Rice Salt-Stress RNA-seq Analysis: DESeq2 Case Study

At a glance

Public rice RNA-seq data, reanalyzed from raw reads

The analysis starts from public sequencing reads rather than a processed expression matrix. Alignment, gene quantification, filtering, and statistical thresholds are therefore part of the visible provenance rather than hidden assumptions.

DatasetPRJNA895747

Design4 groups × 3 replicates

Core methodsHISAT2 + DESeq2

Compute time1 h 30 min

01 / Scientific context

Salt-alkali stress can reshape the rice transcriptome over time

Salt-alkali stress limits rice growth through ionic, osmotic, and high-pH stress. The experiment examined an early response and a response after longer Na₂CO₃ exposure; the two timepoints were therefore modeled separately.

Scientific question

Which genes change in rice seedlings after one and five days of Na₂CO₃ stress, and how much of the strict differential-expression response is shared between the two timepoints?

Organism	Oryza sativa japonica rice seedlings
Public dataset	NCBI BioProject PRJNA895747
Samples	12 paired-end RNA-seq libraries
Design	Control and Na₂CO₃ treatment at Day 1 and Day 5; three biological replicates per group
Reference	IRGSP-1.0
Contrasts	Day 1 treated vs control; Day 5 treated vs control

Provenance note

The session prompt labels the cultivar Nipponbare, whereas the linked publication identifies Liaoxing No.1. The publication and BioProject are treated as authoritative here, but the source metadata mismatch should be resolved before a formal reproduction claim.

02 / Analysis workflow

FASTQ → fastp → HISAT2 → featureCounts → DESeq2

Both timepoints followed the same count-based workflow and were tested as separate treatment-versus-control contrasts.

Read quality controlfastp 0.23.2 removed adapters, trimmed low-quality bases, and discarded reads shorter than 36 nt.

Reference alignmentHISAT2 2.2.1 aligned filtered reads to IRGSP-1.0; SAMtools sorted and indexed the alignments.

Gene quantificationfeatureCounts 2.0.3 counted properly paired reads overlapping annotated exons.

Differential expressionDESeq2 1.38.0 filtered genes with fewer than 10 total counts and tested timepoint-specific contrasts.

Evidence packageCount matrices, complete results, strict gene lists, overlap tables, ranked inputs, QC summaries, and figures were exported.

Strict DEG ruleBenjamini–Hochberg FDR ≤ 0.01 and |log₂ fold change| ≥ 0.585, equivalent to at least a 1.5-fold change.

03 / Results

Day 5 produced the larger strict differential-expression response

DESeq2 tested 28,175 genes at Day 1 and 28,973 at Day 5. Under the strict threshold, 1,451 genes were differentially expressed at Day 1 and 2,005 at Day 5—a 38% increase in the later response.

Day 1

649 upregulated and 802 downregulated genes.

Day 5

869 upregulated and 1,136 downregulated genes.

Shared response

331 strict DEGs overlapped between timepoints.

Temporal specificity

1,120 genes were Day-1-specific and 1,674 were Day-5-specific.

Charts comparing tested genes, strict differential expression, direction, and overlap for rice salt stress — **Figure 1. Differential-expression response at Day 1 and Day 5.** The later contrast contained more strict DEGs; the overlap uses the recorded FDR and fold-change thresholds.

Sequencing depth and gene detection were consistent across libraries

The report records a mean library size of 19.8 ± 2.0 million paired-end reads per sample, 29,456 ± 504 detected genes, 96–97% HISAT2 alignment, and more than 98% Q20 bases after trimming.

Sequencing depth and detected genes across 12 rice RNA-seq libraries — **Figure 2. Library size and gene-detection summary.** Libraries contained approximately 17–23 million reads while detected-gene counts remained near 29,000–30,000.

Both contrasts contained large, statistically supported changes

Side-by-side DESeq2 volcano plots for rice salt stress at Day 1 and Day 5 — **Figure 3. DESeq2 volcano plots.** Dashed vertical lines mark the 1.5-fold-change boundary. Associations do not establish that individual genes cause stress tolerance.

04 / Published comparison

The reanalysis reproduced the direction of the time-dependent response

The publication and Pipette analysis both found more DEGs at Day 5 than Day 1 and a modest predominance of downregulation. Pipette returned fewer strict DEGs at both timepoints.

Metric	Published study	Pipette analysis
Alignment rate	95.24–95.94%	96–97%
Day 1 DEGs	1,780	1,451
Day 5 DEGs	2,315	2,005
Day 5 : Day 1 ratio	1.30	1.38

Software versions, reference annotations, filtering, and threshold implementation can change exact counts. The records do not isolate a single cause for the differences.

05 / Limitations

What this analysis cannot establish

Each condition-by-timepoint group contains only three biological replicates.
The source study sampled shoots from one salt-sensitive japonica cultivar; results should not be generalized to other tissues, stages, or rice populations.
The cultivar-name discrepancy between the prompt and publication remains unresolved.
The session does not independently reproduce the publication’s qRT-PCR validation.
Enrichment-ready lists were exported, but the session does not present interpreted GO or KEGG results.
The 331-gene overlap depends on the strict threshold and is not a complete map of shared biological regulation.

Evidence status

Reanalyzed from public data. The direction of the time-dependent response agrees with the publication, but the provenance mismatch and missing enrichment interpretation prevent describing the entire published analysis as reproduced.

06 / Reproducibility and outputs

Artifacts recorded in the Pipette session

The files below are recorded in the session inventory; only the preserved session PDF is published on this website.

count_matrix.csv38,993 genes across 12 samples

de_results_day1/Complete Day 1 DESeq2 result set

de_results_day5/Complete Day 5 DESeq2 result set

overlap_genes_strict.csv331 genes significant at both timepoints

gsea_ranked_day1.csvRanked genes for downstream analysis

sample_qc_table.csvRead depth, detection, and alignment metrics

Original Pipette session report14-page PDF · 1.1 MB · generated February 9, 2026. The HTML article is the primary case study; the PDF is the preserved session record.

View session Download PDF

Rice Salt-Stress RNA-seq Analysis: FASTQ to Differential Expression

Public rice RNA-seq data, reanalyzed from raw reads

Salt-alkali stress can reshape the rice transcriptome over time

Scientific question

FASTQ → fastp → HISAT2 → featureCounts → DESeq2

Day 5 produced the larger strict differential-expression response

Sequencing depth and gene detection were consistent across libraries

Both contrasts contained large, statistically supported changes

The reanalysis reproduced the direction of the time-dependent response

What this analysis cannot establish

Artifacts recorded in the Pipette session

Run a similar RNA-seq analysis

Rice Salt-Stress RNA-seq Analysis: FASTQ to Differential Expression

Public rice RNA-seq data, reanalyzed from raw reads

Salt-alkali stress can reshape the rice transcriptome over time

Scientific question

FASTQ → fastp → HISAT2 → featureCounts → DESeq2

Day 5 produced the larger strict differential-expression response

Sequencing depth and gene detection were consistent across libraries

Both contrasts contained large, statistically supported changes

The reanalysis reproduced the direction of the time-dependent response

What this analysis cannot establish

Artifacts recorded in the Pipette session

Related reading

More RNA-seq case studies

Run a similar RNA-seq analysis