Search

Glycosylation is one of the most common post-translational modifications (PTM) occurring in a large variety of proteins with important biological functions in human and other higher organisms. Liquid chromatography tandem mass spectrometry (LC-MS/MS) has been routinely used to characterize site-specific protein glycosylation at high throughput in complex glycoproteomic samples. Recently, electron transfer/high-energy collision dissociation (EThcD) was introduced for glycopeptide identification, which offers rich structural information on glycopepides with the fragment ions from the cleavages of both the glycan and the peptide backbone. Herein, we present the software GlycoHybridSeq for automated interpretation of EThcD-MS/MS spectra from glycoproteomic data using a customized scoring function, which enables the functionalities of identifying glycopeptides, characterizing glycosylation sites, and distinguishing some isomeric glycans. We evaluate GlycoHybridSeq on glycoproteomic data collected for cancer biomarker discovery. The results showed that it achieved comparable or better performance than that of Byonic and MSFragger. GlycoHybridSeq is released as an open source software and is ready to be used in large-scale glycoproteomic data analyses.

The ability to predict tandem mass (MS/MS) spectra from peptide sequences can significantly enhance our understanding of the peptide fragmentation process and could improve peptide identification in proteomics. However, current approaches for predicting high-energy collisional dissociation (HCD) spectra are limited to predict the intensities of expected ion types, i.e., the a/b/c/x/y/z ions and their neutral loss derivatives (referred to as backbone ions). In practice, backbone ions only account for < 70% of total ion intensities in HCD spectra, indicating many intense ions are ignored by current predictors. In this paper, we present a deep learning approach that can predict the complete spectra (both backbone and non-backbone ions) directly from peptide sequences. We made no assumptions or expectations on which kind of ions to predict but instead predicting the intensities for all possible m/z. Training this model needs no annotations of fragment ion nor any prior knowledge of the fragmentation rules. Our analyses show that the predicted 2+ and 3+ HCD spectra are highly similar to the experimental spectra, with average full-spectrum cosine similarities of 0.820 (±0.088) and 0.786 (±0.085), respectively, very close to the similarities between the experimental replicated spectra. In contrast, the best-performed backbone only models can only achieve an average similarity below 0.75 and 0.70 for 2+ and 3+ spectra, respectively. Furthermore, we developed a multi-task learning (MTL) approach for predicting spectra of insufficient training samples, which allows our model to make accurate predictions for electron transfer dissociation (ETD) spectra and HCD spectra of less abundant charges (1+ and 4+).

Glycosylation is an important post-translational modification of proteins. Many diseases, such as cancer, have proved to be related to aberrant glycosylation. High throughput quantitative methods have gained attention recently in the study of glycomics. With the development of high-resolution mass spectrometry, the sensitivity of detection in glycomics has largely improved; however, most of the commonly used MS-based techniques are focused on relative quantitative analysis, which can hardly provide direct comparative glycomic quantitation results. In this study, we developed a novel multiplex glycomic analysis method on an LC-ESI-MS platform. Reduced glycans were stable isotopic labeled during the permethylation procedure, with the use of iodomethane reagents CH2DI, CHD2I, CD3I, 13CH3I, 13CH2DI, 13CHD2I, 13CD3I, and CH3I. Up to 8-plex glycomic profiling was possible in a single analysis by LC-MS, and a 100 k mass resolution was sufficient to allow a baseline resolution of the mass differences among the 8-plex labeled glycans. The major advantages of this method are that it overcomes quantitative fluctuations caused by nanoESI, it facilitates a level of comparative quantitative glycomic analysis that accurately reflects the quantitative information in samples, and it dramatically shortens analysis time. Quantitation validation was tested on glycans released from bovine fetuin and model glycoprotein mixtures (RNase B, bovine fetuin, and IgG) with good linearity (R2 = 0.9884) and a dynamic range from 0.1 to 10. The 8-plex strategy was successfully applied to a comparative glycomic study of cancer cell lines. The results demonstrate that different distributions of sialylated glycans are related to the metastatic properties of cell lines and provide important clues for a better understanding of breast cancer brain metastasis.

RATIONALE Proteins undergo post-translational modifications and proteolytic processing that can affect their biological function. Processing often involves the loss of single residues. Cleavage of signal peptides from the N-terminus is commonly associated with translocation. Recent reports have suggested that other processing sites also exist. METHODS The secreted proteins from S. aureus N315 were precipitated with trichloroacetic acid (TCA) and amidinated with S-methyl thioacetimidate (SMTA). Amidinated proteins were digested with trypsin and analyzed with a high-resolution orbitrap mass spectrometer. RESULTS Sixteen examples of Staphylococcus aureus secretory proteins that lose an N-terminal signal peptide during their export were identified using this amidination approach. The N-termini of proteins with and without methionine were identified. Unanticipated protein cleavages due to sortase and an unknown protease were also uncovered. CONCLUSIONS A simple N-terminal amidination based mass spectrometry approach is described that facilitates identification of the N-terminus of a mature protein and the discovery of unexpected processing sites.

The retention time provides critical information for glycan annotation and quantification from the Liquid Chromatography Mass Spectrometry (LC-MS) data. However, the variation of the precise retention time of glycans is highly dependent on the experimental conditions such as the specific separating columns, MS instruments and/or the buffer used. This variation hampers the exploitation of retention time for the glycan annotation from LC-MS data, especially when inter-laboratory data are compared. To incorporate the retention time of glycan across experiments, Glucose Unit Index (GUI) can be computed using the dextrin ladder as internal standard. The retention time of glycans are then calibrated with respect to glucose units derived from dextrin ladders. Despite the successful application of the GUI approach, the manual calibration process is quite tedious and often error prone. In this work, we present a standalone software tool GlycanGUI, with a graphic user interface to automatically carry out the GUI-based glycan annotation/quantification and subsequent data analysis. When tested on experimental data, GlycanGUI reported accurate GUI values compared with manual calibration, and thus is ready to be used for automated glycan annotation and quantification using GUI.

When its DNA is damaged,Escherichia coliinduces the SOS response, which consists of about 40 genes that encode activities to repair or tolerate the damage. Certain alleles of the major SOS-control genes,recAandlexA, cause constitutive expression of the response, resulting in an increase in spontaneous mutations. These mutations, historically called “untargeted”, have been the subject of many previous studies. Here we re-examine SOS-induced mutagenesis using mutation accumulation followed by whole-genome sequencing (MA/WGS), which allows a detailed picture of the types of mutations induced as well as their sequence-specificity. Our results confirm previous findings that SOS expression specifically induces transversion base-pair substitutions, with rates averaging about 60-fold above wild-type levels. Surprisingly, the rates of G:C to C:G transversions, normally an extremely rare mutation, were induced an average of 160-fold above wild-type levels. The SOS-induced transversion showed strong sequence specificity, the most extreme of which was the G:C to C:G transversions, 60% of which occurred at the middle base of 5′GGC3′+5′GCC3′ sites although these sites represent only 8% of the G:C base pairs in the genome. SOS-induced transversions were also DNA strand biased, occurring, on average, 2- to 4- times more often when the purine was on the leading strand template and the pyrimidine on the lagging strand template than in the opposite orientation. However, the strand bias was also sequence specific, and even of reverse orientation at some sites. By eliminating constraints on the mutations that can be recovered, the MA/WGS protocol revealed new complexities to the nature of SOS “untargeted” mutations.HighlightsThe SOS response to DNA damage induces “untargeted” mutationsSOS-mutations are revealed by mutation accumulation and whole genome sequencingSOS-mutations are both sequence and DNA-strand biasedG:C to C:G transversions are particularly highly induced by SOSG:C to C:G transversions are extremely sequence and DNA-strand biased

Peptide amidination labeling using S-methyl thioacetimidate (SMTA) is investigated in an attempt to increase the number and types of peptides that can be detected in a bottom-up proteomics experiment. This derivatization method affects the basicity of lysine residues, and is shown here to significantly impact the idiosyncracies of peptide fragmentation and peptide detectability. The unique and highly reproducible fragmentation properties of SMTA-labeled peptides, such as the strong propensity for forming b1 fragment ions, can be further exploited to modify the scoring of peptide-spectrum pairs and improve peptide identification. To this end, we have developed a supervised post-processing algorithm to exploit these characteristics of peptides labeled by SMTA. Our experiments show that, although the overall number of identifications are similar, the SMTA modification enabled the detection of 16-26% peptides not previously observed in comparable CID/HCD tandem mass spectrometry experiments without SMTA labeling.

Chemical cross-linking combined with mass spectrometric analysis has become an important technique for probing protein three-dimensional structure and protein-protein interactions. A key step in this process is the accurate identification and validation of cross-linked peptides from tandem mass spectra. The identification of cross-linked peptides, however, presents challenges related to the expanded nature of the search space (all pairs of peptides in a sequence database) and the fact that some peptide-spectrum matches (PSMs) contain one correct and one incorrect peptide but often receive scores that are comparable to those in which both peptides are correctly identified. To address these problems and improve detection of cross-linked peptides, we propose a new database search algorithm, XLSearch, for identifying cross-linked peptides. Our approach is based on a data-driven scoring scheme that independently estimates the probability of correctly identifying each individual peptide in the cross-link given knowledge of the correct or incorrect identification of the other peptide. These conditional probabilities are subsequently used to estimate the joint posterior probability that both peptides are correctly identified. Using the data from two previous cross-link studies, we show the effectiveness of this scoring scheme, particularly in distinguishing between true identifications and those containing one incorrect peptide. We also provide evidence that XLSearch achieves more identifications than two alternative methods at the same false discovery rate (availability: https://github.com/COL-IU/XLSearch ).

Proofreading during DNA replication and post-replication mismatch repair are two major defenses against mutations. Foster et al. and Niccum et al. used mutation accumulation and whole genome sequencing to assemble a database of thousands... Mismatch repair (MMR) is a major contributor to replication fidelity, but its impact varies with sequence context and the nature of the mismatch. Mutation accumulation experiments followed by whole-genome sequencing of MMR-defective Escherichia coli strains yielded ≈30,000 base-pair substitutions (BPSs), revealing mutational patterns across the entire chromosome. The BPS spectrum was dominated by A:T to G:C transitions, which occurred predominantly at the center base of 5′NAC3′+5′GTN3′ triplets. Surprisingly, growth on minimal medium or at low temperature attenuated these mutations. Mononucleotide runs were also hotspots for BPSs, and the rate at which these occurred increased with run length. Comparison with ≈2000 BPSs accumulated in MMR-proficient strains revealed that both kinds of hotspots appeared in the wild-type spectrum and so are likely to be sites of frequent replication errors. In MMR-defective strains transitions were strand biased, occurring twice as often when A and C rather than T and G were on the lagging-strand template. Loss of nucleotide diphosphate kinase increases the cellular concentration of dCTP, which resulted in increased rates of mutations due to misinsertion of C opposite A and T. In an mmr ndk double mutant strain, these mutations were more frequent when the template A and T were on the leading strand, suggesting that lagging-strand synthesis was more error-prone, or less well corrected by proofreading, than was leading strand synthesis.

Eukaryotic origins of DNA replication are bound by the origin recognition complex (ORC), which scaffolds assembly of a pre-replicative complex (pre-RC) that is then activated to initiate replication. Both pre-RC assembly and activation are strongly influenced by developmental changes to the epigenome, but molecular mechanisms remain incompletely defined. We have been examining the activation of origins responsible for developmental gene amplification in Drosophila. At a specific time in oogenesis, somatic follicle cells transition from genomic replication to a locus-specific replication from six amplicon origins. Previous evidence indicated that these amplicon origins are activated by nucleosome acetylation, but how this affects origin chromatin is unknown. Here, we examine nucleosome position in follicle cells using micrococcal nuclease digestion with Ilumina sequencing. The results indicate that ORC binding sites and other essential origin sequences are nucleosome-depleted regions (NDRs). Nucleosome position at the amplicons was highly similar among developmental stages during which ORC is or is not bound, indicating that being an NDR is not sufficient to specify ORC binding. Importantly, the data suggest that nucleosomes and ORC have opposite preferences for DNA sequence and structure. We propose that nucleosome hyperacetylation promotes pre-RC assembly onto adjacent DNA sequences that are disfavored by nucleosomes but favored by ORC.

Prostate specific antigen (PSA) is currently used as a diagnostic biomarker for prostate cancer. It is a glycoprotein possessing a single glycosylation site at N69. During our previous study of PSA N69 glycosylation, additional glycopeptides were observed in the PSA sample that were not previously reported and did not match glycopeptides of impure glycoproteins existing in the sample. This extra glycosylation site of PSA is associated with a mutation in KLK3 genes. Among single nucleotide polymorphisms (SNPs) of KLKs families, the rs61752561 in KLK3 genes is an unusual missense mutation resulting in the conversion of D102 to N in PSA amino acid sequence. Accordingly, a new N-linked glycosylation site is created with an N102MS motif. Here we report the first qualitative and quantitative glycoproteomic study of PSA N102 glycosylation site by LC–MS/MS. We successfully applied tandem MS to verify the amino acid sequence possessing N102 glycosylation site and associated glycoforms of PSA samples acquired from ...

Neoepitope peptides are newly formed antigens presented by major histocompatibility complex class I (MHC-I) on cell surfaces. The cells presenting neoepitope peptides are recognized and subsequently killed by cytotoxic T-cells. Immunopeptidomic approaches aim to characterize the peptide repertoire (including neoepitope) associated with the MHC-I molecules on the surface of tumor cells using proteomic technologies, providing critical information for designing effective immunotherapy strategies. We developed a novel constrained de novo sequencing algorithm to identify neo-epitope peptides from tandem mass spectra acquired in immunopeptidomic analyses. Our method incorporates prior probabilities to putative peptides according to position specific scoring matrices (PSSMs) representing the sequence preferences recognized by MHC-I molecules. We implemented a dynamic programming algorithm to determine the peptide sequences with an optimal posterior matching score for each given MS/MS spectrum. Similar to the de novo peptide sequencing, the dynamic programming algorithm allows an efficient searching in the entire peptide sequence space. On an LC-MS/MS dataset, we demonstrated the performance of our algorithm in detecting the neoepitope peptides bound by the HLA-C*0501 molecules that were superior to database search approaches and existing general purpose de novo peptide sequencing algorithms.

RATIONALE Liquid chromatography/mass spectrometry (LC/MS) is currently considered to be a conventional glycomics analysis strategy due to the high sensitivity and ability to handle complex biological samples. Interpretation of LC/MS data is a major bottleneck in high-throughput glycomics LC/MS-based analysis. The complexity of LC/MS data associated with biological samples prompts the needs to develop computational tools capable of facilitating automated data annotation and quantitation. METHODS An LC/MS-based automated data annotation and quantitation software, MultiGlycan-ESI, was developed and utilized for glycan quantitation. Data generated by the software from LC/MS analysis of permethylated N-glycans derived from fetuin were initially validated by manual integration to assess the performance of the software. The performance of MultiGlycan-ESI was then assessed for the quantitation of permethylated fetuin N-glycans analyzed at different concentrations or spiked with permethylated N-glycans derived from human blood serum. RESULTS The relative abundance differences between data generated by the software and those generated by manual integration were less than 5%, indicating the reliability of MultiGlycan-ESI in quantitation of permethylated glycans analyzed by LC/MS. Automated quantitation resulted in a linear relationship for all six N-glycans derived from 50 ng to 400 ng fetuin with correlation coefficients (R2) greater than 0.93. Spiking of permethylated fetuin N-glycans at different concentrations in permethylated N-glycan samples derived from a 0.02 μL of HBS also exhibited linear agreement with R2 values greater than 0.9. CONCLUSIONS With a variety of options, including mass accuracy, merged adducts, and filtering criteria, MultiGlycan-ESI allows automated annotation and quantitation of LC/ESI-MS N-glycan data. The software allows the reliable quantitation of glycan LC/MS data. The software is reliable for automated glycan quantitation, thus facilitating rapid and reliable high-throughput glycomics studies. Copyright © 2014 John Wiley & Sons, Ltd.

Glycan moieties of glycoproteins modulate many biological processes in mammals, such as immune response, inflammation, and cell signaling. Numerous studies show that many human diseases are correlated with quantitative alteration of protein glycosylation. In some cases, these changes can occur for certain types of glycans over specific sites in a glycoprotein rather than on the global abundance of the glycoprotein. Conventional analytical techniques that analyze the abundance of glycans cleaved from glycoproteins cannot reveal these subtle effects. Here we present a novel statistical method to quantify the site-specific glycosylation of glycoproteins in complex samples using label-free mass spectrometric techniques. Abundance variations between sites of a glycoprotein as well as different glycoforms, that is, glycopeptides with different glycans attached to the same site, can be detected using these techniques. We applied our method to an esophageal cancer study based on blood serum samples from cancer patients in an attempt to detect potential biomarkers of site-specific N-linked glycosylation. A few glycoproteins, including vitronectin, showed significantly different site-specific glycosylations within cancer/control samples, indicating that our method is ready to be used for the discovery of glycosylated biomarkers.

Glycosylation is an important protein modification that involves enzymatic attachment of sugars to amino acid residues. Understanding the structure of these sugars and the effects of glycosylation are vital for developing indicators of disease development and progression. Although computational methods based on mass spectrometric data have proven to be effective in monitoring changes in the glycome, developing such methods for the glycoproteome are challenging, largely due to the inherent complexity in simultaneously studying glycan structures with their corresponding glycosylation sites. This paper introduces a computational framework for identifying intact N-linked glycopeptides, i.e. glycopeptides with N-linked glycans attached to their glycosylation sites, in complex proteome samples. Scoring algorithms are presented for tandem mass spectra of glycopeptides resulting from collision-induced dissociation (CID), higher-energy C-trap dissociation (HCD), and electron transfer dissociation (ETD) fragmentation modes. An empirical false-discovery rate estimation method, based on a target-decoy search approach, is derived for assigning confidence. The power of our method is further enhanced when multiple data sets are pooled together to increase identification confidence. Using this framework, 103 highly confident N-linked glycopeptides from 53 sites across 33 glycoproteins were identified in complex human serum proteome samples using conventional proteomic platforms with standard depletion of the 7-most abundant proteins. These results indicate that our method is ready to be used for characterizing site-specific protein glycosylation in complex samples.

The correlations between protein glycosylation and many biological processes and diseases are increasing the demand for quantitative glycomics strategies enabling sensitive monitoring of changes in the abundance and structure of glycans. This is currently attained through multiple strategies employing several analytical techniques such as capillary electrophoresis, liquid chromatography, and mass spectrometry. The detection and quantification of glycans often involve labeling with ionic and/or hydrophobic reagents. This step is needed in order to enhance detection in spectroscopic and mass spectrometric measurements. Recently, labeling with stable isotopic reagents has also been presented as a very viable strategy enabling relative quantitation. The different strategies available for reliable and sensitive quantitative glycomics are herein described and discussed.

Searching spectral libraries in tandem mass spectrometry (MS/MS) is an important new approach to improving the quality of peptide and protein identification. The idea relies on the observation that ion intensities in an MS/MS spectrum of a given peptide are generally reproducible across experiments, and thus, matching between spectra from an experiment and the spectra of previously identified peptides stored in a spectral library can lead to better peptide identification compared to the traditional database search. However, the use of libraries is greatly limited by their coverage of peptide sequences: even for well-studied organisms a large fraction of peptides have not been previously identified. To address this issue, we propose to expand spectral libraries by predicting the MS/MS spectra of peptides based on the spectra of peptides with similar sequences. We first demonstrate that the intensity patterns of dominant fragment ions between similar peptides tend to be similar. In accordance with this observation, we develop a neighbor-based approach which first selects peptides that are likely to have spectra similar to the target peptide and then combines their spectra using a weighted K-nearest neighbor method to accurately predict fragment ion intensities corresponding to the target peptide. This approach has the potential to predict spectra for every peptide in the proteome. When rigorous quality criteria are applied, we estimate that the method increases the coverage of spectral libraries available from the National Institute of Standards and Technology by 20–60%, although the values vary with peptide length and charge state. We find that the overall best search performance is achieved when spectral libraries are supplemented by the high quality predicted spectra.

Metaproteomic studies adopt the common bottom-up proteomics approach to investigate the protein composition and the dynamics of protein expression in microbial communities. When matched metagenomic and/or metatranscriptomic data of the microbial communities are available, metaproteomic data analyses often employ a metagenome-guided approach, in which complete or fragmental protein-coding genes are first directly predicted from metagenomic (and/or metatranscriptomic) sequences or from their assemblies, and the resulting protein sequences are then used as the reference database for peptide/protein identification from MS/MS spectra. This approach is often limited because protein coding genes predicted from metagenomes are incomplete and fragmental. In this paper, we present a graph-centric approach to improving metagenome-guided peptide and protein identification in metaproteomics. Our method exploits the de Bruijn graph structure reported by metagenome assembly algorithms to generate a comprehensive database of protein sequences encoded in the community. We tested our method using several public metaproteomic datasets with matched metagenomic and metatranscriptomic sequencing data acquired from complex microbial communities in a biological wastewater treatment plant. The results showed that many more peptides and proteins can be identified when assembly graphs were utilized, improving the characterization of the proteins expressed in the microbial communities. The additional proteins we identified contribute to the characterization of important pathways such as those involved in degradation of chemical hazards. Our tools are released as open-source software on github at https://github.com/COL-IU/Graph2Pro., Author Summary In recent years, meta-omic (including metatranscriptomic and metaproteomic) techniques have been adopted as complementary approaches to metagenomic sequencing to study functional characteristics and dynamics of microbial communities, aiming at a holistic understanding of a community to respond to the changes in the environment. Currently, metaproteomic data are largely analyzed using the bioinformatics tools originally designed in bottom-up proteomics. In particular, recent metaproteomic studies employed a metagenome-guided approach, in which complete or fragmental protein-coding genes were first predicted from metagenomic sequences (i.e., contigs or scaffolds), acquired from the matched community samples, and predicted protein sequences were then used in peptide identification. A key challenge of this approach is that the protein coding genes predicted from assembled metagenomic contigs can be incomplete and fragmented due to the complexity of metagenomic samples and the short reads length in metagenomic sequencing. To address this issue, in this paper, we present a graph-centric approach that exploits the de bruijn graph structure reported by metagenome assembly algorithms to improve metagenome-guided peptide and protein identification in metaproteomics. We show that our method can identify much more peptides and proteins, improving the characterization of the proteins expressed in the microbial communities.

Mass spectrometry has become a routine experimental tool for proteomic biomarker analysis of human blood samples, partly due to the large availability of informatics tools. As one of the most common protein post-translational modifications (PTMs) in mammals, protein glycosylation has been observed to alter in multiple human diseases and thus may potentially be candidate markers of disease progression. While mass spectrometry instrumentation has seen advancements in capabilities, discovering glycosylation-related markers using existing software is currently not straightforward. Complete characterization of protein glycosylation requires the identification of intact glycopeptides in samples, including identification of the modification site as well as the structure of the attached glycans. In this paper, we present GlycoSeq, an open-source software tool that implements a heuristic iterated glycan sequencing algorithm coupled with prior knowledge for automated elucidation of the glycan structure within a glycopeptide from its collision-induced dissociation tandem mass spectrum. GlycoSeq employs rules of glycosidic linkage as defined by glycan synthetic pathways to eliminate improbable glycan structures and build reasonable glycan trees. We tested the tool on two sets of tandem mass spectra of N-linked glycopeptides cell lines acquired from breast cancer patients. After employing enzymatic specificity within the N-linked glycan synthetic pathway, the sequencing results of GlycoSeq were highly consistent with the manually curated glycan structures. Hence, GlycoSeq is ready to be used for the characterization of glycan structures in glycopeptides from MS/MS analysis. GlycoSeq is released as open source software at https://github.com/chpaul/GlycoSeq/ .

We report analysis of N-glycans derived from disease-free individuals and patients with Barrett's esophagus, high-grade dysplasia, and esophageal adenocarcinoma by microchip electrophoresis with laser-induced fluorescence detection. Serum samples in 10-μL aliquots are enzymatically treated to cleave the N-glycans that are subsequently reacted with 8-aminopyrene-1,3,6-trisulfonic acid to add charge and a fluorescent label. Separations at 1250 V/cm and over 22 cm yielded efficiencies up to 700,000 plates for the N-glycans and analysis times under 100 s. Principal component analysis (PCA) and analysis of variance (ANOVA) tests of the peak areas and migration times are used to evaluate N-glycan profiles from native and desialylated samples and to determine differences among the four sample groups. With microchip electrophoresis, we are able to distinguish the three patient groups from each other and from disease-free individuals.

Though many software packages have been developed to perform label-free quantification of proteins in complex biological samples using peptide intensities generated by LC-MS/MS, two critical issues are generally ignored in this field: (i) peptides have multiple elution patterns across runs in an experiment, and (ii) many peptides cannot be used for protein quantification. To address these two key issues, we have developed a novel alignment method to enable accurate peptide peak retention time determination and multiple filters to eliminate unqualified peptides for protein quantification. Repeatability and linearity have been tested using six very different samples, i.e., standard peptides, kidney tissue lysates, HT29-MTX cell lysates, depleted human serum, human serum albumin-bound proteins, and standard proteins spiked in kidney tissue lysates. At least 90.8% of the proteins (up to 1,390) had CVs ≤ 30% across 10 technical replicates, and at least 93.6% (up to 2,013) had R(2) ≥ 0.9500 across 7 concentrations. Identical amounts of standard protein spiked in complex biological samples achieved a CV of 8.6% across eight injections of two groups. Further assessment was made by comparing mass spectrometric results to immunodetection, and consistent results were obtained. The new approach has novel and specific features enabling accurate label-free quantification.

Protein glycosylation is one of the most common post-translational modifications, estimated to occur in over 50% of human proteins. Mass spectrometry (MS)-based approaches involving different fragmentation mechanisms have been frequently used to detect and characterize protein N-linked glycosylations. In addition to the popular Collision-Induced Dissociation (CID), high-energy C-trap dissociation (HCD) fragmentation, which is a feature of a linear ion trap orbitrap hybrid mass spectrometer (LTQ Orbitrap), has been recently used for the fragmentation of tryptic N-linked glycopeptides in glycoprotein analysis. The oxonium ions observed with high mass accuracy in the HCD spectrum of glycopeptides can be combined with characteristic fragmentation patterns in the CID spectrum resulting from consecutive glycosidic bond cleavages, to improve the detection and characterization of N-linked glycopeptides. As a means of automating this process, we describe here GlypID 2.0, a software tool that implements several algorithmic approaches to utilize MS information including accurate precursor mass and spectral patterns from both HCD and CID spectra, thus allowing for an unequivocal and accurate characterization of N-linked glycosylation sites of proteins. Copyright © 2011 John Wiley & Sons, Ltd.

A synthetic approach to model the analytical complexity of biological proteolytic digests has been developed. Combinatorial peptide libraries ranging in length between nine and twelve amino acids that represent typical tryptic digests were designed, synthesized and analyzed. Individual libraries and mixtures thereof were studied by replicate liquid chromatography-ion trap mass spectrometry and compared to a tryptic digest of Deinococcus radiodurans. Similar to complex proteome analysis, replicate study of individual libraries identified additional unique peptides. Fewer novel sequences were revealed with each additional analysis in a manner similar to that observed for biological data. Our results demonstrate a bimodal distribution of peptides sorting to either very low or very high levels of detection. Upon mixing of libraries at equal abundance, a length-dependent bias in favor of longer sequence identification was observed. Peptide identification as a function of site-specific amino acid content was characterized with certain amino acids proving to be of considerable importance. This report demonstrates that peptide libraries of defined character can serve as a reference for instrument characterization. Furthermore, they are uniquely suited to delineate the physical properties that influence identification of peptides which provides a foundation for optimizing the study of samples with less defined heterogeneity.

Glycosylation is one of the most common posttranslational modifications (PTMs) of proteins, the characterization of which is commonly achieved through proteomic protocol, involving trypsin digestion followed by liquid chromatography/tandem mass spectrometry (LC/MS/MS). However, it is often not possible to characterize all glycopeptides in a complex sample because of the high complexity of glycoproteomic samples, and the relative lower abundances of glycopeptides in comparison to the unmodified peptides. We present here a targeted MS/MS analysis approach, which utilizes a previously developed computational tool, GlyPID, to guide multiple experiments, thus permitting a complete characterization of all N-glycosylation sites of glycoproteins present in a complex sample. We have tested our approach using model glycoproteins analyzed by high-resolution LTQ-FT MS. The results demonstrate a potential use of our method for a high-throughput characterization of complex mixtures of glycosylated proteins.

Glycans commonly exhibit variations in their branching structures and glycosidic bond linkages, in addition to their sequence variation. These glycan features are known to be highly correlated with their biological functions. It is relatively straightforward to deduce the composition and sequence of monosaccharides of a glycan from its tandem mass spectra. However, the characterization of the linkage types of each glycosidic bond is still analytically challenging. In this paper, we present a rank-based discriminative model to differentiate between two types of glycosidic linkages (namely, 1-4 and 1-6) based on the cross-ring fragmentation patterns of the corresponding glycans observed under high-energy collision-induced dissociation (CID). To train our models, we acquired tandem mass spectra for three groups of both native and permethylated linear oligoglucoses using a matrix-assisted laser desorption/ionization tandem time-of-flight (MALDI/TOF/TOF) instrument. Based on a 5-fold cross-validation, the prediction accuracies of our model for native glycans are determined to be about 88.4% and 92.9% for 1-6 and 1-4 linkages, respectively. The accuracies determined for permethylated glycans are slightly lower, but comparable: 85.6% and 89.0% for 1-6 and 1-4 linkages, respectively. Our method is implemented as a web-hosted utility, thus making it readily accessible to the public which can be accessed through http://ggdb.informatics.indiana.edu:8080/glycanview.

The rate of cytosine deamination is much higher in single-stranded DNA (ssDNA) than in double-stranded DNA, and copying the resulting uracils causes C to T mutations. To study this phenomenon, the catalytic domain of APOBEC3G (A3G-CTD), an ssDNA-specific cytosine deaminase, was expressed in an Escherichia coli strain defective in uracil repair (ung mutant), and the mutations that accumulated over thousands of generations were determined by whole-genome sequencing. C:G to T:A transitions dominated, with significantly more cytosines mutated to thymine in the lagging-strand template (LGST) than in the leading-strand template (LDST). This strand bias was present in both repair-defective and repair-proficient cells and was strongest and highly significant in cells expressing A3G-CTD. These results show that the LGST is accessible to cellular cytosine deaminating agents, explains the well-known GC skew in microbial genomes, and suggests the APOBEC3 family of mutators may target the LGST in the human genome.

This chapter introduces computational methods used in mass spectrometry-based proteomics, including those for addressing the critical problems such as peptide identification and protein inference, peptide and protein quantification, characterization of posttranslational modifications (PTMs), and data-independent acquisitions (DIA). The chapter concludes with emerging applications of proteomic techniques, such as metaproteomics, glycoproteomics, and proteogenomics.

Glycosylation is one of the most common post-translational modifications in proteins, existing in ~50% of mammalian proteins. Several research groups have demonstrated that mass spectrometry is an efficient technique for glycopeptide identification; however, this problem is still challenging because of the enormous diversity of glycan structures and the microheterogeneity of glycans. In addition, a glycopeptide may contain multiple glycosylation sites, making the problem complex. Current software tools often fail to identify glycopeptides with multiple glycosylation sites, and hence we present GlycoMID, a graph-based spectral alignment algorithm that can identify glycopeptides with multiple hydroxylysine O-glycosylation sites by tandem mass spectra. GlycoMID was tested on mass spectrometry data sets of the bovine collagen α-(II) chain protein, and experimental results showed that it identified more glycopeptide-spectrum matches than other existing tools, including many glycopeptides with two glycosylation sites.

Label-free absolute protein quantification refers to a process of estimating protein abundances in complex biological samples based on the data acquired from a liquid chromatography mass spectrometry (LC-MS) analysis. Most approaches to label-free quantification rely on measuring peak areas from an extracted-ion chromatogram. However, because of the differences in physicochemical properties associated with different peptide ions, observed peak areas in a single experiment are determined not only by peptide abundances, but also the intrinsic biases of analytical platforms. Therefore, accurate modeling of these biases provides an opportunity to developing new computational methods for precise absolute protein quantification. In this work, we developed a new algorithm for absolute quantification of proteins. The approach is based on the concept of peptide response rate, which characterizes the peptide-specific signal detection bias in an LC-MS experiment. We argue that peptide response rate is an intrinsic and reproducible property of peptide ions that can be reliably predicted using non-linear regression and features extracted from the sequence of the parent protein. Protein abundances are estimated using a maximum likelihood model in which the observed peak areas of peptide ions are adjusted using predicted peptide response rates. We evaluate our approach on a large LC-MS dataset as well as simulated data and provide evidence that the accuracy of absolute protein quantification is improved when peptide-specific response rates are taken into account.

We have used a simple and efficient approach by combining the known functional and structural properties of penicillin G acylase (PGA) from E. coli, and tried to mutate PGA of Bacillus megaterium with the goal of increasing the stability of the enzyme in organic solvents or at acidic pH. The PGA mutants Kβ427A, Kβ430A and Kβ427A/Kβ430A obtained have higher stability in DMF than the wild-type PGA.

Prostate specific antigen (PSA) is currently used as a biomarker to diagnose prostate cancer. PSA testing has been widely used to detect and screen prostate cancer. However, in the diagnostic gray zone, the PSA test does not clearly distinguish between benign prostate hypertrophy and prostate cancer due to their overlap. To develop more specific and sensitive candidate biomarkers for prostate cancer, an in-depth understanding of the biochemical characteristics of PSA (such as glycosylation) is needed. PSA has a single glycosylation site at Asn69, with glycans constituting approximately 8% of the protein by weight. Here, we report the comprehensive identification and quantitation of N-glycans from two PSA isoforms using LC–MS/MS. There were 56 N-glycans associated with PSA, whereas 57 N-glycans were observed in the case of the PSA-high isoelectric point (pI) isoform (PSAH). Three sulfated/phosphorylated glycopeptides were detected, the identification of which was supported by tandem MS data. One of these sulfated/phosphorylated N-glycans, HexNAc5Hex4dHex1s/p1 was identified in both PSA and PSAH at relative intensities of 0.52 and 0.28%, respectively. Quantitatively, the variations were monitored between these two isoforms. Because we were one of the laboratories participating in the 2012 ABRF Glycoprotein Research Group (gPRG) study, those results were compared to that presented in this study. Our qualitative and quantitative results summarized here were comparable to those that were summarized in the interlaboratory study.

Glycosylation of proteins and lipids has been recently shown to play important biological roles, including cell signaling, adhesion motility, immune response, and pathogen interaction. Functions of many proteins are modulated through glycosylation. To better understand the biological attributes of glycans, reliable quantitative glycomics is needed. Several methods prompting effective monitoring of glycans in biological systems have been developed. The aim of this article is to highlight recent advancements in quantitative glycomics and the necessary bioinformatics tools.

Glycomics aims to identify the whole set of functional glycans of glycoconjugates (attached to proteins or lipids) in biological samples. Glycoproteomics aims to characterize the complete structure of all glycoproteins in biological samples, including the glycosylation sites of proteins and the various glycan structures attached to each of these sites. Mass spectrometry (MS) and microarray are high-throughput technologies that are commonly used in glycomics and glycoproteomics, which often result in the generation of large experimental datasets. Bioinformatics approaches play an essential role in automated analysis and interpretation of such data. This unit describes and discusses the computational tools currently available for these analyses, and their glycomics and glycoproteomics applications.

One of the principal goals of glycoprotein research is to correlate glycan structure and function. Such correlation is necessary in order for one to understand the mechanisms whereby glycoprotein structure elaborates the functions of myriad proteins. The accurate comparison of glycoforms and quantification of glycosites are essential steps in this direction. Mass spectrometry has emerged as a powerful analytical technique in the field of glycoprotein characterization. Its sensitivity, high dynamic range, and mass accuracy provide both quantitative and sequence/structural information. As part of the 2012 ABRF Glycoprotein Research Group study, we explored the use of mass spectrometry and ancillary methodologies to characterize the glycoforms of two sources of human prostate specific antigen (PSA). PSA is used as a tumor marker for prostate cancer, with increasing blood levels used to distinguish between normal and cancer states. The glycans on PSA are believed to be biantennary N-linked, and it has been observed that prostate cancer tissues and cell lines contain more antennae than their benign counterparts. Thus, the ability to quantify differences in glycosylation associated with cancer has the potential to positively impact the use of PSA as a biomarker. We studied standard peptide-based proteomics/glycomics methodologies, including LC-MS/MS for peptide/glycopeptide sequencing and label-free approaches for differential quantification. We performed an interlaboratory study to determine the ability of different laboratories to correctly characterize the differences between glycoforms from two different sources using mass spectrometry methods. We used clustering analysis and ancillary statistical data treatment on the data sets submitted by participating laboratories to obtain a consensus of the glycoforms and abundances. The results demonstrate the relative strengths and weaknesses of top-down glycoproteomics, bottom-up glycoproteomics, and glycomics methods.

We introduce three software tools, Cartoonist, GlycoWorkbench, and MultiGlycan, for N-glycan profiling of complex biological samples. Detailed instructions for using these tools are provided, and their performances are demonstrated by using real glycan profiling data.

Multisubunit RNA polymerases IV and V (Pols IV and V) mediate RNA-directed DNA methylation and transcriptional silencing of retrotransposons and heterochromatic repeats in plants. We identified genomic sites of Pol V occupancy in parallel with siRNA deep sequencing and methylcytosine mapping, comparing wild-type plants with mutants defective for Pol IV, Pol V, or both Pols IV and V. Approximately 60% of Pol V-associated regions encompass regions of 24-nucleotide (nt) siRNA complementarity and cytosine methylation, consistent with cytosine methylation being guided by base-pairing of Pol IV-dependent siRNAs with Pol V transcripts. However, 27% of Pol V peaks do not overlap sites of 24-nt siRNA biogenesis or cytosine methylation, indicating that Pol V alone does not specify sites of cytosine methylation. Surprisingly, the number of methylated CHH motifs, a hallmark of RNA-directed de novo methylation, is similar in wild-type plants and Pol IV or Pol V mutants. In the mutants, methylation is lost at 50%–60% of the CHH sites that are methylated in the wild type but is gained at new CHH positions, primarily in pericentromeric regions. These results indicate that Pol IV and Pol V are not required for cytosine methyltransferase activity but shape the epigenome by guiding CHH methylation to specific genomic sites.

As a classic topic in bioinformatics, the fragment assembly problem has been studied for over two decades. Fragment assembly algorithms take a set of DNA fragments as input, piece them together into a set of aligned overlapping fragments (i.e., contigs), and output a consensus sequence for each of the contigs. The rapid advance of massively parallel sequencing, often referred to as next-generation sequencing (NGS) technologies, has revolutionized DNA sequencing by reducing both its time and cost by several orders of magnitude in the past few years, but posed new challenges for fragment assembly. As a result, many new approaches have been developed to assemble NGS sequences, which are typically shorter with a higher error rate, but at a much higher throughput, than classic methods provided. In this chapter, we review both classic and new algorithms for fragment assembly, with a focus on NGS sequences. We also discuss a few new assembly problems emerging from the broader applications of NGS techniques, which are distinct from the classic fragment assembly problem.

PSD does not usually generate a complete series of y-type ions, particularly at high mass, and this is a limitation for de novo sequencing algorithms. It is demonstrated that b(2) and b(3) ions can be used to help assign high mass x(N-2) and x(N-3) fragments that are found in vacuum ultraviolet (VUV) photofragmentation experiments. In addition, v(N)-type ion fragments with side chain loss from the N-terminal residue often enable confirmation of N-terminal amino acids. Libraries containing several thousand peptides were examined using photodissociation in a MALDI-TOF/TOF instrument. 1345 photodissociation spectra with a high S/N ratio were interpreted.

We estimated the reproducibility of tandem mass fragmentation spectra for the widely-used collision-induced dissociation (CID) instruments. Using the Pearson correlation coefficient as a measure of spectral similarity, we found that the within-experiment reproducibility of fragment ion intensities is very high (about 0.85). However, across different experiments and instrument types/setups, the correlation decreases by more than 15% (to about 0.70). We further investigated the accuracy of current predictors of peptide fragmentation spectra and found that they are more accurate than the ad-hoc models generally used by search engines (e.g. SEQUEST) and, surprisingly, approaching the empirical upper limit set by the average across-experiment spectral reproducibility (especially for charge +1 and charge +2 precursor ions). These results provide evidence that, in terms of accuracy of modeling, predicted peptide fragmentation spectra provide a viable alternative to spectral libraries for peptide identification, with a higher coverage of peptides and lower storage requirements. Furthermore, using five data sets of proteome digests by two different proteases, we find that PeptideART (a data-driven machine learning approach) is generally more accurate than MassAnalyzer (an approach based on a kinetic model for peptide fragmentation) in predicting fragmentation spectra, but that both models are significantly more accurate than the ad-hoc models. Availability: PeptideART is freely available at www.informatics.indiana.edu/predrag.

Peptide detectability is defined as the probability that a peptide is identified in an LC-MS/MS experiment and has been useful in providing solutions to protein inference and label-free quantification. Previously, predictors for peptide detectability trained on standard or complex samples were proposed. Although the models trained on complex samples may benefit from the large training data sets, it is unclear to what extent they are affected by the unequal abundances of identified proteins. To address this challenge and improve detectability prediction, we present a new algorithm for the iterative learning of peptide detectability from complex mixtures. We provide evidence that the new method approximates detectability with useful accuracy and, based on its design, can be used to interpret the outcome of other learning strategies. We studied the properties of peptides from the bacterium Deinococcus radiodurans and found that at standard quantities, its tryptic peptides can be roughly classified as either detectable or undetectable, with a relatively small fraction having medium detectability. We extend the concept of detectability from peptides to proteins and apply the model to predict the behavior of a replicate LC-MS/MS experiment from a single analysis. Finally, our study summarizes a theoretical framework for peptide/protein identification and label-free quantification.

Acute light-induced photoreceptor degeneration has been studied in experimental animals as a model for photoreceptor cell loss in human retinal degenerative diseases. Light absorption by rhodopsin in rod photoreceptor outer segments (OS) induces oxidative stress and initiates apoptotic cell death. However, the molecular events that induce oxidative stress and initiate the apoptotic cascade remain poorly understood. To better understand the molecular mechanisms of light-induced photoreceptor cell death, we studied the proteomic changes in OS upon intense light exposure by using a proteolytic (18)O labeling method. Of 171 proteins identified, the relative abundance of 98 proteins in light-exposed and unexposed OS was determined. The quantities of 11 proteins were found to differ by more than 2-fold between light-exposed OS and those remaining in darkness. Among the 11 proteins, 8 were phototransduction proteins and 7 of these were altered such that the efficiency of phototransduction would be reduced or quenched during light exposure. In contrast, the amount of OS rhodopsin kinase was reduced by 2-fold after light exposure, suggesting attenuation in the mechanism of quenching phototransduction. Liquid chromatography multiple reaction monitoring (LC-MRM) was performed to confirm this reduction in the quantity of rhodopsin kinase. As revealed by immunofluorescence microscopy, this reduction of rhodopsin kinase is not a result of protein translocation from the outer to the inner segment. Collectively, our findings suggest that the absolute quantity of rhodopsin kinase in rod photoreceptors is reduced upon light stimulation and that this reduction may be a contributing factor to light-induced photoreceptor cell death. This report provides new insights into the proteomic changes in the OS upon intense light exposure and creates a foundation for understanding the mechanisms of light-induced photoreceptor cell death.

In simplifying the evaluation and quantification of high-throughput label-free quantitative proteomic data, we introduce ProteinQuant Suite. It comprises three standalone complementary computer utilities, namely ProtParser, ProteinQuant, and Turbo RAW2MGF. ProtParser is a filtering utility designed to evaluate database search results. Filtering is performed according to different criteria that are defined by the end-user. ProteinQuant then utilizes this parsed list of peptides and proteins in conjunction with mzXML or mzData files generated from the raw files for quantification. This quantification is based on the automatic detection and integration of chromatographic peaks representative of the liquid chromatography/mass spectrometry (LC/MS) elution profiles of identified peptides. Turbo RAW2MGF was developed to extend the applicability of ProteinQuant Suite to data collected from different types of mass spectrometers. It directly processes raw data files generated by Xcalibur, a ThermoElectron data acquisition software, and generates a MASCOT generic file (MGF). This file format is needed since the protein identification results generated by the database search employing this file format include information required for the precise identification and quantification of chromatographic peaks. The performance of ProteinQuant Suite was initially validated using LC/MS/MS generated for a mixture of standard proteins as well as standard proteins spiked in a complex biological matrix such as blood serum. Automated quantification of the collected data resulted in calibration curves with R2 values higher than 0.95 with linearity spanning over more than 2 orders of magnitude with peak quantification reproducibility better than 15% (RSD). ProteinQuant Suite was also applied to confirm the binding preference of standard glycoproteins to Con A lectin using a sample consisting of both standard glycoproteins and proteins. Copyright © 2008 John Wiley & Sons, Ltd.

Motivation: While it is common to refer to ‘the genome sequence’ as if it were a single, complete and contiguous DNA string, it is in fact an assembly of millions of small, partially overlapping DNA fragments. Sophisticated computer algorithms (assemblers and scaffolders) merge these DNA fragments into contigs, and place these contigs into sequence scaffolds using the paired-end sequences derived from large-insert DNA libraries. Each step in this automated process is susceptible to producing errors; hence, the resulting draft assembly represents (in practice) only a likely assembly that requires further validation. Knowing which parts of the draft assembly are likely free of errors is critical if researchers are to draw reliable conclusions from the assembled sequence data. Results: We develop a machine-learning method to detect assembly errors in sequence assemblies. Several in silico measures for assembly validation have been proposed by various researchers. Using three benchmarking Drosophila draft genomes, we evaluate these techniques along with some new measures that we propose, including the good-minus-bad coverage (GMB), the good-to-bad-ratio (RGB), the average Z-score (AZ) and the average absolute Z-score (ASZ). Our results show that the GMB measure performs better than the others in both its sensitivity and its specificity for assembly error detection. Nevertheless, no single method performs sufficiently well to reliably detect genomic regions requiring attention for further experimental verification. To utilize the advantages of all these measures, we develop a novel machine learning approach that combines these individual measures to achieve a higher prediction accuracy (i.e. greater than 90%). Our combined evidence approach avoids the difficult and often ad hoc selection of many parameters the individual measures require, and significantly improves the overall precisions on the benchmarking data sets. Availability: http://people.cgb.indiana.edu/jeochoi/gav/ Contact: jeochoi@indiana.edu Supplementary information: Supplementary data are available at Bioinformatics online.

Motivation: One of the major problems in shotgun proteomics is the low peptide coverage when analyzing complex protein samples. Identifying more peptides, e.g. non-tryptic peptides, may increase the peptide coverage and improve protein identification and/or quantification that are based on the peptide identification results. Searching for all potential non-tryptic peptides is, however, time consuming for shotgun proteomics data from complex samples, and poses a challenge for a routine data analysis. Results: We hypothesize that non-tryptic peptides are mainly created from the truncation of regular tryptic peptides before separation. We introduce the notion of truncatability of a tryptic peptide, i.e. the probability of the peptide to be identified in its truncated form, and build a predictor to estimate a peptide's truncatability from its sequence. We show that our predictions achieve useful accuracy, with the area under the ROC curve from 76% to 87%, and can be used to filter the sequence database for identifying truncated peptides. After filtering, only a limited number of tryptic peptides with the highest truncatability are retained for non-tryptic peptide searching. By applying this method to identification of semi-tryptic peptides, we show that a significant number of such peptides can be identified within a searching time comparable to that of tryptic peptide identification. Contact: predrag@indiana.edu; rarnold@indiana.edu; hatang@indiana.edu

Glycosylation is one of the most common post-translational modifications (PTMs) of proteins, the characterization of which is commonly achieved utilizing mass spectrometry (MS). However, its applicability is currently limited by the lack of computational tools capable of autmoated interpretation of high throughput MS experiments which would allow the characterization of glycosylation sites and their microheterogeneities. We present here a computational approach which overcomes this problem and allows the identification and assignment of the microheterogeneities of glycosylation sites of glycoproteins from liquid chromatography ion-trap-based mass spectrometry (LC/MS) data. This method was implemented in a software tool and tested on several model glycoproteins. The results demonstrate the potential of our computational approach in automating the high throughput identification of glycoproteins.

A major challenge in shotgun proteomics has been the assignment of identified peptides to the proteins from which they originate, referred to as the protein inference problem. Redundant and homologous protein sequences present a challenge in being correctly identified, as a set of peptides may in many cases represent multiple proteins. One simple solution to this problem is the assignment of the smallest number of proteins that explains the identified peptides. However, it is not certain that a natural system should be accurately represented using this minimalist approach. In this paper, we propose a reformulation of the protein inference problem by utilizing the recently introduced concept of peptide detectability. We also propose a heuristic algorithm to solve this problem and evaluate its performance on synthetic and real proteomics data. In comparison to a greedy implementation of the minimum protein set algorithm, our solution that incorporates peptide detectability performs favorably.

Accurate peptide identification from tandem mass spectrometry experiments is the cornerstone of proteomics. Although various approaches for matching database sequences with experimental spectra have been developed to date (e.g. Sequest, Mascot) the sensitivity and specificity of peptide identification have not yet reached their full potential. This is in part due to the tradeoffs between robustness and accuracy of the existing methods with respect to the non-uniform nature of peptide fragmentation and bond cleavages induced by different mass spectrometers. Accordingly, it is expected that new approaches to de novo predicting peptide fragmentation spectra will enable more accurate peptide identification. To address this problem, here we used a data-driven approach to learn peptide fragmentation rules in mass spectrometry, in the form of posterior probabilities, for various fragment-ion types of doubly and triply charged precursor ions. We show that the accuracy of our neural-network based methodology is useful for subsequent peptide database searches and that the most useful rules of fragmentation significantly differ across ion and precursor types.

Motivation: The emerging glycomics and glycoproteomics projects aim to characterize all forms of glycoproteins in different tissues and organisms. Tandem mass spectrometry (MS/MS) is the key experimental methodology for high-throughput glycan identification and characterization. Fragmentation of glycans from high energy collision-induced dissociation generates ions from glycosidic as well as internal cleavages. The cross-ring ions resulting from internal cleavages provide additional information that is important to reveal the type of linkage between monosaccharides. This information, however, is not incorporated into the current programs for analyzing glycan mass spectra. As a result, they can rarely distinguish from the mass spectra isomeric oligosaccharides, which have the same saccharide composition but different types of sequences, branches or linkages. Results: In this paper, we describe a novel algorithm for glycan characterization using MS/MS. This algorithm consists of three steps. First, we develop a scoring scheme to identify potential bond linkages between monosaccharides, based on the appearance pattern of cross-ring ions. Next, we use a dynamic programming algorithm to determine the most probable oligosaccharide structures from the mass spectrum. Finally, we re-evaluate these oligosaccharide structures, taking into account the double fragmentation ions. We also show the preliminary results of testing our algorithm on several MS/MS spectra of oligosaccharides. Availability: The program GLYCH is available upon request from the authors. Contact: hatang@indiana.edu