Check each claim against its experiments, reported results, scope, and stated limitations.Know what the science actually supports before you trust the answer.
Press Enter ↵ to start review
Explore by Goal
"The only way to discover the limits of the possible is to go beyond them into the impossible."
- Arthur C. Clarke
Quick Explanation
Copied
Key takeaways (skeptical, evidence-tethered)
This study uses “secretome analysis” (bioinformatic filtering of microarray DE genes for predicted secreted proteins) plus plasma validation to nominate CX3CL1 (fractalkine) and CCL2 (MCP-1) as acute exercise–induced myokines in humans.
Main skepticism: “secretome analysis” is a prediction gate, not direct secretion/secretome proteomics; candidate reduction thresholds and reliance on gene expression (not protein abundance in muscle) can bias what survives to plasma.
Long Explanation
Paper Review (visual-first): Identification of human exercise-induced myokines using secretome analysis
Evidence scope: acute 1-legged endurance bout + 12-week training; secreted-protein candidates predicted from muscle gene expression; selected proteins validated in plasma and by qPCR.
1) Visual: pipeline (what they actually did)
Secretome gating: they filtered microarray DEGs for predicted secreted proteins using GO “extracellular region/space” categories and SignalP-based signal peptide prediction.
These fold-changes are taken from the paper’s reported shortlist table/figures for secreted candidates after applying their ≥1.5 fold-change criterion (not a full ranked list).
3) Visual: overlap—why acute ≠ training (gene-expression secretome overlap)
The paper states: acute secretome shortlist reduced to 29 after fold-change cutoff; training secretome produced a list of 69; acute vs training overlap was limited with 13 genes shared.
4) Visual: which “promising” candidates survived plasma validation?
The paper reports that plasma levels increased for CCL2 and CX3CL1 during one-legged exercise and recovery, while several known myokines (IL-6, IL-8, IL-15, BDNF) did not change in plasma under their conditions. VEGFA showed a slight increase but did not reach statistical significance.
5) Explanation-first critique: what is solid vs what is uncertain
What they did well (known vs inferred)
Known (from methods/results): paired within-subject design (exercising vs non-exercising legs) reduces between-person variability for acute mechanistic gene expression shifts.
Known: they do not stop at in silico candidates; they perform qPCR and plasma assays for selected candidates, with CX3CL1/CCL2 changes concordant at both mRNA and plasma level in their data.
Known: they report limited acute vs training overlap, consistent with distinct transcriptional programs between acute and chronic exercise stimuli.
Main scientific blind spots & failure modes
Prediction ≠ secretion (critical): “secretome analysis” is an enrichment heuristic from gene expression and signal peptide/extracellular annotation; it does not directly measure secretion rates or muscle-derived protein abundance. Therefore, candidate survival depends strongly on bioinformatic annotation quality and not on actual secretion/processing dynamics.
Expression-to-plasma translation is non-uniform: they observe mRNA changes for some known myokines (e.g., IL-6 in muscle) without plasma changes, demonstrating that transcription is not always a reliable proxy for circulating protein changes in this context.
Cell-source ambiguity: CCL2 and CX3CL1 can be produced by multiple cell types, including immune cells; the study uses muscle biopsies but cannot fully exclude macrophage/other infiltrates contributing to measured mRNA and subsequent plasma changes.
Candidate truncation & thresholds: the ≥1.5 fold-change cutoff and GO/signal-peptide filters shape the candidate universe; smaller but biologically potent changes might be excluded, and some true myokines could be missed (false negatives).
Power/generalization: relatively small sample sizes and exclusion of participants based on QC/response patterns can affect stability of ranked candidates and their reproducibility across populations (e.g., different fitness levels, females, disease states).
Supported by their data: In their acute one-legged cycling conditions, CX3CL1 and CCL2 show concordant increases in muscle mRNA and circulating plasma levels.
Supported by their data: Several “known myokines” measured in plasma (IL-6, IL-8, IL-15, BDNF) show no change in plasma even when some show muscle mRNA changes, indicating condition-dependent translation from gene expression to circulating protein.
Supported by their data: Acute and training secretome candidates overlap only modestly (small shared set), implying that stimulus timing and duration strongly influence the transcriptional “myokine-like secretome” landscape.
What remains uncertain / could disprove the interpretation
Muscle origin vs immune contribution: if follow-up shows CX3CL1/CCL2 plasma changes are primarily driven by immune-cell dynamics rather than muscle contraction itself, then “myokine” attribution would shift toward a broader tissue crosstalk interpretation.
Direct secretion validation missing: if muscle secreted protein proteomics (not prediction) does not show corresponding increases for CX3CL1/CCL2 (or shows much larger contributions from other tissues), then the muscle-secretome causal framing would weaken.
7) Suggested BGPT next steps (fully within the paper’s scope)
Author reviews (click-through)
Explore additional BGPT-authored analyses per valid author listed on the paper.
Note on citations: all substantive claims here are tied directly to the provided paper text via the single paper DOI.
Feedback:
Updated: June 25, 2026
BGPT Paper Review
Study Novelty
60%
Moderate novelty: the paper’s core innovation is the application of an untargeted, microarray-driven “secretome analysis” pipeline with contralateral leg sampling and plasma validation to nominate candidate exercise-induced myokines, but it is built on existing concepts (myokines, bioinformatic secretion prediction) rather than a completely new experimental technology.
Scientific Quality
70%
Good internal coherence and validation steps (qPCR + targeted plasma), and an appropriate within-subject acute design; however, major mechanistic gaps remain because the “secretome” is predicted from gene expression (not measured secretion/proteomics), and causality/source attribution is not directly tested.
Study Generality
50%
Narrower generality: acute study used middle-aged male subjects and a specific 1-legged endurance protocol; training comparison is in healthy males, and plasma validation is limited to selected proteins—so extrapolation to females, different exercise modalities, or clinical populations remains uncertain.
Study Usefulness
70%
High utility as a hypothesis generator: it provides a constrained candidate set and identifies two chemokines with acute plasma effects (CX3CL1/CCL2) under a defined protocol, which can guide follow-up causal and source-attribution studies.
Study Reproducibility
60%
Moderate reproducibility: methods are described (microarray platform, filtering, secretome criteria, qPCR primers, plasma assay types), and the acute gene expression dataset was submitted to GEO; reproducibility is constrained by small sample sizes, selection/exclusion decisions, and prediction-step sensitivity (secretome gating).
Explanatory Depth
60%
Mechanistic depth is limited: they use clustering/upstream-regulator analysis to characterize pathways correlated with plasma levels, but causal links from muscle contraction → secretion → receptor action in vivo are not directly established.
It will extract the paper’s reported acute secretome shortlist fold-changes, organize them into a ranked table, and generate labeled plots of candidate FC vs plasma-validated outcomes (CX3CL1/CCL2 vs non-changing proteins).
Get emailed when your analysis is done!
We'll email you the results when your analysis is finished.
Hypothesis Graveyard
“IL-6 is the dominant acute exercise myokine responsible for systemic signaling” (stronger versions are weakened here): the paper reports IL-6 mRNA changes without plasma IL-6 change under their specific conditions, undermining a universal dominance claim.
“Most known myokines should show plasma increases whenever muscle mRNA increases” (a general strongman version is unlikely): the paper shows multiple known myokines with mRNA-level signals not translating into plasma changes in their protocol, demonstrating non-uniform expression→circulation mapping.
Explore by Goal
Quick Explanation
Long Explanation
Top Data Sources
Ask a Follow-Up
Key Insight
Keep Exploring
Analysis 
Hypothesis Graveyard
Potential Experiments
Discussion