- Open the table and know which three columns matter
- Read the volcano and PCA plots at a glance
- Spot three common mistakes before they cost you
Many provider reports are made with a tool called DESeq2. They look intimidating, but every row answers the same three questions about one gene.
01Three columns, three questions
+1 means twice as high, −1 means half.
Filter on this, not on pvalue. Below 0.05 is a common cut-off.
Very low values can fake big changes.
02The volcano plot in one look
Each dot is a gene. The further right or left, the bigger the change; the higher up, the stronger the evidence. Your interesting genes live in the two top corners.
03Look at the PCA plot first
Each dot is a sample. Replicates should sit together. One sample far from its group can change the whole gene list, so ask your provider about it before you dig in.
Your RNA-seq results are back: 10 checks
A one-page PDF to keep next to your results: the checks to do before you trust the gene list, write it up or plan the next experiment.
Thank you! Your checklist is ready: download the PDF.
We use your email only as described in our privacy policy.
04Three checks before you trust the list
- Which way is "up"?A positive fold change means higher in your group compared with the reference group. DESeq2 picks the reference alphabetically unless told otherwise. Quick test: find the gene you knocked down. If it says "up", the comparison is flipped.
- Ignore tiny giantsA gene with a baseMean of 3 can show a 16-fold change from a handful of reads. Prefer clearly expressed genes, or confirm by qPCR.
- Keep the full tablePathway tools need every gene, not just the top hits. Save the original file.
05Things that look wrong but aren't
You don't need to be a bioinformatician to read this report. Three columns, two plots, three checks, and you'll know what your data is telling you.
Common questions
What's the difference between p-value and padj?
The p-value looks at one gene on its own. padj adjusts for the fact that you tested thousands of genes, so it controls how many of your "significant" genes are likely false. Use padj.
Can I use FPKM or TPM values to compare groups?
They're useful for looking at expression levels, but differential expression tools work on raw counts with their own normalization. Use the provider's DE table for group comparisons.
I got zero significant genes. Did something go wrong?
Not necessarily. It can mean a small effect, high variation between replicates, or too few replicates to detect changes. Check the PCA plot and talk to your provider or a bioinformatician.
The Data Back Guide
Check the data, test the provider's analysis, rerun the comparison you need for free with no code, and write the methods. 36-page guide + results check workbook.
$19 one-time · 14-day money-back guarantee · instant download