# Bulk RNA-seq Differential Expression Workflow · Yila

> A three-row figure that walks bulk RNA-seq data from raw paired-end reads to a differential expression gene list and pathway enrichment, sized for a transcriptomics methods panel, a defense talk or a bioinformatics course handout.

URL: https://yila.ai/templates/rnaseq-analysis-workflow-figure

The figure lays out the standard bulk RNA-seq chain in nine steps. Raw reads are quality-checked and adapter-trimmed, aligned to a reference genome, sorted and indexed, then counted against a gene annotation to give a raw count matrix. Counts are normalized for library size, tested for differential expression, and shown as volcano, MA and heatmap plots before the gene list goes to GO and KEGG enrichment. A reader should come away knowing which step produces which file and where the statistics enter.

- **Read quality check**: Paired-end read files on the left with a quality report beside them, labelled so the reader knows both mates are checked.
- **Adapter trimming**: Label the arrow with what is removed, and shade trimmed files differently from raw ones so the change is visible.
- **Alignment**: Draw the reference genome as a sequence file and give the annotation its own label instead of merging the two.
- **Feature counting**: Connect counting to both the aligned reads and the annotation, and name the output as a raw count matrix.
- **Normalization**: Name the normalization method and use one colour scale for the matrices before and after so they compare directly.
- **Differential expression test**: This panel should carry the test and the multiple-testing threshold, not density curves of normalized counts.
- **Result plots**: A volcano plot puts log fold change on x; an MA plot puts mean expression on x and log fold change on y.
- **Functional enrichment**: Send the DE gene list down GO and KEGG branches and show the enriched categories as dot or bar plots.

- Paper mechanism figure, Defense slides, Course handout, Academic poster

- **MA axes are swapped**: The sample's MA plot puts mean expression on the vertical axis and log fold change on the horizontal one, the reverse of the M-versus-A convention reviewers expect, so swap the axis labels before using it.
- **Step seven shows the wrong thing**: The differential expression panel draws normalized versus unnormalized density curves rather than a test. Replace it with the significance threshold, the correction method and the counts of up- and down-regulated genes.
- **Annotation is not the genome**: The sample labels the GTF as the reference genome, but a GTF is gene annotation. Alignment takes the genome sequence and counting takes the annotation, so the two inputs belong on separate labels.
- **Replicates and batches**: Reviewers ask how many biological replicates each group has and whether batch was modelled. If your figure is a single straight pipeline, add a short sample design note beside the count matrix that states groups, replicates and any batch term.

## FAQ

### Can this go straight into my paper?

It works as the starting point for a methods schematic. It is an AI-generated PNG, and the sample has swapped MA plot axes and a test panel that shows distributions instead, so check each step against your pipeline and keep refining it in chat until it fits the journal's requirements.

### Can I change the steps or tool names?

Not element by element, but you can keep refining it by chatting in Yila: remove trimming, switch alignment to pseudoalignment, add batch correction or drop the KEGG branch, and ask for a new version with those changes.

### What format is delivered?

An AI-generated PNG image, ready for a paper, handout or slide. It is not a vector graphic, and the small labels inside the volcano plot, MA plot and heatmap may blur when shrunk, so preview it at the final column width before you submit.

### Is it free?

Viewing this page and the sample costs nothing. Generating in Yila consumes credits, and every new account receives 2,500 credits, enough to reshape this pipeline around your own aligner, counting method and enrichment databases over several rounds of edits.

### I work with single-cell data. Will it fit?

This figure is built for bulk RNA-seq, where the test assumes sample-level counts. Single-cell work adds barcode demultiplexing, clustering and cell type annotation, so describe a single-cell pipeline in the prompt and let it rearrange the steps.

### Can the plot row show my real results?

You can describe the volcano, MA and heatmap styles more precisely, but the points in a generated figure are illustrative and carry no data. Put your real plots in the results section and keep this row as labelled placeholders in the workflow.

---

- [Bulk RNA-seq Differential Expression Workflow](https://yila.ai/templates/rnaseq-analysis-workflow-figure)
- [llms.txt](https://yila.ai/llms.txt)
