Putting humpty dumpty back together again: what does protein quantification mean in bottom-up proteomics?
Open accessAbstract
Bottom-up proteomics provides peptide measurements and has been invaluable for moving proteomics into large-scale analyses. Commonly, a single quantitative value is reported for each protein-coding gene by aggregating peptide quantities into protein groups following protein inference or parsimony. However, given the complexity of both RNA splicing and post-translational protein modification, it is overly simplistic to assume that all peptides that map to a singular protein-coding gene will demonstrate the same quantitative response. By assuming that all peptides from a protein-coding sequence are representative of the same protein, we may miss the discovery of important biological differences. To capture the contributions of existing proteoforms, we need to reconsider the practice of aggregating protein values to a single quantity per protein-coding gene.
AI snapshot
Written from the paper’s full text.
TL;DR
Bottom-up proteomics typically aggregates peptide measurements into a single quantity per protein coding gene, but this assumes all peptides from a gene respond similarly despite extensive splicing and post-translational modification. The paper argues that this aggregation can average away biologically meaningful proteoform-specific differences and recommends reporting peptide-level information or multiple peptide groups rather than blindly merging to one protein value.
Key findings
- It is estimated there may be up to 100 proteoforms per protein on average, based on possible coding-sequence variation and PTMs; a given tryptic peptide can derive from multiple proteoforms, so peptide measurements are convolutions of the proteoforms present.
- Aggregating peptides to a protein value suppresses outliers/noise and reduces the number of hypothesis tests, but protein groups with greater peptide coverage tend to be statistically different less often, with fold-change consistently trending toward zero (Figure 3).
- Examples show discordant peptide behavior: amyloid precursor protein/amyloid-beta has Aβ40 and Aβ42 plus over 20 additional amyloid-beta proteoforms detected in Alzheimer’s brain samples; tau has at least 95 discovered PTMs; proglucagon is processed into up to nine polypeptide products including glucagon and GLP1.
- The paper recommends no longer blindly merging all peptides into a single gene-level quantity; alternatives include keeping peptides as independent measurements or combining peptides that map to the same gene and co-vary across conditions into 1 to N quantities per protein, while keeping individual peptide quantities accessible.
Methods
- Not stated in the paper; this is a perspective/commentary rather than a primary research study, with no new experimental design, sample, or statistical analysis reported.
- The text synthesizes prior literature and uses illustrative examples and figures, including technical replicate injections of cerebrospinal fluid digests, Alzheimer’s disease biomarker examples, and a Parkinson’s disease cerebrospinal fluid proteomics example.
Limitations
- Stated: Keeping all peptides as independent measurements may be difficult for biological interpretation, especially in less studied systems, and reduced statistical power for differential abundance testing on tens of thousands of peptides compared with thousands of protein groups will likely yield fewer significant differences.
- Stated: A co-variation-based peptide grouping approach could depend heavily on multiple conditions with enough biological replicates and high reproducibility, and may not be suitable for proteins with low peptide coverage.
- Evident: As a perspective/commentary, the paper reports no primary data or formal validation of its proposed alternatives; its arguments rely on cited examples and illustrative figures.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
0 of 33 cited sources checked · 0 supported · 0 flagged
- Findingown data
In the same cerebrospinal fluid experiment, peptides detected and quantified from GAPDH show the same trend across peptides.
“In contrast, if we look at peptides detected and quantified from GAPDH protein in the same CSF experiment we observe the same trend across peptides.”
Quote in the paperSubstantial claim - Findingnone given
Across different proteomics data types, data scales, and sum-based or reference-based quantification, fold-change consistently trends toward zero.
“Despite the different types of proteomics data, the difference in scale of the data, and using either a sum-based or reference-based quantification, the fold-change consistently trends towards zero.”
Quote in the paperNot a substantial claim - Findingreasoning
Protein groups with more observed peptides tend to be statistically different less often than protein groups with fewer peptides.
“Interestingly, protein groups with greater numbers of peptides observed tend to be statistically different less often than protein groups with fewer peptides (Figure 3).”
Quote in the paperSubstantial claim - Findingown data
In tryptic peptide data, some biologically relevant differences would not be accurately represented if peptide measures were combined into a single protein-level value.
“If we look at data from tryptic peptides, we see that some biologically relevant differences would not be accurately represented if our peptide measures are combined to a singular protein level (Figure 4).”
Quote in the paperSubstantial claim - Findingown data
In technical replicate injections of cerebrospinal fluid digests, peptide-level values show more variability than protein-level values.
“For example, we observe more variability in the peptide level values compared to the protein level values in technical replicate injections of cerebrospinal fluid (CSF) digests (Figure 2).”
Quote in the paperSubstantial claim
- Findingown data
Tryptic peptides mapping to the amyloid beta region show a different abundance profile from tryptic peptides mapping to other regions of the protein.
“Specifically, in tryptic peptides mapping to the region of the amyloid beta sequence we observe a different abundance profile compared to tryptic peptides mapping to other regions of the protein.”
Quote in the paperSubstantial claim - Findingown data
Summing all peptide measures mapping to the secretogranin 2 protein-coding sequence misses differences between experimental groups for several individual peptides.
“As demonstrated in figure 5, if we sum all peptide measures that map to the protein coding sequence of secretogranin 2, then we miss the differences between experimental groups for several of the individual peptides.”
Quote in the paperSubstantial claim - Interpretationreasoning
Aggregating all tryptic peptide measures assumes they are all derived from the unprocessed state, which may not be accurate for peptides mapping to amyloid precursor protein.
“If we aggregate all the tryptic peptide measures, we are assuming they are all derived from the unprocessed state, which may not be the most accurate assumption for peptides mapping to amyloid precursor protein.”
Quote in the paperSubstantial claim - Interpretationreasoning
Because RNA splicing and post-translational modification create complexity, it is too simple to assume that every peptide mapping to one protein-coding gene will show the same quantitative response.
“However, given the complexity of both RNA splicing and post-translational protein modification, it is overly simplistic to assume that all peptides that map to a singular protein coding gene will demonstrate the same quantitative response.”
Quote in the paperSubstantial claim - Interpretationreasoning
The loss of biological effect when aggregating peptides into a single protein value is analogous to single-cell versus bulk tissue analysis.
“The problem of reducing the biological effect when aggregating peptide quantities into a single protein value is analogous to single cell versus bulk tissue analysis.”
Quote in the paperSubstantial claim - Premisecitation
A study of cerebrospinal fluid in Parkinson's disease found specific tryptic peptides differentially abundant in affected individuals compared with healthy age-matched controls.
“For example, a study of cerebrospinal fluid in Parkinson’s disease found that specific tryptic peptides are differentially abundant in affected individuals compared to healthy, age-matched controls.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [20] source
Not checked in the free run.
- Premisecitation
Top-down analysis has observed that GAPDH does not have many proteoforms.
“Interestingly, GAPDH has been observed not to have many proteoforms by top-down analysis.(21)”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [21] source
Not checked in the free run.
- Interpretationreasoning
The same averaging problem applies when protein-level quantities are reported from peptides.
“The same is true with reporting protein level quantities from peptides.”
Quote in the paperSubstantial claim - Interpretationreasoning
Even assays using multiple affinity reagents or many tryptic peptides to different domains or modified sites likely undersample the proteoform species in a sample.
“Even assays that use multiple affinity reagents or many tryptic peptides to different domains or modified sites of a protein will likely provide an undersampling of the proteoform species in the sample.”
Quote in the paperSubstantial claim - Interpretationreasoning
Aggregating peptides to one measure per protein-coding sequence accurately reflects peptide-level measurements only when all peptides agree.
“Instead, aggregating peptides to a single measure per protein coding sequence only accurately reflects the peptide level measurements if all peptides are in agreement (Figure 5).”
Quote in the paperSubstantial claim - Interpretationreasoning
When interpreting bottom-up proteomics data, conclusions can be made only about detected peptides, not about the proteoforms from which they originate.
“When interpreting bottom-up proteomics data we are only able to make conclusions about the peptides we detect, not the proteoforms from which they originate.”
Quote in the paperSubstantial claim - Interpretationreasoning
When coverage is low, differences in peptides specific to a subset of proteoforms may or may not be captured.
“Conversely, if coverage is low, differences in peptides specific to a subset of proteoforms may or may not be captured.”
Quote in the paperSubstantial claim - Interpretationreasoning
Differences could disappear entirely in a bulk sample because the effect on each cell could be very different.
“Furthermore, differences could disappear entirely in the bulk sample because the effect on each cell could be very different.”
Quote in the paperSubstantial claim - Interpretationreasoning
Peptides mapping to a protein-coding sequence need not be present at the same quantity in a biological system.
“These examples highlight that just because peptides map to a protein coding sequence, it does not mean that the peptides will be present at the same quantity in a biological system.”
Quote in the paperSubstantial claim - Premisecitation
On average, there may be as many as 100 proteoforms per protein, based on possible coding-sequence variations and post-translational modifications.
“it is estimated that there may be up to 100 proteoforms per protein on average.(4,5)”
- Also stated in: § Limitations of assuming a single quantity per protein coding gene · hedged
“The estimate of an average of 100 proteoforms per protein coding gene may seem large until one investigates just how many modified peptides have been detected for most proteins.(4)”
Quote in the paperSubstantial claim0 of 2 sources checked - Also stated in: § Limitations of assuming a single quantity per protein coding gene · hedged
- Premisecitation
Genomic sequence variation can lead to differing peptide sequences in the population.
“Genomic sequence variation can lead to differing peptide sequences in the population.(25–27)”
Quote in the paperSubstantial claim0 of 3 sources checked - Interpretationreasoning
To the extent that peptides from a protein-coding sequence are not present at the same quantity, statistical power to connect phenotype to proteomic data will be reduced.
“To the extent this is true, statistical power will be reduced in connecting phenotype to proteomic data.”
Quote in the paperSubstantial claim - Interpretationreasoning
Reporting a protein-level value makes differences difficult to compare across studies because the measured peptides may be important for interpreting the results.
“If a value is reported at a protein level it makes these differences difficult to compare across studies since the peptides measured may be important for interpreting the results.”
Quote in the paperSubstantial claim - Interpretationreasoning
Averaging bulk tissue results loses the ability to assess the degree of heterogeneity in the effect.
“By averaging the results from bulk tissue, the ability to assess the degree of heterogeneity on the effect will be lost.”
Quote in the paperSubstantial claim - Premisecitation
Several proposed methods exclude peptide measurements from the aggregate protein quantity when they are outliers relative to other peptides mapping to the same protein-coding gene.
“The approach of several proposed methods focuses on the exclusion of peptide measurements from inclusion in the aggregate protein quantity if they are outliers from other peptide measurements mapping to the same protein coding gene.(28–33)”
Quote in the paperSubstantial claim0 of 6 sources checked - Interpretationreasoning
Any method that constrains complex proteoforms into a single quantitative value per protein-coding gene may miss many underlying differences.
“It should be noted that any method that constrains complex proteoforms into a single quantitative value per protein coding gene may miss many of the underlying differences.”
Quote in the paperSubstantial claim - Interpretationreasoning
A change might be reflected only in a proteoform that is best represented by the quantity of a single peptide.
“A change might only be reflected in a proteoform that is best reflected in the quantity of a single peptide.”
Quote in the paperNot a substantial claim - Premisecitation
More than 20 additional amyloid-beta proteoforms have been detected in Alzheimer's brain samples from endogenous cleavage and post-translational modifications.
“In addition to the widely known amyloid-beta 40 and amyloid-beta 42 peptides, over 20 additional amyloid-beta proteoforms have been detected in samples of Alzheimer’s brain samples arising from endogenous cleavage and post-translational modifications.(12,13)”
Quote in the paperSubstantial claim0 of 2 sources checked - Interpretationreasoning
Because amyloid precursor protein is heavily processed, it is difficult to determine whether its tryptic peptides come from unprocessed APP or from processed forms.
“Knowing that the amyloid precursor protein is heavily processed, it is difficult to determine the origin of many of its tryptic peptides - whether they are derived from an unprocessed amyloid precursor protein, or from one of many processed forms.”
Quote in the paperSubstantial claim - Premisecitation
Peptides in the C-terminal or N-terminal regions of granin family proteins were found to be decreased in Parkinson's disease.
“Specifically, peptides in the C-terminal or N-terminal regions of granin family proteins were found to be decreased in Parkinson’s.(20)”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [20] source
Not checked in the free run.
- Interpretationreasoning
All possible sequence variations and modifications can occur in numerous combinations, which further complicates interpretation of bottom-up data.
“All possible sequence variation and modifications can occur in numerous combinations, further complicating the interpretation of bottom-up data (Figure 1).”
Quote in the paperSubstantial claim - Interpretationown data
Aggregating peptides that map to a protein-coding sequence into one measure, especially by averaging or summing, suppresses outliers or noisy signals.
“By aggregating peptides mapping to a protein coding sequence into a single measure, especially by common methods that average or sum peptide measurements, outliers or noisy signals are suppressed.”
Quote in the paperSubstantial claim - Interpretationreasoning
Even measuring every unmodified and modified tryptic peptide along the predicted protein-coding gene sequence with a mass spectrometer or sequence-specific affinity reagent cannot reconstruct the original proteoforms.
“Thus, even the measurement of every unmodified and modified tryptic peptide along the predicted protein coding gene sequence using either a mass spectrometer or a sequence specific affinity reagent can’t put Humpty Dumpty back together again.”
Quote in the paperSubstantial claim - Premisecitation
The correlation between gene expression and an individual protein measurement is relatively poor.
“Such comparisons have demonstrated that the correlation between gene expression and an individual protein measurement is relatively poor.(7)”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [7] source
Not checked in the free run.
- Premisereasoning
A given tryptic peptide can come from multiple different proteoforms because related proteoforms share most of their amino acid sequence.
“As most of the amino acid sequence is shared among related proteoforms, a given tryptic peptide can be derived from multiple different proteoforms (Figure 1).”
Quote in the paperSubstantial claim - Interpretationreasoning
Treating all peptides from a protein-coding sequence as representatives of the same protein can cause important biological differences to be missed.
“By assuming all peptides from a protein coding sequence are representative of the same protein we may miss the discovery of important biological differences.”
Quote in the paperSubstantial claim - Interpretationreasoning
Ambiguity from modified or processed protein biomarkers is general to human biology and disease, not unique to Alzheimer's disease.
“Ambiguity due to modified or processed protein biomarkers is not a problem unique to Alzheimer’s disease, but rather is general to human biology and therefore human disease.”
Quote in the paperSubstantial claim - Interpretationreasoning
Discarding outlier peptide measurements may lose true biological signals.
“If those outlier peptide measurements are discarded, then true biological signals may be lost.”
Quote in the paperSubstantial claim
- May combine two claims. 18
- Classified as probably not a substantial claim. 5
May combine two claims.
PremisecitationProglucagon is processed into up to nine polypeptide products, including glucagon and GLP1, which have distinct metabolic roles and are drug targets for diabetes and obesity.
“Proglucagon is processed to form up to nine different polypeptide products, including the better-known glucagon and GLP1. Both polypeptides have distinct roles in metabolism, and both are drug targets for diabetes and obesity.(18)”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [18] source
Not checked in the free run.
May combine two claims.
InterpretationreasoningAggregating to a single protein measure reduces the number of hypotheses tested, which makes the analysis more sensitive to detecting changes in protein abundance.
“Additionally, by aggregating to a single protein measure we reduce the number of hypotheses tested, therefore making the analysis more sensitive to finding changes in protein abundance.”
Quote in the paperSubstantial claimMay combine two claims.
InterpretationreasoningUnless all proteoforms change similarly across conditions, aggregating more peptides into one protein value can average away the biological effect.
“Unless all those proteoforms change similarly among conditions, aggregating more peptides to a single protein value can average away the biological effect.”
Quote in the paperSubstantial claimMay combine two claims.
PremisecitationAmyloid-beta peptides vary in length depending on cleavage site, but commonly occur as 40 or 42 amino acids.
“The amyloid-beta peptides can be variable lengths depending on specific cleavage site, but commonly occur as a peptide of either 40 or 42 amino acids.(11)”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [11] source
Not checked in the free run.
May combine two claims.
Interpretationown dataFrom the detected peptides, one cannot conclude that only one GAPDH proteoform is present; one can conclude only that all detected peptides share the same abundance trend.
“Although there are known proteoforms, from the peptides we detect we cannot conclude that only one proteoform of GAPDH is present in our samples. Instead, we can only conclude that all the peptides we detect share the same abundance trend.”
Quote in the paperSubstantial claimMay combine two claims.
InterpretationreasoningAfter digestion, the direct link between a peptide and the proteoform it came from is lost, so individual peptide measurements are convolutions of the proteoforms present.
“Once digested in a mixture, the direct connection between a peptide and its originating proteoform(s) is lost, such that the measurements of individual peptides are convolutions of the proteoforms the peptides are present.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationPhosphorylated tau proteoforms in cerebrospinal fluid have gained acceptance as diagnostic biomarkers of disease.
“In addition to amyloid beta, phosphorylated tau proteoforms in the cerebrospinal fluid of patients have also gained acceptance as diagnostic biomarkers of disease.(14)”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [14] source
Not checked in the free run.
May combine two claims.
InterpretationreasoningA decrease in the magnitude of change can still be statistically significant, but that does not mean the differences represent every peptide measured from that protein.
“While the decreased magnitude of change can still be statistically significant, it doesn’t mean those differences are representative of every peptide measured from that protein.”
Quote in the paperSubstantial claimMay combine two claims.
PremisecitationIn Alzheimer's disease, cleavage of amyloid precursor protein produces several soluble forms, C-terminal fragments, and amyloid-beta peptides that contribute to plaques.
“In Alzheimer’s disease a series of cleavage events lead to several shorter soluble forms of amyloid precursor protein (sAPPα, sAPPβ), C-terminal fragments (AICD50, CTF 83, CTF 89, CTF 99, p3) and amyloid-beta peptides, which contribute to forming the characteristic plaques observed in the brains of diseased individuals.(10)”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [10] source
Not checked in the free run.
May combine two claims.
PremisecitationTau is often measured with antibodies for phosphorylation at threonine 181 and total-tau, but at least 95 post-translational modifications have been discovered on tau, with potentially hundreds of proteoforms from combinations.
“For example, the microtubule-associated protein tau is often measured using antibodies that represent phosphorylation at threonine 181 and one that measures so called “total-tau”. However, at least 95 post-translational modifications have been discovered on tau with potentially many 100s of possible proteoforms resulting from the combinations.(16)”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [16] source
Not checked in the free run.
May combine two claims.
PremisecitationIn large multiplexed experiments compared with label-free approaches, the missing-data pattern is distinct, but the overall relationship between abundance and missingness is similar.
“Evaluating large, multiplexed experiments in comparison to label-free approaches, the pattern of the missing data appears to be very distinct, but the macrostructure overall is similar in regard to the relationship between abundance and missing data.(8)”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [8] source
Not checked in the free run.
May combine two claims.
PremisecitationSpecific tau phosphosites may be better indicators of disease progression, which emphasizes the importance of distinguishing pTau isoforms and proteoforms.
“Additional studies indicate that specific tau phosphosites may be better indicators of disease progression, emphasizing the importance of distinguishing between different pTau isoforms and proteoforms.(15,16)”
Quote in the paperSubstantial claim0 of 2 sources checkedMay combine two claims.
InterpretationreasoningAggregating peptides into one protein-level measurement will either misinterpret a proteoform-specific difference as an effect of the whole protein or average it away and miss it.
“By aggregating the peptides into a single protein level measurement this difference will be 1) misinterpreted as an effect of the entire protein or 2) averaged away and missed entirely.”
Quote in the paperSubstantial claimMay combine two claims.
InterpretationreasoningGreater peptide coverage likely spans more proteoforms, so a measured peptide quantity could come from multiple proteoforms containing that peptide.
“Greater peptide coverage will likely span more proteoforms, meaning a measured peptide quantity could be derived from multiple proteoforms containing that peptide.”
Quote in the paperSubstantial claimMay combine two claims.
PremisecitationMultiplexing methods are limited in how many samples can be tagged and analyzed together, and the same peptides are sampled less frequently across batches than proteins are.
“These multiplexing methods are still limited in the number of samples that can be uniquely tagged, combined, and analyzed at once, and while multiple batches of samples can be acquired, the same peptides are less frequently sampled in different batches compared to proteins.(9)”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [9] source
Not checked in the free run.
May combine two claims.
PremisecitationC-peptide and insulin are both derived from proinsulin, and C-peptide is a valuable measure of insulin secretion and pancreatic beta cell function.
“C-peptide and insulin are both derived from proinsulin, with C-peptide being a valuable measure of insulin secretion and therefore pancreatic beta cell function.(17)”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [17] source
Not checked in the free run.
May combine two claims.
InterpretationcitationOutlier-exclusion methods improve protein concentration estimates but still report only one protein quantity and ignore peptides that do not agree with that value.
“While these all demonstrate improved protein concentration estimates, they still only report a single protein quantity and ignore peptides that do not agree with that single value.”
Quote in the paperSubstantial claim0 of 6 sources checkedProbably not a substantial claim: may not pass S4 (based).
Interpretationnone givenOnce a molecule has been digested into peptides, it is impossible to track the relationship between a peptide and its protein molecule.
“however, once a molecule is digested to peptides it is impossible to track the peptide-protein molecule relationship.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S2 (checkable). May combine two claims.
InterpretationreasoningDifferent peptides will likely reflect different combinations of proteoforms, so for a given protein the same peptides should be used to create the protein-level quantity across all samples.
“Furthermore, for a specific “protein” it is critical that the same peptides are used to create the protein level quantity for all samples as different peptides will likely reflect different combinations of proteoforms.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S4 (based).
InterpretationreasoningReduced statistical power for differential abundance testing on tens of thousands of peptides compared with thousands of protein groups will likely result in fewer significant differences.
“Additionally, reduced statistical power for differential abundance testing on tens of thousands of peptides compared to thousands of protein groups will also likely result in fewer significant differences.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S4 (based). May combine two claims.
InterpretationreasoningAccurate merging of peptides is impossible without detailed knowledge of all proteoforms in the sample, so keeping peptides as independent measurements is a possible solution.
“A solution to the presence of discordant peptides could be to keep all peptides as independent measurements because it is impossible to accurately merge peptides without detailed knowledge of all proteoforms in the sample.”
Quote in the paperNot a substantial claim
The paper relies on this without support.
Bottom-up proteomics commonly reports a single quantitative value per protein-coding gene by aggregating peptide quantities into protein groups after protein inference or parsimony.
“Commonly, a single quantitative value is reported for each protein coding gene by aggregating peptide quantities into protein groups following protein inference and/or parsimony.”
The paper relies on this without support.
Most strategies for aggregating peptide quantities to the protein level assume that peptides belonging to the same protein behave similarly.
“Many strategies have been created to accomplish this, with most assuming that peptides belonging to the same protein will behave similarly.”
The paper relies on this without support.
The ability to detect differences in quantitative proteomics is affected by the size of the biological effect, biological and technical variability, and the number of hypothesis tests.
“In quantitative proteomics, our ability to find differences is affected by three parameters: 1) the size of the biological effect, 2) the biological and technical variability, and 3) the number of hypothesis tests that are made within the experiment.”
The paper relies on this without support.
Data-dependent acquisition samples stochastically, leading to more missing data at the peptide level when the same precursor is not selected in all experimental runs.
“With data-dependent acquisition the sampling is stochastic, leading to more missing data at a peptide level if the same precursor is not selected in all the experimental runs.”
The paper relies on this without support.
Tissues are heterogeneous, and a large change in one cell or a small change across all cells would be indistinguishable in a bulk analysis but biologically different.
“It is well known that tissues are heterogeneous and that you could have a large change occurring in a single cell or a small change occurring across all cells -- these would be indistinguishable in a bulk analysis but clearly very different results biologically.”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Additional sources of uncertainty may arise from processes such as protein quantification in bottom-up proteomics because measurements are made at the peptide level and isoforms can be difficult to identify.
“additional sources of uncertainty may arise from processes such as protein quantification in bottom-up proteomics (Rifai et al., 2006; Plubell et al., 2022) as measurements are made at the peptide level and isoforms can be difficult to identify (Forshed et al., 2011; Webb-Robertson et al., 2014).”
From Missing data in multi-omics integration: Recent advances through artificial intelligence (2023) · cites this paper as Plubell et al., 2022
Cites
A statistical model for identifying proteins by tandem mass spectrometry.
Nesvizhskii, A., Keller, A., Kolker, E., & Aebersold, R. (2003) · Analytical Chemistry · doi:10.1021/ac0341261 · Cited by 4,902 · see citing works
Correlation between Protein and mRNA Abundance in Yeast
Gygi, S. P., Rochon, Y., Franza, B. R., & Aebersold, R. (1999) · Molecular and Cellular Biology · doi:10.1128/mcb.19.3.1720 · Cited by 3,819 · see citing works
On the Dependency of Cellular Protein Levels on mRNA Abundance
Liu, Y., Beyer, A., & Aebersold, R. (2016) · Cell · doi:10.1016/j.cell.2016.03.014 · Cited by 3,221 · see citing works
Amyloid Precursor Protein Processing and Alzheimer's Disease
O'Brien, R. J., & Wong, P. C. (2011) · Annual Review of Neuroscience · doi:10.1146/annurev-neuro-061010-113613 · Cited by 2,061 · see citing works
Assarsson, E., Lundberg, M., Holmquist, G., Björkesten, J., Bucht Thorsen, S., Ekman, D., Eriksson, A., Rennel Dickens, E., Ohlsson, S., Edfeldt, G., et al. (2014) · PLoS ONE · doi:10.1371/journal.pone.0095192 · Cited by 1,874 · see citing works
Aptamer-Based Multiplexed Proteomic Technology for Biomarker Discovery
Gold, L., Ayers, D., Bertino, J., Bock, C., Bock, A., Brody, E. N., Carter, J., Dalby, A. B., Eaton, B. E., Fitzwater, T., et al. (2010) · PLoS ONE · doi:10.1371/journal.pone.0015004 · Cited by 1,754 · see citing works
Tabb, D. L., McDonald, W. H., & Yates, J. R. (2002) · Journal of Proteome Research · doi:10.1021/pr015504q · Cited by 1,369 · see citing works
How many human proteoforms are there?
Aebersold, R., Agar, J. N., Amster, I. J., Baker, M. S., Bertozzi, C. R., Boja, E. S., Costello, C. E., Cravatt, B. F., Fenselau, C., Garcia, B. A., et al. (2018) · Nature Chemical Biology · doi:10.1038/nchembio.2576 · Cited by 907 · see citing works
Self-citationA practical review of C-peptide testing in diabetes
Leighton, E., Sainsbury, C. A., & Jones, G. C. (2017) · Diabetes Therapy · doi:10.1007/s13300-017-0265-4 · Cited by 773 · see citing works
Tau PTM Profiles Identify Patient Heterogeneity and Stages of Alzheimer’s Disease
Wesseling, H., Mair, W., Kumar, M., Schlaffner, C. N., Tang, S., Beerepoot, P., Fatou, B., Guise, A. J., Cheng, L., Takeda, S., et al. (2020) · Cell · doi:10.1016/j.cell.2020.10.029 · Cited by 688 · see citing works
the Dominantly Inherited Alzheimer Network, Barthélemy, N. R., Li, Y., Joseph-Mathurin, N., Gordon, B. A., Hassenstab, J., Benzinger, T. L. S., Buckles, V., Fagan, A. M., Perrin, R. J., et al. (2020) · Nature Medicine · doi:10.1038/s41591-020-0781-z · Cited by 651 · see citing works
Weitz, J. I., Fredenburgh, J. C., & Eikelboom, J. W. (2017) · Journal of the American College of Cardiology · doi:10.1016/j.jacc.2017.09.024 · Cited by 612 · see citing works
Physiology of Proglucagon Peptides: Role of Glucagon and GLP-1 in Health and Disease
Sandoval, D. A., & D'Alessio, D. A. (2015) · Physiological Reviews · doi:10.1152/physrev.00013.2014 · Cited by 498 · see citing works
IDPicker 2.0: Improved Protein Assembly with High Discrimination Peptide Identification Filtering
Ma, Z.-Q., Dasari, S., Chambers, M., Litton, M. D., Sobecki, S. M., Zimmerman, L. J., Halvey, P. J., Schilling, B., Drake, P. M., Gibson, B. W., et al. (2009) · Journal of Proteome Research · doi:10.1021/pr900360j · Cited by 328 · see citing works
Proteoforms as the next proteomics currency
Smith, L. M., & Kelleher, N. L. (2018) · Science · doi:10.1126/science.aat1884 · Cited by 322 · see citing works
Self-citationThe Human Proteoform Project: Defining the human proteome
Smith, L. M., Agar, J. N., Chamot-Rooke, J., Danis, P. O., Ge, Y., Loo, J. A., Paša-Tolić, L., Tsybin, Y. O., Kelleher, N. L., & The Consortium for Top-Down Proteomics (2021) · Science Advances · doi:10.1126/sciadv.abk0734 · Cited by 242 · see citing works
Self-citationLing, Y., Morgan, K., & Kalsheker, N. (2003) · The International Journal of Biochemistry & Cell Biology · doi:10.1016/s1357-2725(03)00133-x · Cited by 224 · see citing works
Ngo, D., Sinha, S., Shen, D., Kuhn, E. W., Keyes, M. J., Shi, X. D., Benson, M. D., O’Sullivan, J., Keshishian, H., Farrell, L. A., et al. (2016) · Circulation · doi:10.1161/circulationaha.116.021803 · Cited by 214 · see citing works
Portelius, E., Westman-Brinkmalm, A., Zetterberg, H., & Blennow, K. (2006) · Journal of Proteome Research · doi:10.1021/pr050475v · Cited by 206 · see citing works
Multibatch TMT reveals false positives, batch effects and missing values
Brenes, A., Hukelmann, J., Bensaddek, D., & Lamond, A. I. (2019) · Molecular & Cellular Proteomics · doi:10.1074/mcp.ra119.001472 · Cited by 196 · see citing works
Diversity of Amyloid-beta Proteoforms in the Alzheimer’s Disease Brain
Wildburger, N. C., Esparza, T. J., LeDuc, R. D., Fellers, R. T., Thomas, P. M., Cairns, N. J., Kelleher, N. L., Bateman, R. J., & Brody, D. L. (2017) · Scientific Reports · doi:10.1038/s41598-017-10422-x · Cited by 164 · see citing works
Self-citationTop-down characterization of endogenous protein complexes with native proteomics
Skinner, O. S., Haverland, N. A., Fornelli, L., Melani, R. D., Do Vale, L. H. F., Seckler, H. S., Doubleday, P. F., Schachner, L. F., Srzentić, K., Kelleher, N. L., et al. (2018) · Nature Chemical Biology · doi:10.1038/nchembio.2515 · Cited by 155 · see citing works
Self-citationTau: From research to clinical development
Holtzman, D. M., Carrillo, M. C., Hendrix, J. A., Bain, L. J., Catafau, A. M., Gault, L. M., Goedert, M., Mandelkow, E., Mandelkow, E., Miller, D. S., et al. (2016) · Alzheimer's & Dementia · doi:10.1016/j.jalz.2016.03.018 · Cited by 139 · see citing works
Goeminne, L. E., Gevaert, K., & Clement, L. (2016) · Molecular & Cellular Proteomics · doi:10.1074/mcp.m115.055897 · Cited by 110 · see citing works
Vitamin D–Binding Protein Concentrations Quantified by Mass Spectrometry
Hoofnagle, A. N., Eckfeldt, J. H., & Lutsey, P. L. (2015) · New England Journal of Medicine · doi:10.1056/nejmc1502602 · Cited by 103 · see citing works
Saltzman, A. B., Leng, M., Bhatt, B., Singh, P., Chan, D. W., Dobrolecki, L. E., Chandrasekaran, H., Choi, J. M., Jain, A., Jung, S. Y., et al. (2018) · Molecular & Cellular Proteomics · doi:10.1074/mcp.tir118.000850 · Cited by 102 · see citing works
Cerebrospinal fluid proteomics implicates the granin family in Parkinson’s disease
Rotunno, M. S., Lane, M., Zhang, W., Wolf, P., Oliva, P., Viel, C., Wills, A.-M., Alcalay, R. N., Scherzer, C. R., Shihabuddin, L. S., et al. (2020) · Scientific Reports · doi:10.1038/s41598-020-59414-4 · Cited by 94 · see citing works
Systematic detection of functional proteoform groups from bottom-up proteomic datasets
Bludau, I., Frank, M., Dörig, C., Cai, Y., Heusel, M., Rosenberger, G., Picotti, P., Collins, B. C., Röst, H., & Aebersold, R. (2021) · Nature Communications · doi:10.1038/s41467-021-24030-x · Cited by 88 · see citing works
Matzke, M. M., Brown, J. N., Gritsenko, M. A., Metz, T., Pounds, J. G., Rodland, K., Shukla, A. K., Smith, R. D., Waters, K. M., McDermott, J., et al. (2012) · PROTEOMICS · doi:10.1002/pmic.201200269 · Cited by 87 · see citing works
Self-citationIdentification of genetic variants influencing the human plasma proteome
Johansson, Å., Enroth, S., Palmblad, M., Deelder, A. M., Bergquist, J., & Gyllensten, U. (2013) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.1217238110 · Cited by 82 · see citing works
Covariation of Peptide Abundances Accurately Reflects Protein Concentration Differences
Zhang, B., Pirmoradian, M., Zubarev, R., & Käll, L. (2017) · Molecular & Cellular Proteomics · doi:10.1074/mcp.o117.067728 · Cited by 68 · see citing works
Self-citationA Review of Imputation Strategies for Isobaric Labeling-Based Shotgun Proteomics
Bramer, L. M., Irvahn, J., Piehowski, P. D., Rodland, K. D., & Webb-Robertson, B.-J. M. (2021) · Journal of Proteome Research · doi:10.1021/acs.jproteome.0c00123 · Cited by 47 · see citing works
Self-citationKinetics of plasma apolipoprotein E isoforms by LC-MS/MS: a pilot study
Blanchard, V., Ramin-Mangata, S., Billon-Crossouard, S., Aguesse, A., Durand, M., Chemello, K., Nativel, B., Flet, L., Chétiveaux, M., Jacobi, D., et al. (2018) · Journal of Lipid Research · doi:10.1194/jlr.p083576 · Cited by 43 · see citing works
Peptide Correlation Analysis (PeCorA) Reveals Differential Proteoform Regulation
Dermit, M., Peters-Clarke, T. M., Shishkova, E., & Meyer, J. G. (2021) · Journal of Proteome Research · doi:10.1021/acs.jproteome.0c00602 · Cited by 40 · see citing works
Bayesian Proteoform Modeling Improves Protein Quantification of Global Proteomic Measurements
Webb‐Robertson, B., Matzke, M. M., Datta, S., Payne, S., Kang, J., Bramer, L., Nicora, C., Shukla, A. K., Metz, T., Rodland, K., et al. (2014) · Molecular & Cellular Proteomics · doi:10.1074/mcp.m113.030932 · Cited by 39 · see citing works
Self-citationArshad, O. A., Danna, V., Petyuk, V., Piehowski, P., Liu, T., Rodland, K., & McDermott, J. (2019) · Molecular & Cellular Proteomics · doi:10.1074/mcp.ra119.001540 · Cited by 39 · see citing works
Integrated Identification and Quantification Error Probabilities for Shotgun Proteomics
The, M., & Käll, L. (2019) · Molecular & Cellular Proteomics · doi:10.1074/mcp.ra118.001018 · Cited by 37 · see citing works
Self-citationForshed, J., Johansson, H. J., Pernemalm, M., Branca, R. M. M., Sandberg, A., & Lehtiö, J. (2011) · Molecular & Cellular Proteomics · doi:10.1074/mcp.m111.010264 · Cited by 32 · see citing works
Tsai, T.-H., Choi, M., Banfai, B., Liu, Y., MacLean, B. X., Dunkley, T., & Vitek, O. (2020) · Molecular & Cellular Proteomics · doi:10.1074/mcp.ra119.001792 · Cited by 31 · see citing works
Deducing the presence of proteins and proteoforms in quantitative proteomics
Bamberger, C., Martínez‐Bartolomé, S., Montgomery, M., Pankow, S., Hulleman, J. D., Kelly, J. W., & Yates, J. R. (2018) · Nature Communications · doi:10.1038/s41467-018-04411-5 · Cited by 30 · see citing works
Schaffer, L. V., Millikin, R., Shortreed, M. R., Scalf, M., & Smith, L. (2020) · Journal of Proteome Research · doi:10.1021/acs.jproteome.0c00332 · Cited by 29 · see citing works
Self-citationQuantifying Homologous Proteins and Proteoforms
Malioutov, D., Chen, T., Airoldi, E., Jaffe, J., Budnik, B., & Slavov, N. (2019) · Molecular & Cellular Proteomics · doi:10.1074/mcp.tir118.000947 · Cited by 21 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Missing data in multi-omics integration: Recent advances through artificial intelligence
Flores, J. E., Claborne, D. M., Weller, Z. D., Webb-Robertson, B.-J. M., Waters, K. M., & Bramer, L. M. (2023) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2023.1098308 · Cited by 181 · see citing works
“Finally, additional sources of uncertainty may arise from processes such as protein quantification in bottom-up proteomics (Rifai et al., 2006; Plubell et al., 2022) as measurements are made at the peptide level and isoforms can be difficult to identify (Forshed et al., 2011; Webb-Robertson et al.,…”
BackgroundSelf-citation
- Authors
- Plubell, D. L., Käll, L., Webb-Robertson, B.-J., Bramer, L. M., Ives, A., Kelleher, N. L., Smith, L. M., Montine, T. J., Wu, C. C., & MacCoss, M. J.
- Affiliations
- University of Washington (US)
- Science for Life Laboratory (SE)
- Pacific Northwest National Laboratory, Richland, Washington 99352, United States
- Pacific Northwest National Laboratory (US)
- Northwestern University (US)
- University of Wisconsin–Madison (US)
- Stanford University (US)
- Corresponding author
- MacCoss, Michael J.
- ORCID
- Venue
- Journal of Proteome Research
- Publisher
- American Chemical Society (ACS)
- Published
- 2022-02-27
- Volume
- 21
- Issue
- 4
- Pages
- 891-898
- ISSN
- 1535-3893 (ISSN-L)
- 1535-3907
- PMC
- PMC8976764
- OpenAlex
- W4214530719
- Semantic Scholar
- 522cb8341d32843d715a39c0000ef929b9f8fdcd
- Cited by
- 85
- References
- 42
- Citations by year
- 2026: 14 · 2025: 25 · 2024: 19 · 2023: 19 · 2022: 8
- FWCI
- 6.42
- Type
- journal article
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Licence
- Locations
- Journal of Proteome Research (publishedVersion)
- PubMed (publishedVersion)
- PubMed Central (acceptedVersion)
- OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) (submittedVersion)
- Funders
- Vetenskapsr?det (2017-04030)
- U.S. Department of Energy (U011CA184783)
- National Institutes of Health (F31AG069420, P41GM103533, P41GM108569, R13AG057087, R35GM126914, RF1AG053959, U19AG065156, UH3CA246635)
- Topics
- Physical Sciences › Chemistry › Spectroscopy › Advanced Proteomics Techniques and Applications
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Advanced Biosensing Techniques and Applications
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
- Keywords
- protein quantification, shotgun proteomics, peptide aggregation, protein groups, protein inference, parsimony, proteoforms, RNA splicing, post-translational modification, protein-coding genes
- MeSH terms
- Peptides, Protein Processing, Post-Translational, Proteins, Proteome, Proteomics
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
