Hi everybody! I am looking for somebody with experience and a piece of advice!
What I am trying to accomplish here is a volcano plot in which I am plotting in the x-axis log2fold-change and y-axis is meant to -log10pvalue (pvalue came from t.test within subject comparison pre and post)
My 1st doubt, is kind of naive, because I am not working with RNAseq, I am working with RT-qPCR, with specific set of genes run through individuals in pre and post intervention. So, even though multiple experiments go through qPCR to confirm or compare, I didn't find so much of volcano plot for small set (~ 50) of genes in RT-qPCR. I guess technically and statistically is possible to do it, but I am not sure if it is strange to do it in this context
My 2nd doubt is that my volcano plot is not really "significant", it has no the typical shape in which you can perfectly observe up versus downregulation. My plot is a few upregulations versus no change. The upregulation was defined by 2 points I got from a blog:
log2fold-change > 0.6 & p.value_t.test < 0.05
Javier Hernando, Regarding Volcano Plot Applicability for RT-qPCR:
- Appropriateness: It's not unusual to use volcano plots with RT-qPCR data, even for smaller gene sets. They effectively visualize @fold-change and statistical significance, regardless of the technology used.
- Key Considerations:
- Normalization: Ensure proper normalization of RT-qPCR data to account for technical variations.
- Sample Size: Adequate sample size is crucial for reliable statistical significance.
Regarding Cutoff Values:
- Flexibility: The default cutoffs you mentioned (log2fold-change > 0.6, p-value < 0.05) are common starting points, but they can be adjusted based on:
- Biological Context: Consider the expected magnitude of changes for your genes.
- Experimental Design: Factor in sample size and variability.
- Field-Specific Standards: Adhere to common practices in your research area.
Addressing the "Non-Significant" Plot:
- Interpretation: A plot lacking a clear separation of upregulated and downregulated genes might indicate:
- Limited Number of DEGs: There might genuinely be few differentially expressed genes (DEGs) in your experiment.
- Insufficient Sensitivity: The cutoffs might be too stringent, or the sample size might be too small to detect more subtle changes.
- Biological Variability: Biological variation might be obscuring clear patterns.
Recommendations:
1. Explore Alternative Cutoffs: Experiment with different fold-change and p-value thresholds to see if patterns emerge.
2. Consider Biological Relevance: Focus on genes with meaningful fold-changes, even if not meet strict cutoffs.
3. Increase Statistical Power: If possible, increase the sample size or consider alternative statistical methods.
4. Incorporate Validation: Validate findings with independent experiments or techniques.
5. Consult Experts: Discuss with experienced researchers in your field for guidance on appropriate cutoffs and interpretation.
Additional Considerations:
- Replicates: Ensure sufficient technical and biological replicates to enhance data reliability.
- Statistical Considerations: Explore alternative statistical tests if assumptions of the t-test are not met.
- Graphical Options: Consider other visualization techniques like heatmaps or MA plots to complement volcano plot insights.
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction