# Identifying Ubiquitination Sites from Mass Spectrometry Data: A Practical Guide to Di-Gly Remnant Searching

Ubiquitination is a reversible post-translational modification that controls protein degradation, DNA repair, cell signaling, and inflammatory pathways. When a ubiquitin molecule is conjugated to a lysine residue on a target protein, trypsin digestion leaves a two-glycine remnant (Gly-Gly or di-Gly) attached to that lysine. This remnant produces a characteristic mass shift of 114.0429 Da that mass spectrometers can detect. Searching for this di-Gly modification in proteomics data is the standard computational approach for identifying ubiquitination sites at scale. This article explains how to configure database searches for di-Gly remnants, control false discovery rates, validate candidate sites, and interpret results within the limits of current mass spectrometry technology.

The workflow described here applies to researchers who have already generated raw mass spectrometry files from ubiquitination-enriched samples or whole proteomes. It covers the decisions that determine whether a search returns biologically meaningful ubiquitination sites or a list of false positives. The practical outcome is a reproducible search strategy that produces confident site identifications suitable for downstream validation.

## The Biological Basis of Di-Gly Remnant Detection

Ubiquitination involves the covalent attachment of the C-terminal glycine of ubiquitin to the epsilon-amino group of a lysine residue on a substrate protein. The ubiquitin molecule itself is 76 amino acids long and contains seven internal lysines that can be ubiquitinated to form polyubiquitin chains. When trypsin cleaves after arginine and lysine residues, it cuts both the substrate protein and the attached ubiquitin. The digestion products include peptides where the modified lysine retains the last two glycine residues of ubiquitin. This di-Gly remnant is the diagnostic signature used in mass spectrometry searches.

The mass shift created by the di-Gly remnant is 114.0429 Da. This is a specific and measurable change that search engines look for when matching tandem mass spectra to peptide sequences. The remnant is stable under standard collision-induced dissociation and higher-energy collisional dissociation fragmentation methods, which makes it detectable in routine proteomics workflows. The same di-Gly signature also appears on peptides derived from NEDD8 and ISG15 modifications, which share the C-terminal Gly-Gly motif with ubiquitin. Researchers must account for this cross-reactivity when interpreting results.

Mass spectrometry-based proteomics has become the standard approach for systematic characterization of post-translational modifications on a global scale. The development of enrichment strategies, high-performance liquid chromatography fractionation, and bioinformatic tools has enabled researchers to profile ubiquitination sites across entire proteomes. The di-Gly remnant approach is one of several methods for identifying ubiquitination sites, but it is the most widely used because it works with standard trypsin digestion protocols and does not require genetic engineering of the ubiquitin pathway.

The study of post-translational modifications by mass spectrometry has advanced substantially over the past two decades. Methods now exist for identifying, quantifying, and localizing multiple types of modifications including phosphorylation, glycosylation, ubiquitination, SUMOylation, acetylation, and methylation. Each modification type requires specific enrichment strategies, fractionation approaches, detection methods, and bioinformatic tools. The di-Gly remnant search is one component of this larger analytical framework, and understanding its place within the broader proteomics workflow helps researchers make informed decisions about experimental design.

## Database Setup for Di-Gly Modification Searches

The first decision in a di-Gly search is which protein database to use. The database must contain the complete set of proteins expected in the sample, including isoforms and common contaminants. The National Center for Biotechnology Information maintains sequence databases and search systems that researchers can use to build custom protein databases for their organism of interest. The NCBI provides official descriptions of its databases, search systems, and sequence resources that can guide database selection.

For human samples, the UniProt human proteome or the RefSeq human protein database are common choices. For model organisms, the corresponding proteome databases should be used. The database should be reviewed for completeness because missing protein sequences will prevent peptide matching and produce false negative results. Common contaminants such as keratins, trypsin, and serum proteins should be included in the database so that spectra from these proteins are not incorrectly assigned to ubiquitinated peptides.

The search parameters must include the di-Gly modification as a variable modification on lysine. The monoisotopic mass shift is 114.0429 Da. Oxidation of methionine is typically included as a variable modification, and carbamidomethylation of cysteine is included as a fixed modification when iodoacetamide was used during sample preparation. The enzyme specificity should be set to trypsin with allowance for up to two missed cleavages. The precursor mass tolerance depends on the mass spectrometer used, with high-resolution instruments typically using 10 to 20 ppm for precursor ions and 0.02 to 0.05 Da for fragment ions.

The choice between variable and fixed modification settings affects search sensitivity and specificity. Setting di-Gly as a variable modification allows the search engine to consider both modified and unmodified versions of each peptide. This is necessary because not every lysine in a protein is ubiquitinated, and the search must determine which peptides carry the modification. Variable modification searches are computationally more expensive but produce more accurate results than forcing the modification on all lysines.

Database construction requires attention to several practical details. The database should be downloaded from a reliable source and the version recorded. The number of protein entries and the date of download should be documented because databases are updated regularly and results may differ between versions. For organisms with alternative splicing, the database should include all known isoforms because a peptide from one isoform may not match the canonical sequence. For samples that may contain proteins from multiple species, such as host-pathogen interaction studies, the database should include all expected species.

The inclusion of common contaminants is a frequently overlooked step. Keratins from skin and hair, trypsin from the digestion process, and serum proteins from cell culture media can produce abundant spectra that consume search capacity and increase false discovery rates. Including these sequences in the database allows the search engine to assign spectra to contaminants instead of incorrectly matching them to ubiquitinated peptides from the sample. The common Repository of Adventitious Proteins database provides a standardized list of contaminants that can be appended to the protein database.

## False Discovery Rate Control in Di-Gly Searches

False discovery rate control is the most critical quality check in any proteomics search, and di-Gly searches are particularly prone to false positives because the modification is relatively rare. The target-decoy approach is the standard method for estimating and controlling false discovery rates. A decoy database is created by reversing or shuffling the protein sequences in the target database. The search is run against the combined target and decoy database, and the number of matches to decoy sequences is used to estimate the false discovery rate.

The false discovery rate is calculated as the number of decoy hits divided by the number of target hits, with adjustments for the database size. A common threshold is a 1 percent false discovery rate at the peptide level and a 1 percent false discovery rate at the protein level. For ubiquitination site identification, an additional site-level false discovery rate should be calculated because a single peptide can contain multiple lysines, and the search engine must determine which lysine carries the di-Gly remnant.

The localization of the di-Gly modification to a specific lysine residue is a separate challenge from peptide identification. Search engines such as MaxQuant, Proteome Discoverer, and MSFragger provide localization scores that indicate the confidence that the modification is on the reported lysine instead of on another lysine in the same peptide. A localization probability of at least 0.75 is commonly used as a threshold, but higher thresholds such as 0.99 are recommended for reporting individual ubiquitination sites. The false discovery rate for site localization should be controlled separately from the peptide and protein false discovery rates.

The Galaxy Training Network provides accessible workflow training and analysis tutorials that cover false discovery rate estimation and quality control in proteomics data analysis. These resources can help researchers implement reproducible quality control steps in their di-Gly search workflows.

Several practical considerations affect the reliability of false discovery rate estimation. The decoy database should be generated using a method that produces sequences with similar properties to the target sequences, such as reversal instead of random shuffling. The search should be run against the concatenated target and decoy database instead of separate searches, because this ensures that the competition between target and decoy matches is fair. The false discovery rate calculation should account for the fact that the decoy database is the same size as the target database, and the number of decoy hits should be multiplied by a factor to account for the search space.

The choice of false discovery rate threshold depends on the purpose of the experiment. For discovery experiments that aim to identify as many candidate sites as possible, a more permissive threshold such as 5 percent may be acceptable. For experiments that will guide functional validation experiments, a stricter threshold such as 1 percent or even 0.1 percent is appropriate. The threshold should be set before the search is run and documented in the methods section of any publication.

## Search Engine Options and Workflow Choices

Several search engines support di-Gly modification searches, and the choice of engine affects sensitivity, speed, and the format of results. MaxQuant is widely used for label-free quantification and supports di-Gly as a variable modification. Proteome Discoverer with the Sequest HT or MS Amanda nodes is common in laboratories using Thermo Fisher instruments. MSFragger is an open-source search engine that offers fast searches and is integrated into the FragPipe computational platform. Each engine has its own scoring model, and results should not be directly compared across engines without careful consideration of the scoring differences.

The choice of search engine should be based on the laboratory's existing infrastructure, the instrument used for data acquisition, and the need for downstream quantification. MaxQuant provides integrated label-free quantification and is well documented for di-Gly searches. FragPipe offers a modular workflow that includes peptide identification, protein inference, and quantification in a single pipeline. The Bioconductor project provides packages for reproducible genomic and proteomic analysis, including tools for processing search engine results and performing statistical analysis of ubiquitination sites.

The search workflow should be documented and version-controlled to ensure reproducibility. The nf-core documentation describes community standards for pipeline usage, configuration, and reproducible workflow context. Applying these standards to proteomics data analysis ensures that searches can be repeated with the same parameters and that results can be compared across experiments. The Carpentries lessons provide foundational training in computing, data management, shell, and Git that supports reproducible research practices.

Search engine parameters require careful consideration beyond the modification settings. The precursor mass tolerance should match the mass accuracy of the instrument. High-resolution instruments such as Orbitrap and Q-TOF systems can use tolerances of 10 to 20 ppm for precursor ions, while lower-resolution instruments require wider tolerances. The fragment mass tolerance depends on the fragmentation method and the instrument type. The number of missed cleavages allowed should be balanced against the search time and the false discovery rate, with two missed cleavages being a common setting for trypsin digestion.

The scoring model of each search engine affects how peptide spectrum matches are ranked. Some engines use probability-based scoring, while others use empirical scoring based on the number of matched fragment ions. The choice of scoring model can affect the sensitivity and specificity of the search, and researchers should understand the scoring model used by their chosen engine. The results should be reviewed for consistency with the expected fragmentation patterns of the modified peptides.

## Enrichment Strategies and Sample Preparation Considerations

The detection of ubiquitination sites by mass spectrometry depends on the abundance of ubiquitinated peptides relative to unmodified peptides. Ubiquitinated proteins are often low in abundance, and the di-Gly remnant is present on only a small fraction of the total peptide population. Enrichment strategies are therefore essential for deep profiling of ubiquitination sites.

The most common enrichment approach uses antibodies that recognize the di-Gly remnant. These antibodies are used to immunoprecipitate peptides carrying the di-Gly modification after trypsin digestion. This approach, sometimes called di-Gly capture or UbiScan, enriches ubiquitinated peptides from complex mixtures and enables the identification of thousands of ubiquitination sites in a single experiment. The antibody-based approach is specific for the di-Gly remnant and also captures peptides from NEDD8 and ISG15 modifications, which must be considered during interpretation.

Alternative approaches include the use of ubiquitin-binding domains to enrich ubiquitinated proteins before digestion, followed by standard proteomics analysis. These approaches do not provide the same site-level resolution as di-Gly capture because the enrichment occurs at the protein level instead of the peptide level. The choice of enrichment strategy depends on the research question. Protein-level enrichment is suitable for identifying ubiquitinated proteins, while peptide-level di-Gly capture is required for site-level identification.

The NESSI-MS method represents an alternative strategy for studying SUMOylation, a related modification that also targets lysine residues. This method uses antibody-based purification of peptides bearing endogenous SUMO2/3 and allows exact mapping of SUMOylated lysine residues using mass spectrometry. The approach avoids the use of ectopically expressed or mutant SUMO, which can alter the native modification landscape. While NESSI-MS is specific for SUMO instead of ubiquitin, it illustrates the principle that antibody-based peptide enrichment followed by mass spectrometry can map lysine modifications with site-level resolution.

Sample preparation quality directly affects the success of di-Gly searches. The digestion must be complete to produce peptides within the detectable mass range, and the digestion efficiency should be monitored by checking the number of missed cleavages in the search results. The alkylation of cysteine residues must be complete to prevent disulfide bond formation and to ensure that the carbamidomethylation modification is consistent across all peptides. The desalting and fractionation steps must be reproducible to allow comparison across samples.

The amount of starting material is a critical consideration. Ubiquitinated peptides are present at low abundance, and the enrichment step loses material. The starting amount should be sufficient to yield detectable ubiquitinated peptides after enrichment. For cell culture samples, this typically means starting with several milligrams of protein. For tissue samples, the amount depends on the protein content of the tissue and the abundance of ubiquitinated proteins.

## Quantification of Ubiquitination Sites

Quantitative analysis of ubiquitination sites requires a strategy for comparing modification levels across conditions. Label-free quantification is the simplest approach and is supported by MaxQuant and FragPipe. The intensity of the precursor ion for each ubiquitinated peptide is measured and compared across samples. Label-free quantification requires consistent sample preparation and acquisition because variations in digestion efficiency, injection volume, and instrument performance affect peptide intensities.

Stable isotope labeling approaches provide more accurate quantification. SILAC (stable isotope labeling by amino acids in cell culture) involves growing cells in media containing heavy or light amino acids and mixing samples before digestion and analysis. Tandem mass tag labeling allows multiplexed quantification of up to 16 samples in a single experiment. These approaches reduce technical variability but require additional sample preparation steps and are more expensive than label-free quantification.

The quantification of ubiquitination sites is complicated by the fact that the modification is substoichiometric. A protein may be ubiquitinated at a specific lysine in only a small fraction of its molecules, and the ubiquitinated peptide may be present at very low abundance. The dynamic range of the mass spectrometer limits the ability to detect and quantify low-abundance ubiquitinated peptides in the presence of abundant unmodified peptides. Enrichment reduces this problem but introduces its own biases, as the efficiency of the enrichment antibody may vary for different peptide sequences.

The temporal dynamics of post-translational modifications can be studied using quantitative mass spectrometry approaches similar to those developed for phosphorylation. A general mass spectrometric technology for identification and quantitation of phosphorylation sites as a function of stimulus, time, and subcellular location has been applied to detect thousands of phosphorylation sites and determine their temporal dynamics. Similar approaches can be applied to ubiquitination sites, although the enrichment and detection challenges are greater because of the lower abundance of ubiquitinated peptides.

Quantitative comparisons require careful experimental design. The samples to be compared should be processed in parallel to minimize batch effects. The order of acquisition should be randomized to avoid instrument drift affecting the results. Technical replicates should be included to assess the variability of the measurement. The statistical analysis should account for the missing values that are common in label-free quantification, where a peptide may be detected in one sample but not in another.

The normalization of quantitative data is an important step. The total peptide intensity can vary between samples due to differences in the amount of material injected or the efficiency of digestion. Normalization methods adjust for these differences to allow comparison of modification levels across samples. The choice of normalization method depends on the experimental design and the distribution of the data.

## Site Localization and Validation

The identification of the specific lysine residue that carries the di-Gly remnant requires careful evaluation of the tandem mass spectrum. The fragment ions must be consistent with the modification being on one lysine instead of another. Search engines provide localization scores that quantify the confidence in the site assignment, but these scores should be interpreted with caution.

The presence of fragment ions that bracket the modified lysine provides direct evidence for the site assignment. If the spectrum contains fragment ions on both sides of the modified lysine, the localization is confident. If the modified lysine is near the peptide terminus or if the peptide contains multiple lysines, the localization may be ambiguous. Manual inspection of spectra for high-value sites is recommended, particularly for sites that will be subjected to functional validation.

Validation of ubiquitination sites can be performed using targeted mass spectrometry approaches such as parallel reaction monitoring or selected reaction monitoring. These approaches use the mass spectrometer to specifically detect and quantify the ubiquitinated peptide in a complex mixture. Targeted approaches provide higher sensitivity and specificity than data-dependent acquisition and can confirm the presence of a specific ubiquitination site across multiple samples.

Site-directed mutagenesis is the gold standard for functional validation of ubiquitination sites. Mutation of the identified lysine to arginine prevents ubiquitination at that site and allows assessment of the functional consequences. This approach has been used extensively to study the role of ubiquitination in protein degradation, signaling, and other cellular processes. The identification of ubiquitination sites by mass spectrometry provides the direct evidence needed to design these mutagenesis experiments, which previously relied on indirect evidence from mutation studies.

The validation process should be planned before the discovery experiment is completed. The criteria for selecting sites for validation should be defined in advance, including the localization score threshold, the false discovery rate threshold, and the biological relevance of the candidate protein. The validation experiments should be designed to test specific hypotheses about the function of the ubiquitination site.

## Common Failure Patterns in Di-Gly Searches

Several recurring problems can compromise di-Gly searches. The most common is the misidentification of the di-Gly modification on lysine when the modification is actually on the peptide N-terminus or on another residue. The di-Gly remnant can also be confused with other modifications that produce similar mass shifts. The mass shift of 114.0429 Da is close to the mass of several amino acid combinations, and search engines may assign the modification to the wrong residue if the spectrum quality is poor.

Another common failure is the inclusion of too many variable modifications in the search. Each variable modification increases the search space and the number of candidate peptides, which increases the false discovery rate. The di-Gly modification should be the only variable modification on lysine unless there is a specific reason to include others. Oxidation of methionine and acetylation of protein N-termini are common additions, but each additional modification requires careful consideration of its effect on the false discovery rate.

The use of an incomplete protein database is a frequent cause of false negatives. If the database does not contain the protein sequence for a ubiquitinated protein, the spectra from that protein will not be matched. This is particularly problematic for organisms with incomplete genome annotations or for samples that contain proteins from multiple species. The database should be reviewed for completeness and updated when new annotations become available.

The failure to account for the di-Gly remnant on peptides from NEDD8 and ISG15 can lead to incorrect biological interpretation. These modifications produce the same di-Gly remnant after trypsin digestion, and the search engine cannot distinguish between them. Researchers studying ubiquitination should be aware that a subset of the identified sites may be from NEDD8 or ISG15 substrates. Additional experiments, such as immunodepletion of NEDD8 or ISG15, may be needed to confirm the ubiquitin-specific sites.

Poor spectrum quality is a common underlying cause of search failures. Spectra with low signal-to-noise ratios, incomplete fragmentation, or incorrect charge state assignment produce unreliable peptide spectrum matches. The quality of the spectra should be assessed before the search, and low-quality spectra should be removed or the acquisition parameters should be adjusted. The number of spectra that match to decoy sequences is a useful indicator of spectrum quality, with high decoy match rates suggesting poor quality.

The choice of mass tolerance can also cause failures. If the precursor mass tolerance is too narrow, peptides with mass calibration errors will not be matched. If the tolerance is too wide, the search space increases and the false discovery rate rises. The mass calibration should be checked before the search, and the tolerance should be set based on the observed mass accuracy of the instrument.

## Limitations of Di-Gly Remnant Searching

The di-Gly remnant approach has inherent limitations that affect the interpretation of results. The approach only detects ubiquitination sites that are present in the sample at sufficient abundance and that produce peptides within the mass range detectable by the mass spectrometer. Very large or very small peptides may not be detected, and peptides with certain sequences may not ionize efficiently.

The approach cannot distinguish between different types of ubiquitin chains. The di-Gly remnant is the same whether the ubiquitin is attached as a monomer or as part of a polyubiquitin chain, and it does not reveal the linkage type of polyubiquitin chains. Determining ubiquitin chain architectures requires additional mass spectrometry approaches that analyze the ubiquitin peptides themselves instead of the substrate peptides. The identification of ubiquitination sites and determination of ubiquitin chain architectures are related but distinct analytical challenges.

The approach does not provide information about the dynamics of ubiquitination. A single time point measurement reveals which sites are ubiquitinated at that moment but does not indicate the rate of ubiquitination or deubiquitination. Time course experiments are needed to study the dynamics of ubiquitination, and these experiments require careful experimental design to capture the relevant time points.

The sensitivity of the approach is limited by the abundance of ubiquitinated peptides and the efficiency of enrichment. Low-abundance ubiquitination sites may be missed, and the results may be biased toward highly abundant ubiquitinated proteins. The false discovery rate control helps to ensure that reported sites are confident, but it does not address the issue of false negatives.

The approach provides limited information about the stoichiometry of ubiquitination. The detection of a ubiquitination site indicates that the modification is present but does not quantify the fraction of the protein that is modified. The stoichiometry of ubiquitination is important for understanding the functional significance of the modification, but it is difficult to measure with current mass spectrometry approaches.

The interpretation of ubiquitination sites requires consideration of the broader biological context. Ubiquitination can signal protein degradation through the proteasome, but it also regulates DNA repair, cell cycle progression, inflammatory signaling, and other processes. The identification of a ubiquitination site on a protein does not indicate the functional consequence of that modification. The integration of ubiquitination site data with functional experiments is necessary to understand the biological role of the modification.

## Records and Quality Control Documentation

The documentation of search parameters and quality control metrics is essential for reproducible di-Gly searches. The search parameters should be recorded in a format that allows the search to be repeated exactly. This includes the database version, the search engine version, the modification settings, the mass tolerances, and the false discovery rate thresholds.

The quality control metrics should include the number of spectra acquired, the number of peptide spectrum matches, the number of unique peptides, the number of proteins identified, and the false discovery rates at each level. The distribution of peptide scores and the number of decoy hits should be reviewed to ensure that the search performed as expected. The retention time alignment and mass calibration should be checked for consistency across samples.

The raw mass spectrometry data and the search results should be stored in a repository that supports data sharing and reanalysis. The PRIDE database and the MassIVE repository are common choices for proteomics data deposition. The deposition of raw data and search results enables other researchers to reanalyze the data and verify the reported ubiquitination sites.

The EMBL-EBI Training resources provide learning pathways for bioinformatics data resources and practical analysis education. These resources can help researchers develop the skills needed to document and share proteomics data effectively. The training materials cover data deposition, quality control, and reproducible analysis practices.

The documentation should include a description of the sample preparation protocol, the mass spectrometry acquisition parameters, and the data analysis workflow. The version numbers of all software tools should be recorded, along with the parameters used for each step. The documentation should be sufficient for another researcher to reproduce the analysis from the raw data.

The quality control metrics should be reviewed at each step of the workflow. The number of spectra acquired should be consistent with the expected performance of the instrument. The number of peptide spectrum matches should be reasonable for the sample type and the enrichment strategy. The false discovery rates should be within the expected range, and any deviations should be investigated before proceeding with the analysis.

## Professional Escalation Criteria

Researchers should seek expert assistance when the di-Gly search results are inconsistent with biological expectations or when quality control metrics indicate problems with the search. The following situations warrant consultation with a proteomics bioinformatician or a mass spectrometry facility director:

The false discovery rate cannot be controlled at the desired threshold despite adjustment of search parameters. This may indicate problems with the sample preparation, the mass spectrometry acquisition, or the database.

The localization scores for the identified ubiquitination sites are consistently low, suggesting that the modification sites cannot be confidently assigned. This may require changes to the fragmentation method or the use of alternative search engines.

The number of identified ubiquitination sites is much lower than expected based on the literature for the sample type. This may indicate problems with the enrichment, the digestion, or the mass spectrometry acquisition.

The results contain a high proportion of ubiquitination sites on proteins that are not expected to be ubiquitinated based on their cellular localization or function. This may indicate contamination or cross-reactivity of the enrichment antibody.

The quantification results are not reproducible across technical replicates. This may indicate problems with sample preparation, acquisition, or data processing that require expert troubleshooting.

The interpretation of ubiquitination sites in the context of specific biological pathways may require consultation with experts in ubiquitin biology. The identification of a ubiquitination site is the first step, and the functional significance of the site requires experimental validation.

## At a Glance

| Decision Point | Recommended Setting | Purpose |
|---|---|---|
| Protein database | Complete proteome for the organism, including isoforms and contaminants | Ensures all possible peptide matches are considered |
| Di-Gly modification | Variable modification on lysine, mass shift 114.0429 Da | Allows detection of ubiquitinated and unmodified peptides |
| False discovery rate | 1 percent at peptide and protein level, site-level localization probability at least 0.75 | Controls false positives while maintaining sensitivity |
| Search engine | MaxQuant, Proteome Discoverer, or MSFragger based on laboratory infrastructure | Provides peptide identification and localization scores |
| Enrichment | Anti-di-Gly antibody capture after trypsin digestion | Enriches ubiquitinated peptides for deep profiling |
| Validation | Manual spectrum inspection and targeted mass spectrometry for high-value sites | Confirms site localization before functional studies |

## Practical Workflow for Di-Gly Searching

The following workflow provides a structured approach to di-Gly searching that can be adapted to different search engines and laboratory protocols.

Step 1: Review the raw mass spectrometry data for quality. Check the number of spectra acquired, the precursor mass accuracy, and the fragmentation quality. Remove samples with poor acquisition metrics before proceeding with the search.

Step 2: Select the protein database. Download the complete proteome for the organism of interest from a reliable source such as NCBI or UniProt. Add common contaminants and verify that the database contains the expected proteins.

Step 3: Configure the search parameters. Set trypsin as the digestion enzyme with up to two missed cleavages. Set carbamidomethylation of cysteine as a fixed modification. Set oxidation of methionine and di-Gly on lysine as variable modifications. Set the precursor mass tolerance based on the instrument specifications.

Step 4: Run the search against the target and decoy databases. Verify that the search completed without errors and that the number of peptide spectrum matches is reasonable for the sample type.

Step 5: Apply false discovery rate filtering. Set the peptide and protein false discovery rate thresholds to 1 percent. Review the number of decoy hits to confirm that the false discovery rate estimation is reliable.

Step 6: Evaluate site localization. Review the localization scores for the identified ubiquitination sites. Apply a localization probability threshold of at least 0.75 for reporting sites. Manually inspect spectra for sites that will be used in downstream experiments.

Step 7: Perform quantification if required. Use label-free or stable isotope labeling approaches to compare ubiquitination levels across conditions. Verify that the quantification results are reproducible across technical replicates.

Step 8: Document the search parameters and results. Record the database version, search engine version, modification settings, mass tolerances, and false discovery rate thresholds. Store the raw data and search results in a repository.

Step 9: Validate high-value sites. Use targeted mass spectrometry or site-directed mutagenesis to confirm the ubiquitination sites that are most important for the research question.

## Interpretation of Ubiquitination Site Data

The interpretation of ubiquitination site data requires consideration of the biological context. Ubiquitination can signal protein degradation through the proteasome, but it also regulates DNA repair, cell cycle progression, inflammatory signaling, and other processes. The identification of a ubiquitination site on a protein does not indicate the functional consequence of that modification.

The stoichiometry of ubiquitination is an important consideration. Most ubiquitination sites are substoichiometric, meaning that only a fraction of the protein molecules are ubiquitinated at a given site at a given time. The detection of a ubiquitination site by mass spectrometry indicates that the modification is present but does not quantify the fraction of the protein that is modified.

The relationship between ubiquitination and other post-translational modifications should be considered. Proteins can be modified by phosphorylation, acetylation, methylation, and ubiquitination at the same or nearby residues. The interplay between these modifications can affect protein function in complex ways. The study of post-translational modifications by mass spectrometry has revealed that many proteins contain multiple modification sites that show different kinetics and respond to different stimuli.

The identification of ubiquitination sites on proteins involved in specific pathways can provide insights into the regulation of those pathways. For example, the identification of ubiquitination sites on proteins involved in adipogenesis has revealed mechanisms of obesity regulation. The study of annexin A1 in adipose tissue demonstrated that the protein competes with MYCBP2 for interaction with PDLIM7, which affects the ubiquitination and degradation of SMAD4 and regulates adipogenesis. This type of mechanistic insight requires the integration of ubiquitination site data with functional experiments.

The interpretation of ubiquitination data should also consider the subcellular localization of the modified proteins. Ubiquitination occurs in multiple cellular compartments, and the functional consequences of the modification may depend on the location of the protein. The enrichment strategy and the sample preparation method should be designed to capture the relevant subcellular fractions.

## Related Analytical Approaches

The di-Gly remnant approach is one of several mass spectrometry strategies for studying ubiquitination and related modifications. The choice of approach depends on the research question and the available resources.

The identification of ubiquitination sites and determination of ubiquitin chain architectures by mass spectrometry is a related but distinct analytical challenge. The di-Gly remnant approach identifies the sites of ubiquitination on substrate proteins, while the analysis of ubiquitin chain architectures requires the identification of ubiquitin peptides that contain the di-Gly remnant on one of the seven internal lysines of ubiquitin. The linkage type of polyubiquitin chains determines the functional outcome of ubiquitination, with K48-linked chains typically signaling degradation and K63-linked chains involved in signaling and DNA repair.

The study of SUMOylation requires different analytical approaches because the SUMO remnant after trypsin digestion is larger than the di-Gly remnant. The NESSI-MS method uses antibody-based purification of peptides bearing endogenous SUMO2/3 and allows exact mapping of SUMOylated lysine residues. This approach avoids the artifacts associated with ectopically expressed or mutant SUMO and provides a more physiologically relevant view of SUMOylation.

The analysis of other lysine modifications such as acetylation and methylation requires different search parameters and enrichment strategies. Acetylation produces a mass shift of 42.0106 Da on lysine, and methylation produces a mass shift of 14.0157 Da. These modifications can be searched using similar workflows with the appropriate modification settings and enrichment antibodies.

The choice between different analytical approaches should be based on the specific research question. The di-Gly remnant approach is appropriate for identifying ubiquitination sites on a proteome-wide scale. The analysis of ubiquitin chain architectures is appropriate for understanding the functional consequences of ubiquitination. The NESSI-MS approach is appropriate for studying SUMOylation in its native context.

## Training and Reproducibility Resources

The successful implementation of di-Gly searching requires training in proteomics data analysis and reproducible research practices. The EMBL-EBI Training resources provide learning pathways for bioinformatics data resources and practical analysis education. These resources cover the use of public databases, data deposition, and analysis tools.

The Galaxy Training Network provides accessible workflow training and analysis tutorials that cover proteomics data analysis and quality control. These tutorials provide step-by-step instructions for running searches and interpreting results in the Galaxy platform, which supports reproducible analysis through its workflow system.

The Bioconductor project provides packages for reproducible genomic and proteomic analysis. These packages can be used to process search engine results, perform statistical analysis, and generate visualizations of ubiquitination site data. The Bioconductor documentation covers installation, usage, and workflow development.

The nf-core documentation describes community standards for pipeline usage, configuration, and reproducible workflow context. These standards can be applied to proteomics data analysis to ensure that workflows are portable, well documented, and reproducible across different computing environments.

The Carpentries lessons provide foundational training in computing, data management, shell, and Git. These skills are essential for managing the large data files generated by mass spectrometry experiments and for implementing reproducible analysis workflows.

The training resources should be used to develop a standard operating procedure for di-Gly searches in the laboratory. The standard operating procedure should include the specific parameters used, the quality control checks performed, and the criteria for accepting or rejecting results. The standard operating procedure should be reviewed and updated regularly as new tools and methods become available.

## Frequently Asked Questions

### What is the di-Gly remnant and why is it used for ubiquitination site identification?

The di-Gly remnant is the two-glycine residue that remains attached to a lysine after trypsin digestion of a ubiquitinated protein. Trypsin cleaves after arginine and lysine residues in both the substrate protein and the attached ubiquitin molecule. The digestion leaves the last two glycines of ubiquitin covalently attached to the modified lysine. This remnant produces a characteristic mass shift of 114.0429 Da that mass spectrometry search engines can detect. The di-Gly remnant is the standard diagnostic signature for ubiquitination sites because it is stable under standard fragmentation methods and can be searched as a variable modification on lysine.

### How do I choose the right protein database for a di-Gly search?

The protein database should contain the complete set of proteins expected in the sample, including isoforms and common contaminants. For human samples, the UniProt human proteome or the RefSeq human protein database are common choices. The database should be reviewed for completeness because missing protein sequences will prevent peptide matching and produce false negative results. Common contaminants such as keratins, trypsin, and serum proteins should be included so that spectra from these proteins are not incorrectly assigned to ubiquitinated peptides. The NCBI provides official descriptions of its databases and search systems that can guide database selection.

### What false discovery rate thresholds should I use for di-Gly searches?

A common threshold is 1 percent false discovery rate at the peptide level and 1 percent at the protein level. For ubiquitination site identification, an additional site-level false discovery rate should be calculated because a single peptide can contain multiple lysines. A localization probability of at least 0.75 is commonly used as a threshold for reporting sites, with higher thresholds such as 0.99 recommended for individual sites that will be subjected to functional validation. The false discovery rate is estimated using the target-decoy approach, where a decoy database is created by reversing or shuffling the protein sequences.

### How can I distinguish ubiquitination from NEDD8 and ISG15 modifications?

The di-Gly remnant is identical for ubiquitin, NEDD8, and ISG15 because all three modifications share the C-terminal Gly-Gly motif. The search engine cannot distinguish between these modifications based on the di-Gly remnant alone. Researchers studying ubiquitination should be aware that a subset of the identified sites may be from NEDD8 or ISG15 substrates. Additional experiments, such as immunodepletion of NEDD8 or ISG15, may be needed to confirm the ubiquitin-specific sites.

### What is the difference between site identification and site localization?

Site identification refers to the assignment of a peptide spectrum match to a peptide sequence that contains the di-Gly modification. Site localization refers to the assignment of the modification to a specific lysine residue within that peptide. A peptide can contain multiple lysines, and the search engine must determine which lysine carries the di-Gly remnant. Localization scores quantify the confidence in the site assignment. The presence of fragment ions that bracket the modified lysine provides direct evidence for the site assignment, while ambiguous localization requires manual spectrum inspection.

### How do I validate ubiquitination sites identified by mass spectrometry?

Validation can be performed using targeted mass spectrometry approaches such as parallel reaction monitoring or selected reaction monitoring. These approaches specifically detect and quantify the ubiquitinated peptide in a complex mixture. Site-directed mutagenesis is the gold standard for functional validation. Mutation of the identified lysine to arginine prevents ubiquitination at that site and allows assessment of the functional consequences. Manual inspection of spectra for high-value sites is recommended, particularly for sites that will be subjected to functional validation.

### What are the main limitations of di-Gly remnant searching?

The approach only detects ubiquitination sites that are present at sufficient abundance and that produce peptides within the mass range detectable by the mass spectrometer. The approach cannot distinguish between different types of ubiquitin chains, and it does not reveal the linkage type of polyubiquitin chains. The approach does not provide information about the dynamics of ubiquitination, and time course experiments are needed to study the dynamics. The sensitivity is limited by the abundance of ubiquitinated peptides and the efficiency of enrichment, and low-abundance sites may be missed.

### When should I seek expert assistance with di-Gly searching?

Expert assistance is warranted when the false discovery rate cannot be controlled at the desired threshold, when localization scores are consistently low, when the number of identified sites is much lower than expected, when results contain a high proportion of sites on proteins not expected to be ubiquitinated, or when quantification results are not reproducible across technical replicates. Consultation with a proteomics bioinformatician or a mass spectrometry facility director can help troubleshoot these problems.

## Related Bioinformatics Guides

- [Proteomics Mass Spectrometry: From Sample Preparation to Data Analysis](/knowledge/bioinformatics/proteomics-mass-spectrometry-from-sample-preparation-to-data-analysis)
- [Mass Spectrometry-Based Proteomics: Data Analysis Pipelines and Tools](/knowledge/bioinformatics/mass-spectrometry-based-proteomics-data-analysis-pipelines-and-tools)
- [Mass Spectrometry Protein Identification: From Raw Spectra to Confident Hits](/knowledge/bioinformatics/mass-spectrometry-protein-identification-from-raw-spectra-to-confident-hits)
- [Metabolomics Data Analysis in R: A Practical Workflow](/knowledge/bioinformatics/metabolomics-data-analysis-in-r-a-practical-workflow)
- [Microbiome Data Analysis in R: A Practical Guide for Compositional Data](/knowledge/bioinformatics/microbiome-data-analysis-in-r-a-practical-guide-for-compositional-data)

## References and Further Reading

- [NCBI Data Resources](https://www.ncbi.nlm.nih.gov/). National Center for Biotechnology Information.
- [EMBL-EBI Training](https://www.ebi.ac.uk/training). European Bioinformatics Institute.
- [Bioconductor](https://bioconductor.org/). Bioconductor Project.
- [Galaxy Training Network](https://training.galaxyproject.org/). Galaxy Project.
- [nf-core Documentation](https://nf-co.re/docs). nf-core.
- [The Carpentries Lessons](https://carpentries.org/lessons). The Carpentries.
- [Native and Endogenous SUMO Site Identification Using Mass Spectrometry (NESSI-MS).](https://pubmed.ncbi.nlm.nih.gov/40875113). Methods in molecular biology (Clifton, N.J.), 2025.
- [Identification, Quantification, and Site Localization of Protein Posttranslational Modifications via Mass Spectrometry-Based Proteomics.](https://pubmed.ncbi.nlm.nih.gov/27975226). Advances in experimental medicine and biology, 2016.
- [Global, in vivo, and site-specific phosphorylation dynamics in signaling networks.](https://pubmed.ncbi.nlm.nih.gov/17081983). Cell, 2006.
- [Annexin A1 binds PDZ and LIM domain 7 to inhibit adipogenesis and prevent obesity.](https://pubmed.ncbi.nlm.nih.gov/39174522). Signal transduction and targeted therapy, 2024.
- [Identification of ubiquitination sites and determination of ubiquitin-chain architectures by mass spectrometry.](https://pubmed.ncbi.nlm.nih.gov/16338362). Methods in enzymology, 2005.

> This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.