4  A first analysis in EpiNexus

This chapter walks you through one complete analysis in EpiNexus, from an empty project to the results. We use the ChIP-seq example of this book: H3K27ac in the young mouse heart (GEO series GSE155100) (Boogerd et al. 2023). The next chapters explain what happens at each step.

4.1 The question

How do I go from reads to results in EpiNexus? There are five steps:

1. Create a project.
2. Add the data: upload your files or import public data.
3. Describe the samples.
4. Set up and start a run.
5. Follow the run and open the results.

NoteIn EpiNexus 1.2

This whole chapter describes EpiNexus 1.2 (October 2026). Button and field names are written in italics, as they appear on screen.

4.2 Step 1: Create a project

A project holds the data, the sample descriptions and the runs of one study.

1. In the sidebar, click All Projects, then New Project.
2. Enter a Project Name, for example “Arid1a heart H3K27ac”. The Description is optional.
3. Choose the Reference Genome. For our example, choose mouse mm10, the genome the authors used. Only installed genomes are listed. If yours is missing, ask the EpiNexus team.
4. Click Create Project.

The genome you choose here is the default for all runs in the project. Chapter 3 explains how to choose a genome.

Figure 4.1: The Create New Project form, filled in for our example, with mouse mm10 as the reference genome. EpiNexus 1.2 interface; not covered by the book’s licence.

4.3 Step 2: Add the data

Open Upload Data in the sidebar. There are two tabs.

From my computer. Drag your FASTQ files into the drop zone, or click Browse Files. EpiNexus also accepts BAM files with their peak files. Files are sent in pieces of 50 MB. If the connection breaks, the upload continues where it stopped. You can close the browser and come back later.

From a public repository (GEO / SRA / ENA). For our example, this is the tab to use.

1. Enter the accession numbers. For GSE155100, you can enter the series number, or the GSM numbers of the eight samples you want (four H3K27ac and four inputs).
2. Click Look up. EpiNexus lists the samples and the files it would import.
3. Check each row (Figure 4.2). EpiNexus fills in the Target, Condition and Rep (replicate) from the archive text. For the inputs, set the target role to input. Make sure each H3K27ac sample and its input have the same condition and replicate number. To rename a condition for all its samples at once, edit it in the box above the table.
4. Click Import N sample(s). The files go straight from the archive into your project. The bar above the table shows how much of your storage the import will use.

Figure 4.2: The import preview for GSE155100, after Look up. Each row is one sample, with the target and the condition filled in from GEO; the four inputs have the target role input. The table continues to the right with the replicate, the reads, the size and the source. EpiNexus 1.2 interface; not covered by the book’s licence.

A progress card shows each file. When the import is done, the samples appear on the Samples page.

4.4 Step 3: Describe the samples

Open Samples in the sidebar. Each row is one sample:

Table 4.1: Sample fields.

Column	What to enter	Example
Sample Name	A short, unique name	H3K27Ac_Control_rep1
Files	The FASTQ files of the sample	(filled in by the import)
Target	The histone mark or protein, or input / IgG for controls	H3K27ac, input
Condition	The group the sample belongs to	Arid1a f/f Control
Rep	The replicate number within the condition	1

These fields drive the whole analysis:

• The target decides the peak mode (narrow or broad) and marks controls.
• The condition decides which samples are compared.
• The condition and replicate decide which input belongs to which ChIP sample. EpiNexus pairs an input with the ChIP sample that has the same condition and replicate number. A sample without such an input uses the input of its condition with the lowest replicate number, or, when all inputs come from one condition, one of those (Chapter 15).

The cards above the table summarise the samples (Figure 4.3). Check them before you start a run: our example has two targets (H3K27ac and input), two conditions and two replicates per group.

Figure 4.3: The Samples page of the GSE155100 project. The cards show the targets, the conditions and the number of replicates. Each row shows the sample’s file, target, condition and replicate. EpiNexus 1.2 interface; not covered by the book’s licence.

If you uploaded your own files, register the samples here. Click Add Sample for one sample, or Import CSV for many. A sample sheet is a comma-separated file with these columns: SampleID, fastq_r1, fastq_r2, Factor, Condition, Replicate. Leave fastq_r2 empty for single-end reads, and write input in the Factor column for controls.

WarningPitfall: importing a sample sheet

Import CSV replaces all samples in the project, including samples imported from GEO. Use it before an import, not after.

4.5 Step 4: Set up and start a run

Click Run Pipeline in the sidebar. The wizard Create New Run has four steps.

Assay & Pipeline. Choose the assay: ChIP-seq, CUT&Tag, CUT&RUN, ATAC-seq or DNA Methylation. EpiNexus sets good defaults for each assay. Then choose the pipeline:

• Full Pipeline: alignment, peak calling, quality control, differential analysis and the analyses that follow. Choose this for FASTQ files.
• Analysis Only: quality control, differential analysis and the analyses that follow, from BAM files you uploaded together with their peak files. EpiNexus does not call peaks from BAM files.
• Preprocessing Only: in EpiNexus 1.2 this runs the same steps as Full Pipeline.

In version 1.2 the files you select decide what runs: FASTQ files go through the whole pipeline, and BAM files through the analysis steps.

For our example: ChIP-seq and Full Pipeline.

Samples & Genome. Select the samples to analyse. A sample can be selected only when all its files are uploaded. Check the Reference Genome. It starts as the project’s genome, but you can change it for one run. The project in our figures has mm39 as its genome (Figure 4.3); for this run we chose mm10, the genome the authors used.

Parameters. Review the settings. For ChIP-seq, the defaults are:

Table 4.2: Default parameters for ChIP-seq in EpiNexus 1.2.

Setting	Default for ChIP-seq	Explained in
Differential comparisons	Every pair of conditions, ticked	Chapter 9
Peak Caller	MACS2/3	Chapter 7
Normalization	TMM	Chapter 9
Q-value Threshold	0.05	Chapter 7
Use spike-in normalization	Off	Chapter 10
Samples that fail QC	Keep them and flag the results	Chapter 6
Consensus peaks for differential	Each sample’s peaks	Chapter 9

For our example, check the comparison: aMHC-Cre;Arid1a f/f cKO vs Arid1a f/f Control, with the control as the reference. A positive fold change then means “higher in the knockout”.

Review. EpiNexus shows a summary and checks the design (Chapter 2). Errors appear under Fix before launching. You must fix them before the run can start. Warnings appear under Design notes. Read them, then click Launch Pipeline.

Figure 4.4: Part of the Review step for our example: the genome, the eight samples with their targets, conditions and replicates, the comparison and the key settings. Here the design checks passed. EpiNexus 1.2 interface; not covered by the book’s licence.

4.6 Step 5: Follow the run

The run page has three tabs.

• Overview shows the settings, the times, and a list of steps. Each step shows its state: pending, running, done, failed or skipped. A skipped step shows the reason. When the run is finished, the list shows how long each step took and any warnings.
• Logs shows what the pipeline wrote while it ran.
• Results lists all output files. You can download each file. View analysis opens the results pages. MultiQC report opens a quality report for all samples.

A run with FASTQ files can take hours. You can close the page and come back later. Our example, eight single-end libraries, took 1 h 30 min; alignment and peak calling took half of that time (Figure 4.5).

Figure 4.5: Part of the Overview tab of the finished GSE155100 run. The warnings say that no cheaper Spot machine was free, and that multi-mark integration was skipped because the run has only one mark. The table shows how long each step took. EpiNexus 1.2 interface; not covered by the book’s licence.

Three buttons manage a finished run:

• Run again starts a new run with the same samples and settings, using the current software.
• Delete run removes the run. You can restore it for 7 days from the Runs page.
• Cancel stops a run that is still running.

4.7 Where to find the results

When the run is finished, the Analysis pages in the sidebar show the results. At the top of each page, Viewing results from shows which run you are looking at. You can switch between runs there.

Table 4.3: Results pages and the chapters that explain them.

Page	Shows	Chapter
Quality Control	Quality metrics per sample, graded	Chapter 6
Annotation	Where peaks are, relative to genes	Chapter 11
Differential	Regions that change between conditions	Chapter 9
Signal Profiles	Signal around peaks, and sample correlation	Chapter 8
Genome Browser	Tracks and peaks along the genome	Chapter 8
GO Enrichment	Biological processes near the regions	Chapter 11
TF Analysis, Motifs	Transcription factor motifs	Chapter 12
Super-Enhancers, Peak-Gene Links	Strong enhancers and their likely genes	Chapter 13
Integration	Two marks combined, gene by gene	Not in this edition

The Output pages let you take the results out of EpiNexus:

• Reports writes a report in HTML, PDF or Markdown, with a methods section you can use in a paper.
• Export downloads result files as one zip file. Signal tracks, aligned reads and methylation calls are too large for it; they are offered as download links, valid for 24 hours. It also makes a GEO submission package, and three files to reproduce the run: a shell script, the parameters, and the methods text.

Chapter 18 explains how to use these to publish your results.

4.8 Further reading

The rest of the book explains each step of the run. Chapter 5 starts with the reads.

Boogerd, Cornelis J., Ilaria Perini, Eirini Kyriakopoulou, Su Ji Han, Phit La, et al. 2023. ‘Cardiomyocyte Proliferation Is Suppressed by ARID1A-Mediated YAP Inhibition During Cardiac Maturation’. Nature Communications 14 (1): 4716. https://doi.org/10.1038/s41467-023-40203-2.