Canonical correlation analysis for multi-omics: Application to cross-cohort analysis.

Integrative approaches that simultaneously model multi-omics data have gained increasing popularity because they provide holistic system biology views of multiple or all components in a biological system of interest. Canonical correlation analysis (CCA) is a correlation-based integrative method designed to extract latent features shared between multiple assays by finding the linear combinations of features-referred to as canonical variables (CVs)-within each assay that achieve maximal across-assay correlation. Although widely acknowledged as a powerful approach for multi-omics data, CCA has not been systematically applied to multi-omics data in large cohort studies, which has only recently become available. Here, we adapted sparse multiple CCA (SMCCA), a widely-used derivative of CCA, to proteomics and methylomics data from the Multi-Ethnic Study of Atherosclerosis (MESA) and Jackson Heart Study (JHS). To tackle challenges encountered when applying SMCCA to MESA and JHS, our adaptations include the incorporation of the Gram-Schmidt (GS) algorithm with SMCCA to improve orthogonality among CVs, and the development of Sparse Supervised Multiple CCA (SSMCCA) to allow supervised integration analysis for more than two assays. Effective application of SMCCA to the two real datasets reveals important findings. Applying our SMCCA-GS to MESA and JHS, we identified strong associations between blood cell counts and protein abundance, suggesting that adjustment of blood cell composition should be considered in protein-based association studies. Importantly, CVs obtained from two independent cohorts also demonstrate transferability across the cohorts. For example, proteomic CVs learned from JHS, when transferred to MESA, explain similar amounts of blood cell count phenotypic variance in MESA, explaining 39.0% ~ 50.0% variation in JHS and 38.9% ~ 49.1% in MESA. Similar transferability was observed for other omics-CV-trait pairs. This suggests that biologically meaningful and cohort-agnostic variation is captured by CVs. We anticipate that applying our SMCCA-GS and SSMCCA on various cohorts would help identify cohort-agnostic biologically meaningful relationships between multi-omics data and phenotypic traits.

Genetic interactions drive heterogeneity in causal variant effect sizes for gene expression and complex traits.

Despite the growing number of genome-wide association studies (GWASs), it remains unclear to what extent gene-by-gene and gene-by-environment interactions influence complex traits in humans. The magnitude of genetic interactions in complex traits has been difficult to quantify because GWASs are generally underpowered to detect individual interactions of small effect. Here, we develop a method to test for genetic interactions that aggregates information across all trait-associated loci. Specifically, we test whether SNPs in regions of European ancestry shared between European American and admixed African American individuals have the same causal effect sizes. We hypothesize that in African Americans, the presence of genetic interactions will drive the causal effect sizes of SNPs in regions of European ancestry to be more similar to those of SNPs in regions of African ancestry. We apply our method to two traits: gene expression in 296 African Americans and 482 European Americans in the Multi-Ethnic Study of Atherosclerosis (MESA) and low-density lipoprotein cholesterol (LDL-C) in 74K African Americans and 296K European Americans in the Million Veteran Program (MVP). We find significant evidence for genetic interactions in our analysis of gene expression; for LDL-C, we observe a similar point estimate, although this is not significant, most likely due to lower statistical power. These results suggest that gene-by-gene or gene-by-environment interactions modify the effect sizes of causal variants in human complex traits.

Rapid characterization of adeno-associated virus (AAV) capsid proteins using microchip ZipChip CE-MS.

Adeno-associated viruses (AAVs) are viral vectors used as delivery systems for gene therapies. Intact protein characterization of AAV viral capsid proteins (VPs) and their post-translational modifications is critical to ensuring product quality. In this study, microchip-based ZipChip capillary electrophoresis-mass spectrometry (CE-MS) was applied for the rapid characterization of AAV intact VPs, specifically full and empty viral capsids of serotypes AAV6, AAV8 and AAV9, which was accomplished using 5 min of analysis time. Low levels of dimethyl sulfoxide (4%) in the background electrolyte (BGE) improved MS signal quality and component detection. A sensitivity evaluation revealed consistent detection of VP proteoforms when as little as 2.64 × 106 viral particles (≈26.4 picograms) were injected. Besides the traditional VP proteoforms used for serotype identification, multiple VP3 variants were detected, including truncated VP3 variants most likely generated by leaky scanning as well as unacetylated and un-cleaved VP3 proteoforms. Phosphorylation, known to impact AAV transduction efficiency, was also seen in all serotypes analysed. Additionally, low abundant fragments originating from either N- or C-terminus truncation were detected. As the aforementioned VP components can impact product quality and efficacy, the ZipChip's ability to rapidly characterize them illustrates its strength in monitoring product quality during AAV production.

Development of a Rapid Adeno-Associated Virus (AAV) Identity Testing Platform through Comprehensive Intact Mass Analysis of Full-Length AAV Capsid Proteins.

Adeno-associated viruses (AAVs) are commonly used as vectors for the delivery of gene therapy targets. Characterization of AAV capsid proteins (VPs) and their post-translational modifications (PTMs) have become a critical attribute monitored to evaluate product quality. Liquid chromatography-mass spectrometry (LC-MS) analysis of intact AAV VPs provides both quick and reliable serotype identification as well as proteoform information on each VP. Incorporating these analytical strategies into rapid good manufacturing practice (GMP)-compliant workflows containing robust, but simplified, data processing methods is necessary to ensure effective product quality control (QC) during production. Here, we present a GMP-compliant LC-MS workflow for the rapid identification and in-depth characterization of AAVs. Hydrophilic interaction liquid chromatography (HILIC) MS with difluoroacetic acid as a mobile phase modifier is utilized to achieve the intact separation and identification of AAV VPs and their potential proteoforms. Peptide mapping is performed to confirm PTMs identified during intact VP analysis and for in-depth PTM characterization. The intact separations platform is then incorporated into a data processing workflow developed using GMP-compliant software capable of rapid AAV serotype identification and, if desired, specific serotype PTM monitoring and characterization. Such a platform provides product QC capabilities that are easily accessible in a regulatory setting.

Molecular examination of the endogenous opioid system in rhesus macaque monkeys with self-injurious behavior.

Self-injurious behavior (SIB) can lead to serious injury and occurs in approximately 1%-4% of the adult population, with higher incidences in adolescent and institutionalized populations, as well as in children with developmental disorders such as Autism. SIB also spontaneously occurs in a low percentage of captive monkeys. Rhesus macaque (Macaca mulatta) monkeys are evolutionarily and physiologically similar to humans, share 93% genetic sequence similarity to humans, and have long been used as testing subjects for vaccine and clinical trials. Previous studies hypothesized that altered endogenous opioid expression occurs in the brains of individuals and animals that self-injure. We examined the regional mRNA expression of opioid signaling genes in sixteen rhesus macaques that exhibited SIB and eight sex- and age- matched controls. The brain regions examined are linked to reward reinforcement and stress adaptation including the hypothalamus, orbital frontal cortex, nucleus accumbens, hippocampus, caudate, and the amygdala. We found decreased µ-opioid receptor (OPRM1) in the amygdala of monkeys with SIB, and reduced prodynorphin (PDYN) in the hypothalamus. Our data suggest dysfunction in the regulation of opioid peptide precursors and calls for further investigation of the endogenous opioid system in SIB.

An FBN1 deep intronic variant is associated with pseudoexon formation and a variable Marfan phenotype in a five generation family.

Exome sequencing of genes associated with heritable thoracic aortic disease (HTAD) failed to identify a pathogenic variant in a large family with Marfan syndrome (MFS). A genome-wide linkage analysis for thoracic aortic disease identified a peak at 15q21.1, and genome sequencing identified a novel deep intronic FBN1 variant that segregated with thoracic aortic disease in the family (LOD score 2.7) and was predicted to alter splicing. RT-PCR and bulk RNA sequencing of RNA harvested from fibroblasts explanted from the affected proband revealed an insertion of a pseudoexon between exons 13 and 14 of the FBN1 transcript, predicted to lead to nonsense mediated decay (NMD). Treating the fibroblasts with an NMD inhibitor, cycloheximide, greatly improved the detection of the pseudoexon-containing transcript. Family members with the FBN1 variant had later onset aortic events and fewer MFS systemic features than typical for individuals with haploinsufficiency of FBN1. Variable penetrance of the phenotype and negative genetic testing in MFS families should raise the possibility of deep intronic FBN1 variants and the need for additional molecular studies.

Distinct COPD subtypes in former smokers revealed by gene network perturbation analysis.

BACKGROUND: Chronic obstructive pulmonary disease (COPD) varies significantly in symptomatic and physiologic presentation. Identifying disease subtypes from molecular data, collected from easily accessible blood samples, can help stratify patients and guide disease management and treatment. METHODS: Blood gene expression measured by RNA-sequencing in the COPDGene Study was analyzed using a network perturbation analysis method. Each COPD sample was compared against a learned reference gene network to determine the part that is deregulated. Gene deregulation values were used to cluster the disease samples. RESULTS: The discovery set included 617 former smokers from COPDGene. Four distinct gene network subtypes are identified with significant differences in symptoms, exercise capacity and mortality. These clusters do not necessarily correspond with the levels of lung function impairment and are independently validated in two external cohorts: 769 former smokers from COPDGene and 431 former smokers in the Multi-Ethnic Study of Atherosclerosis (MESA). Additionally, we identify several genes that are significantly deregulated across these subtypes, including DSP and GSTM1, which have been previously associated with COPD through genome-wide association study (GWAS). CONCLUSIONS: The identified subtypes differ in mortality and in their clinical and functional characteristics, underlining the need for multi-dimensional assessment potentially supplemented by selected markers of gene expression. The subtypes were consistent across cohorts and could be used for new patient stratification and disease prognosis.

Does the BioBLU 0.3f single-use scale to the BioFlo® 320 reuseable bioreactor on a matched volumetric oxygen mass transfer coefficient?

The volumetric oxygen mass transfer coefficient ([Formula: see text]) is an essential parameter in aerobic high-cell density fermentation where the availability of oxygen to growing microorganisms is a limiting factor. Bioprocess teams looking to scale-up/down between the Eppendorf BioBLU 0.3f single-use vessel and the BioFlo® 320 reusable vessel bioreactors may find it challenging using a matched [Formula: see text]. The maximum [Formula: see text] of the BioFlo® 320 reusable bioreactor was 109 h-1, which was approximately twice that of the BioBLU 0.3f single-use vessel. The results here show no overlap in [Formula: see text] values when both bioreactors were compared and thus conclude that scalability based on [Formula: see text] is not viable. The maximum [Formula: see text] of the Eppendorf BioBLU 0.3f single-use reported here was 47 h-1 compared to that of the manufacturer's value of 2500 h-1, indicating a 53-fold difference. This discrepancy was attributed to the incompatible sulfite addition method used by the manufacturer for estimation.

Elucidating Molecular Design Principles for Charge-Alternating Peptides.

The therapeutic potential of protein drugs has been hindered by difficulties with long-term stability and rapid clearance from the body. Recombinant fusion proteins provide a scalable platform for engineered biologics, whereby a polypeptide domain is appended to alter the physical characteristics of a therapeutic protein and enhance its pharmaceutical viability. Two simple design principles for recombinant fusion proteins, based on the physical properties of the polypeptide domain, have been separately applied to address issues with the stability and delivery of biologics. "Conformationally disordered" peptides, exemplified by the homo amino acid peptide polyG, have been shown to increase the circulation half-life and bioactivity of protein therapeutics in vivo. Superhydrophilic peptides, exemplified by the alternating-charge peptide poly(EK), have been shown to increase the thermostability of proteins in vitro. The combination of superhydrophilicity and conformational disorder in a single fusion peptide could simultaneously address concerns regarding the stability and therapeutic lifetime of biologics. In the current work, we use enhanced sampling molecular dynamics (MD) simulations to investigate the conformational ensemble of poly(EK) and glycine-substituted poly(EK) variants and validate our structural predictions with circular dichroism (CD). We find the (EK)15 peptide exhibits a high propensity for forming antiparallel ß-strand secondary structures, which are stabilized by extensive salt bridging of the positive and negative side chains. MD simulations predict that limited glycine substitutions effectively disrupt the secondary structure and promote disordered conformations at physiologically relevant temperatures. We conclude that the conformational disorder of alternating-charge peptides should be taken into account to improve their suitability for drug delivery applications. We also contribute a computational approach to quantify conformational disorder in polypeptides, which should facilitate the de novo design of effective fusion proteins.

The Prevalence of Traumatic Brain Injury and On-Campus Service Utilization Among Undergraduate Students.

OBJECTIVE: (1) To examine the prevalence of college-aged students on campus who had a history of concussion(s) or traumatic brain injury (TBI) and associated symptoms and (2) to evaluate the rate of campus service utilization among students who had a TBI. SETTING: A representative sampling, randomly selected campus-wide e-mail was sent out with a Qualtrics online survey. PARTICIPANTS: College-aged students (N = 466, n = 429) at a northeastern US university responded and consented to participate in the online survey. DESIGN: Descriptive study using an online survey to evaluate the prevalence of TBI and subsequent utility of campus services. MAIN OUTCOME MEASURES: The survey assessed concussion history and the diagnosing clinician, activity in which it occurred, level of consciousness, associated symptoms, and utilization of campus services for academic accommodation. RESULTS: Prevalence of concussion/TBI in the study population (51.7%) is notably higher than previous reporting in adult and college populations (∼10%-25%). Campus disability and accommodative services were underutilized (10%-13%) and were accessed more readily by the female population. In addition, there were reportable increases in symptom prevalence and service utilization with a greater number of previous concussions. CONCLUSIONS: Awareness and utilization of accommodative services, especially with a history of concussions, need to be improved in order to ensure academic and occupational success of students.

Twoplex 12/13 C6 aniline stable isotope and linkage-specific sialic acid labeling 2D-LC-MS workflow for quantitative N-glycomics.

Quantitative glycomics represents an actively expanding research field ranging from the discovery of disease-associated glycan alterations to the quantitative characterization of N-glycans on therapeutic proteins. Commonly used analytical platforms for comparative relative quantitation of complex glycan samples include MALDI-TOF-MS or chromatographic glycan profiling with subsequent data alignment and statistical evaluation. Limitations of such approaches include run-to-run technical variation and the potential introduction of subjectivity during data processing. Here, we introduce an offline 2D LC-MSE workflow for the fractionation and relative quantitation of twoplex isotopically labeled N-linked oligosaccharides using neutral 12 C6 and 13 C6 aniline (Δmass = 6 Da). Additional linkage-specific derivatization of sialic acids using 4-(4,6-dimethoxy-1,3,5-trizain-2-yl)-4-methylmorpholinium chloride offered simultaneous and advanced in-depth structural characterization. The potential of the method was demonstrated for the differential analysis of structurally defined N-glycans released from serum proteins of patients diagnosed with various stages of colorectal cancer. The described twoplex 12 C6 /13 C6 aniline 2D LC-MS platform is ideally suited for differential glycomic analysis of structurally complex N-glycan pools due to combination and analysis of samples in a single LC-MS injection and the associated minimization in technical variation.

Redefining the Protein-Protein Interface: Coarse Graining and Combinatorics for an Improved Understanding of Amino Acid Contributions to the Protein-Protein Binding Affinity.

The ability to intervene in biological pathways has for decades been limited by the lack of a quantitative description of protein-protein interactions (PPIs). Herein we generate and compare millions of simple PPI models for insight into the mechanisms of specific recognition and binding. We use a coarse-grained approach whereby amino acids are counted in the interface, and these counts are used as binding affinity predictors. We perform lasso regression, a modern regression technique aimed at interpretability, with every possible amino acid combination (over 106 unique feature sets) to select only those amino acid predictors that provide more information than noise. This approach circumvents arbitrary binning and assumptions about the binding environment that obscure other binding affinity models. Aggregated analysis of these models trained at various interfacial cutoff distances informs the roles of specific amino acids in different binding contexts. We find that a simple amino acid count model outperforms detailed intermolecular contact and binned residue type models. We identify the prevalence of serine, glycine, and tryptophan in the interface as particularly important for predicting binding affinity across a range of distance cutoffs. Although current sample size limitations prevent a robust consensus model for binding affinity prediction, our approach underscores the relevance of a residue-based description of the protein-protein interface to increase our understanding of specific interactions.

Quantitative glycomics using liquid phase separations coupled to mass spectrometry.

Post-translational modification of proteins by the attachment of glycans is governed by a variety of highly specific enzymes and is associated with fundamental impacts on the parent protein's physical, chemical and biological properties. The inherent connection between cellular physiology and specific glycosylation patterns has been shown to offer potential for diagnostic and prognostic monitoring of altered glycosylation in the disease state. Conversely, glycoprotein based biopharmaceuticals have emerged as dominant therapeutic strategies in the treatment of intricate diseases. Glycosylation present on these biopharmaceuticals represents a major critical quality attribute with impacts on both pharmacokinetics and pharmacodynamics. The structural variety of glycans, based upon their non-template driven assembly, poses a significant analytical challenge for both qualitative and quantitative analysis. Labile monosaccharide constituents, isomeric species and often low sample availability from biological sources necessitates meticulous sample handling, ultra-high-resolution analytical separation and sensitive detection techniques, respectively. In this article a critical review of analytical quantitation approaches using liquid phase separations coupled to mass spectrometry for released glycans of biopharmaceutical and biomedical significance is presented. Considerations associated with sample derivatisation strategies, ionisation, relative quantitation through isotopic as well as isobaric labelling, metabolic/enzymatic incorporation and targeted analysis are all thoroughly discussed.

Inferring clonal composition from multiple sections of a breast cancer.

Cancers arise from successive rounds of mutation and selection, generating clonal populations that vary in size, mutational content and drug responsiveness. Ascertaining the clonal composition of a tumor is therefore important both for prognosis and therapy. Mutation counts and frequencies resulting from next-generation sequencing (NGS) potentially reflect a tumor's clonal composition; however, deconvolving NGS data to infer a tumor's clonal structure presents a major challenge. We propose a generative model for NGS data derived from multiple subsections of a single tumor, and we describe an expectation-maximization procedure for estimating the clonal genotypes and relative frequencies using this model. We demonstrate, via simulation, the validity of the approach, and then use our algorithm to assess the clonal composition of a primary breast cancer and associated metastatic lymph node. After dividing the tumor into subsections, we perform exome sequencing for each subsection to assess mutational content, followed by deep sequencing to precisely count normal and variant alleles within each subsection. By quantifying the frequencies of 17 somatic variants, we demonstrate that our algorithm predicts clonal relationships that are both phylogenetically and spatially plausible. Applying this method to larger numbers of tumors should cast light on the clonal evolution of cancers in space and time.

SP3-based host cell protein monitoring in AAV-based gene therapy products using LC-MS/MS.

Residual host cell proteins (HCPs) represent a critical quality attribute of biotherapeutic drug products. Workflows enabling reliable HCP detection in monoclonal antibodies and recombinant proteins have been developed, which facilitated process optimization to improve product stability and safety, and allowed setting of acceptance limits for HCP content. However, the detection of HCPs in gene therapy products such as adeno-associated viral (AAV) vectors has been limited. Here, the use of SP3 sample preparation followed by liquid chromatography-mass spectrometry (LC-MS) analysis for HCP profiling in various AAV samples is reported. Suitability of the workflow is demonstrated and provided data constitutes an important reference for future work aiming towards a knowledge-driven improvement of manufacturing conditions and characterization of AAV vector products.

Factors affecting adherence to intranasal treatment for allergic rhinitis: A qualitative study.

Objective: To determine the facilitators of and barriers to adherence to use of intranasal pharmacotherapy (daily intranasal corticosteroids and/or antihistamine, and nasal saline irrigation [NSI]), for allergic rhinitis (AR). Methods: Patients were recruited from an academic tertiary care rhinology and allergy clinic. Semi-structured interviews were conducted after the initial visit and/or 4-6 weeks following treatment. Transcribed interviews were analyzed using a grounded theory, inductive approach to elucidate themes regarding patient adherence to AR treatment. Results: A total of 32 patients (12 male, 20 female; age 22-78) participated (seven at initial visit, seven at follow-up visit, and 18 at both). Memory triggers, such as linking nasal routine to existing daily activities or medications, were identified by patients as the most helpful strategy for adherence at initial and follow-up visits. Logistical obstacles related to NSI (messy, takes time, etc.) was the most common concept discussed at follow-up. Patients modified the regimen based on side effects experienced or perceived efficacy. Conclusions: Memory triggers help patients adhere to nasal routines. Logistical obstacles related to NSI can deter from use. Health care providers should address both concepts during patient counseling. Nudge-based interventions that incorporate these concepts may help improve adherence to AR treatment. Level of Evidence: 2.

The functional impact of rare variation across the regulatory cascade.

Each human genome has tens of thousands of rare genetic variants; however, identifying impactful rare variants remains a major challenge. We demonstrate how use of personal multi-omics can enable identification of impactful rare variants by using the Multi-Ethnic Study of Atherosclerosis, which included several hundred individuals, with whole-genome sequencing, transcriptomes, methylomes, and proteomes collected across two time points, 10 years apart. We evaluated each multi-omics phenotype's ability to separately and jointly inform functional rare variation. By combining expression and protein data, we observed rare stop variants 62 times and rare frameshift variants 216 times as frequently as controls, compared to 13-27 times as frequently for expression or protein effects alone. We extended a Bayesian hierarchical model, "Watershed," to prioritize specific rare variants underlying multi-omics signals across the regulatory cascade. With this approach, we identified rare variants that exhibited large effect sizes on multiple complex traits including height, schizophrenia, and Alzheimer's disease.

Analysis of Human Papilloma Virus Content and Integration in Mucoepidermoid Carcinoma.

Mucoepidermoid Carcinomas (MEC) represent the most common malignancies of salivary glands. Approximately 50% of all MEC cases are known to harbor CRTC1/3-MAML2 gene fusions, but the additional molecular drivers remain largely uncharacterized. Here, we sought to resolve controversy around the role of human papillomavirus (HPV) as a potential driver of mucoepidermoid carcinoma. Bioinformatics analysis was performed on 48 MEC transcriptomes. Subsequent targeted capture DNA sequencing was used to annotate HPV content and integration status in the host genome. HPV of any type was only identified in 1/48 (2%) of the MEC transcriptomes analyzed. Importantly, the one HPV16+ tumor expressed high levels of p16, had high expression of HPV16 oncogenes E6 and E7, and displayed a complex integration pattern that included breakpoints into 13 host genes including PIK3AP1, HIPI, OLFM4,SIRT1, ARAP2, TMEM161B-AS1, and EPS15L1 as well as 9 non-genic regions. In this cohort, HPV is a rare driver of MEC but may have a substantial etiologic role in cases that harbor the virus. Genetic mechanisms of host genome integration are similar to those observed in other head and neck cancers.

Analyzing the Long Time-Scale Dynamics of Uremic Toxins Bound to Sudlow Site II in Human Serum Albumin.

Protein bound uremic toxins (PBUTs), a series of chemicals that remain a challenge for removal strategies used on patients suffering with chronic kidney disease, could be strong candidates for MD study in order to better understand the interactions and time scales associated with binding mode transitions. Currently, traditional dialysis methods cannot satisfactorily remove PBUTs from the bloodstream. This is at least partly due to these toxin's high level of affinity for protein binding sites, particularly the prominent human serum albumin (HSA) and two of its drug binding sites (Sudlow site I and II). We investigate the dynamics of binding site transitions and interactions by MD simulations targeting four well-known toxins: indoxyl sulfate (IS), p-cresyl sulfate (PCS), indole-3-acetic acid (IAA), and hippurate acid (HIP). Long-time scale dynamics are obtained by the use of time-structure independent component analysis (tICA) for dimensionality reduction followed by spectral analysis of a Markov state model (MSM) scored using the generalized matrix Rayleigh quotient (GMRQ). Our results add new insights to prior findings related to the key role of charge-pairing in governing toxin-protein interactions. We find that IAA, the bulkiest hydrophobic toxin studied, observes the slowest process of at least 3 times slower than the smaller, less hydrophobic toxins. In general, we find that the processes slower than 15 ns are correlated with a transition from dominantly hydrophobic interactions deep in the binding pocket to a gain in hydrogen bonding partners near the mouth of the pocket. Our results indicate that aromatic residues such as PHE play a part in a type of toxin stabilization akin to π-stacking. In conclusion, this work presents mechanistic descriptions of interactions/transitions for a set of important PBUTs that bind Sudlow site II on time scales relevant to the underlying binding kinetics of most interest.

2-Dimensional ultra-high performance liquid chromatography and DMT-MM derivatization paired with tandem mass spectrometry for comprehensive serum N-glycome characterization.

Glycosylation is a prominent co- and post-translational modification which contributes to a variety of important biological functions. Protein glycosylation characteristics, particularly N-glycosylation, are influenced by changes in one's pathological state, such as through the presence of disease, and as such, there is great interest in N-glycans as potential disease biomarkers. Human serum is an attractive source for N-glycan based biomarker studies as circulatory proteins are representative of one's physiology, with many serum proteins containing N-glycosylation. The difficulty in comprehensively characterizing the serum N-glycome arises from the absence of a biosynthetic template resulting in great structural heterogeneity and complexity. To help overcome these challenges we developed a 2-dimensional liquid chromatography platform which utilizes offline weak anion exchange (WAX) chromatography in the first dimension and hydrophilic interaction liquid chromatography (HILIC) in the second dimension to separate N-glycans by charge, corresponding to degree of sialylation, and size, respectively. Performing these separations offline enables subsequent derivatization with 4-(4,6-Dimethoxy-1,3,5-triazin-2-yl)-4-methylmorpholinium chloride (DMT-MM) for sialic acid linkage determination and the identification of sialic acid linkage isomers. Subsequent tandem mass spectrometry analysis revealed the identification of 212 complete and partial N-glycan structures including low abundant N-glycans containing acetyl and sulphate modifications. The identifications obtained through this platform were then applied to N-glycans released from a set of stage 3 gastric cancer serum samples obtained from patients before (pre-op) and after (post-op) tumour resection to investigate how the serum N-glycome can facilitate differentiation between the two pathological states.