramine. This assumption was supported by SIRIUS software showing a SIRIUS score of 99.43% for N-acetyltyramine. Finally, the putative structure annotation was confirmed by a reference compound. The described false-positive MDA findings could be attributed to the presence of N-acetyltyramine in putrefied blood samples. Being an isomer of MDA, N-acetyltyramine could not be distinguished by high-resolution data of the protonated molecules. The presented results once again highlight that false-positive findings may occur even in hyphenatedhigh-resolution mass spectrometry (HRMS) when using full-scan information only.

AnnoSM: An Automated Annotation Tool for Determining the Substituent Modes on the Parent Skeleton Based on a Characteristic MS/MS Fragment Ion Library.

Mass spectrometry (MS) is a powerful technology for the structural elucidation of known or unknown small molecules. However, the accuracy of MS-based structure annotation is still limited due to the presence of numerous isomers in complex matrices. There are still challenges in automatically interpreting the fine structure of molecules, such as the types and positions of substituents (substituent modes, SMs) in the structure. In this study, we employed flavones, flavonols, and isoflavones as examples to develop an automated annotation method for identifying the SMs on the parent molecular skeleton based on a characteristic MS/MS fragment ion library. Importantly, user-friendly software AnnoSM was built for the convenience of researchers with limited computational backgrounds. It achieved 76.87% top-1 accuracy on the 148 authentic standards. Among them, 22 sets of flavonoid isomers were successfully differentiated. Moreover, the developed method was successfully applied to complex matrices. One such example is the extract of Ginkgo biloba L. (EGB), in which 331 possible flavonoids with SM candidates were annotated. Among them, 23 flavonoids were verified by authentic standards. The correct SMs of 13 flavonoids were ranked first on the candidate list. In the future, this software can also be extrapolated to other classes of compounds.

De Novo Genome Assembly of the Whitespot Parrotfish (Scarus forsteni): A Valuable Scaridae Genomic Resource.

Scarus forsteni, a whitespot parrotfish from the Scaridae family, is a herbivorous fish inhabiting coral reef ecosystems. The deterioration of coral reefs has highly affected the habitats of the parrotfish. The decline in genetic diversity of parrotfish emphasizes the critical importance of conserving their genetic variability to ensure the resilience and sustainability of marine ecosystems for future generations. In this study, a genome of S. forsteni was assembled de novo through using Illumina and Nanopore sequencing. The 1.71-Gb genome of S. forsteni, was assembled into 544 contigs (assembly level: contig). It exhibited an N50 length of 17.97 Mb and a GC content percentage of 39.32%. Our BUSCO analysis revealed that the complete protein of the S. forsteni genome had 98.10% integrity. Combined with structure annotation data, 34,140 (74.81%) genes were functionally annotated out of 45,638 predicted protein-coding genes. Upon comparing the genome size and TE content of teleost fishes, a roughly linear relationship was observed between these two parameters. However, TE content is not a decisive factor in determining the genome size of S. forsteni. Population history analysis results indicate that S. forsteni experienced two major population expansions, both of which occurred before the last interglacial period. In addition, through a comparative genomic analysis of the evolutionary relationship of other species, it was found that S. forsteni had the closest relationship with Cheilinus undulatus, another member of the Labridae family. Our expansion and contraction analysis of the gene family showed that the expansion genes were mainly associated with immune diseases, organismal systems, and cellular processes. At the same time, cell transcription and translation, sex hormone regulation, and other related pathways were also more prominent in the positive selection genes. The genomic sequence of S. forsteni offers valuable resources for future investigations on the conservation, evolution, and behavior of fish species.

Analyses of high spatial resolution datasets identify genes associated with multi-layered secondary cell wall thickening in Pinus bungeana.

Secondary cell wall (SCW) thickening is a major cellular developmental stage determining wood structure and properties. Although the molecular regulation of cell wall deposition during tracheary element differentiation has been well established in primary growth systems, less is known about the gene regulatory processes involved in the multi-layered SCW thickening of mature trees. Using third-generation [long-read single-molecule real-time (SMRT)] and second-generation [short-read sequencing by synthesis (SBS)] sequencing methods, we established a Pinus bungeana transcriptome resource with comprehensive functional and structural annotation for the first time. Using these approaches, we generated high spatial resolution datasets for the vascular cambium, xylem expansion regions, early SCW thickening, late SCW thickening and mature xylem tissues of 71-year-old Pinus bungeana trees. A total of 79 390 non-redundant transcripts, 31 808 long non-coding RNAs and 5147 transcription factors were annotated and quantified in different xylem tissues at all growth and differentiation stages. Furthermore, using this high spatial resolution dataset, we established a comprehensive transcriptomic profile and found that members of the NAC, WRKY, SUS, CESA and LAC gene families are major players in early SCW formation in tracheids, whereas members of the MYB and LBD transcription factor families are highly expressed during late SCW thickening. Our results provide new molecular insights into the regulation of multi-layered SCW thickening in conifers. The high spatial resolution datasets provided can serve as important gene resources for improving softwoods.

A spectroscopic test suggests that fragment ion structure annotations in MS/MS libraries are frequently incorrect

Modern untargeted mass spectrometry (MS) analyses quickly detect and resolve thousands of molecular compounds. Although features are readily annotated with a molecular formula in high-resolution small-molecule MS applications, the large majority of them remains unidentified in terms of their full molecular structure. Collision-induced dissociation tandem mass spectrometry (CID-MS2) provides a diagnostic molecular fingerprint to resolve the molecular structure through a library search. However, for de novo identifications, one must often rely on in silico generated MS2 spectra as reference. The ability of different in silico algorithms to correctly predict MS2 spectra and thus to retrieve correct molecular structures is a topic of lively debate, for instance in the CASMI contest. Underlying the predicted MS2 spectra are the in silico generated product ion structures, which are normally not used in de novo identification, but which can serve to critically assess the fragmentation algorithms. Here we evaluate in silico generated MSn product ion structures by comparison with structures established experimentally by infrared ion spectroscopy (IRIS). For a set of three dozen product ion structures from five precursor molecules, we find that virtually all fragment ion structure annotations in three major in silico MS2 libraries (HMDB, METLIN, mzCloud) are incorrect and caution the reader against their use for structure annotation of MS/MS ions.