RNA-Seq vs ChIP-Seq: Complementary Approaches for Gene Regulation

RNA sequencing (RNA-seq) and chromatin immunoprecipitation sequencing (ChIP-seq) answer different biological questions. RNA-seq measures the abundance of transcribed RNA across the genome, providing a snapshot of gene expression levels in a cell population. ChIP-seq identifies genomic regions where specific proteins, such as transcription factors or modified histones, are bound. These two methods do not compete with one another. They measure distinct molecular layers, and combining them allows researchers to connect the presence of a regulatory protein at a genomic location with the resulting changes in gene output. This article explains the technical basis of each method, the biological questions each can address, and a practical framework for integrating the two data types in regulatory genomics studies.

The intended readers are students, researchers, analysts, and life-science professionals who need to design experiments, interpret published studies, or build analysis pipelines. The practical outcome is a conceptual framework for deciding when to use RNA-seq, when to use ChIP-seq, and how to integrate the resulting data to test hypotheses about gene regulation.

What Each Method Measures

RNA-seq quantifies the transcriptome. The workflow begins with RNA extraction, conversion to complementary DNA, library preparation, and high-throughput sequencing. The resulting reads are aligned to a reference genome or transcriptome, and counts per gene are used to estimate expression levels. Differential expression analysis compares conditions to identify genes whose transcript abundance changes.

ChIP-seq identifies protein-DNA interactions. The workflow involves crosslinking proteins to DNA, fragmenting the chromatin, immunoprecipitating the protein of interest with a specific antibody, reversing the crosslinks, and sequencing the enriched DNA fragments. The resulting reads are aligned to the genome, and peak calling identifies regions of significant enrichment relative to an input or immunoglobulin G control.

The European Bioinformatics Institute provides training materials that cover the experimental design and analysis steps for both sequencing approaches, including quality control, alignment, and interpretation of results. These resources are useful for analysts who need to understand the assumptions built into each step of the pipeline.

The Regulatory Question That Requires Both Methods

Gene expression is controlled at multiple levels. A transcription factor can only influence a gene if it binds near that gene's regulatory elements, such as promoters or enhancers. Histone modifications can mark chromatin as active or repressed. RNA-seq alone shows the outcome of these regulatory events but does not reveal which proteins caused the outcome. ChIP-seq alone shows where a protein binds but does not reveal whether that binding changed transcription.

The integration of RNA-seq and ChIP-seq data addresses the question of causality. If a transcription factor binds near a gene and that gene shows altered expression in the same condition, the evidence supports a direct regulatory relationship. If a gene changes expression but shows no nearby binding of the factor under study, the change is likely indirect. This logic underpins many published studies that combine the two methods.

A review in the International Journal of Molecular Sciences discusses how ChIP-seq data can strengthen information generated from RNA-seq in plant research. The authors describe how integrating the two data types helps unravel transcriptional regulatory networks, particularly when researchers need to associate differentially expressed transcription factors with their downstream responsive genes. The same logic applies to animal and human studies.

Core Principles of Experimental Design

Biological Replicates

Both RNA-seq and ChIP-seq require biological replicates to support statistical inference. A single sample cannot distinguish biological variation from technical noise. The number of replicates depends on the expected effect size, the heterogeneity of the sample, and the statistical method used. For RNA-seq, three to five replicates per condition are common in cell culture studies. For ChIP-seq, two to three replicates are often used, though more may be needed when peak strength is weak or the protein of interest binds a small number of genomic regions.

Controls

RNA-seq experiments compare conditions against a baseline, such as untreated cells or a wild-type genotype. The choice of control defines the biological question. ChIP-seq experiments require an input DNA control to account for sequencing biases and chromatin accessibility. Without an input control, peak callers cannot distinguish true binding from background noise.

Antibody Validation

ChIP-seq quality depends on antibody specificity. An antibody that cross-reacts with other proteins produces false peaks. Researchers should validate antibodies by western blot, immunoprecipitation followed by mass spectrometry, or use of a knockout or knockdown cell line. The antibody validation data should be reported alongside the ChIP-seq results.

Sequencing Depth

The required sequencing depth depends on the genome size and the biological question. For RNA-seq, deeper sequencing allows detection of low-abundance transcripts. For ChIP-seq, deeper sequencing improves the resolution of binding sites but reaches a point of diminishing returns. The European Bioinformatics Institute training materials provide guidance on sequencing depth and quality metrics for both methods.

Practical Workflow for Integrated Analysis

Step 1: Define the Biological Question

Write a clear hypothesis before collecting data. For example, a researcher might ask whether a specific transcription factor directly activates a set of target genes in response to a stimulus. This question requires both ChIP-seq to identify binding sites and RNA-seq to measure expression changes.

Step 2: Collect Matched Samples

The RNA-seq and ChIP-seq samples should come from the same cell population and treatment condition. Mismatched samples introduce confounding variables. For example, if RNA-seq uses cells treated for 24 hours and ChIP-seq uses cells treated for 6 hours, the binding events may not correspond to the expression changes.

Step 3: Process Each Data Type Independently

Run quality control, alignment, and quantification for each data type before attempting integration. For RNA-seq, this includes read trimming, alignment to the reference genome, and generation of a count matrix. For ChIP-seq, this includes alignment, removal of duplicate reads, and peak calling. Poor quality in either data set will compromise the integrated analysis.

Step 4: Integrate the Data

The simplest integration approach is to overlap ChIP-seq peaks with the promoters or enhancers of differentially expressed genes. A gene is considered a candidate direct target if it shows both a nearby binding event and a significant expression change. More sophisticated approaches use statistical models to test whether the observed overlap exceeds what would be expected by chance.

Step 5: Validate Selected Targets

Integrated analysis generates candidate regulatory relationships. These candidates require experimental validation. Common validation methods include quantitative PCR after chromatin immunoprecipitation to confirm binding, reporter assays to test enhancer activity, and knockdown or overexpression of the transcription factor to confirm that the predicted target genes respond.

Options and Tradeoffs in Data Integration

Peak-to-Gene Assignment

ChIP-seq peaks must be assigned to genes before integration with RNA-seq data. The simplest approach assigns a peak to the nearest gene. This method is easy to implement but can be incorrect when a peak lies in an enhancer that regulates a distant gene. More sophisticated approaches use chromatin conformation data, correlation of accessibility across cell types, or expression quantitative trait loci to link peaks to genes.

Correlation of Binding and Expression

A direct target gene should show a correlation between the strength of protein binding and the magnitude of expression change. However, the relationship is not always linear. Some transcription factors bind many sites without strongly affecting expression, while others have large effects at a small number of sites. The interpretation should account for the possibility that binding is necessary but not sufficient for expression change.

Time Course Considerations

Gene regulation is dynamic. A transcription factor may bind a target early in a response, with expression changes following hours later. A single time point may miss the relevant window. Time course experiments, where both RNA-seq and ChIP-seq are collected at multiple time points, provide a more complete picture of the regulatory cascade.

Cell Type Specificity

Both gene expression and protein binding are cell type specific. A regulatory relationship identified in one cell type may not hold in another. Studies that profile multiple cell types, such as the single-cell work in human developmental hematopoiesis, show that chromatin accessibility and gene expression patterns differ substantially across cell states.

Observations and Measurements That Support Integration

Transcription Factor Motif Enrichment

ChIP-seq peaks can be analyzed for enrichment of known transcription factor binding motifs. If the immunoprecipitated protein is a transcription factor, its motif should be enriched in the peak regions. This analysis provides confidence that the antibody captured the intended protein and that the peaks represent functional binding sites.

Histone Modification Patterns

ChIP-seq can target histone modifications instead of transcription factors. Active promoters are marked by H3K4me3, active enhancers by H3K27ac, and repressed regions by H3K27me3. Integrating histone modification ChIP-seq with RNA-seq reveals whether expression changes correspond to changes in chromatin state. A study of human lung cancer cells used integrative RNA-seq and H3 trimethylation ChIP-seq analysis to examine gene regulation in cells isolated by laser microdissection.

Chromatin Accessibility

ATAC-seq measures chromatin accessibility and is often used alongside RNA-seq. Open chromatin regions are candidate regulatory elements. The integration of ATAC-seq and RNA-seq follows the same logic as ChIP-seq and RNA-seq integration, with accessibility serving as a proxy for regulatory potential. Studies in systemic lupus erythematosus, asthma exacerbation, and hepatocellular carcinoma have used this approach to identify transcription factors and target genes in disease contexts.

Single-Cell Resolution

Single-cell RNA-seq and single-cell ATAC-seq allow regulatory analysis at the level of individual cells. A study of human developmental hematopoiesis applied both methods to over 8,000 blood cells from fetal liver and bone marrow. The authors identified progenitor populations and observed opposing patterns of chromatin accessibility and differentiation that coincided with dynamic changes in the activity of lineage-specific transcription factors. The integrative analysis revealed extensive epigenetic but not transcriptional priming of hematopoietic stem cells prior to lineage commitment.

A similar approach in pig embryonic myogenesis profiled gene expression and chromatin accessibility in developing somites and myotomes. The authors identified myogenic cells, constructed a differentiation trajectory, and found that dynamic changes in gene expression and chromatin accessibility coincided with the activities of cell type specific transcription factors. Integrative analysis identified EGR1 and RHOB as critical regulators of pig embryonic myogenesis.

Records and Documentation

Metadata Standards

Every sequencing experiment should be documented with complete metadata. This includes the cell type, treatment conditions, time points, antibody lot numbers, library preparation kits, sequencing instrument, and analysis software versions. The FAIR Guiding Principles describe the importance of making data findable, accessible, interoperable, and reusable. Following these principles ensures that others can reproduce the analysis or reanalyze the data with new methods.

Data Repositories

Raw sequencing data should be deposited in public repositories. The National Center for Biotechnology Information hosts the Sequence Read Archive and Gene Expression Omnibus, which are standard repositories for RNA-seq and ChIP-seq data. The National Institutes of Health Genomic Data Sharing Policy describes the expectations for data sharing in NIH funded research. Researchers should check the requirements of their funding agency and journal before starting the project.

Analysis Scripts

The computational steps used to process and integrate the data should be recorded. Version controlled scripts, container images, and workflow management systems improve reproducibility. The analysis log should include the reference genome version, alignment parameters, peak calling thresholds, and integration criteria.

Common Failure Patterns

Antibody Failure

ChIP-seq experiments fail when the antibody does not specifically enrich the target protein. Signs of antibody failure include low numbers of peaks, no enrichment of the expected motif, and poor signal relative to the input control. Researchers should test antibodies in a pilot experiment before committing to a large study.

Batch Effects

RNA-seq and ChIP-seq experiments performed at different times or in different batches can show systematic differences unrelated to the biological variable. Batch effects can produce false positives and false negatives in differential analysis. Experimental design should balance conditions across batches, and computational methods can be used to detect and correct for batch effects.

Misalignment of Time Points

If RNA-seq and ChIP-seq samples are collected at different times after a treatment, the integration may miss the regulatory relationship. Binding events that occur early and resolve before the expression measurement will not be detected. Conversely, expression changes that occur after the binding measurement will appear unconnected.

Overinterpretation of Correlation

A correlation between binding and expression does not prove causation. The transcription factor may bind a site without affecting transcription, or the expression change may be caused by another factor that binds the same region. Functional validation is required to establish a direct regulatory relationship.

Peak Calling Artifacts

ChIP-seq peak callers can produce false peaks in regions of high copy number, repetitive DNA, or extreme chromatin accessibility. The input control helps mitigate these artifacts, but some false peaks will remain. Manual inspection of candidate peaks in a genome browser is a useful quality check.

Limitations of Each Method

RNA-seq Limitations

RNA-seq measures steady-state RNA levels, which reflect the balance of transcription and degradation. A change in RNA abundance does not distinguish increased transcription from decreased RNA stability. RNA-seq also does not capture post-transcriptional regulation, such as alternative splicing, RNA editing, or translation efficiency, unless specialized protocols are used.

ChIP-seq Limitations

ChIP-seq requires a high quality antibody and sufficient cell numbers. The crosslinking step can introduce artifacts, and the resolution of binding sites is limited by the fragment size. ChIP-seq does not reveal whether a bound protein is activating or repressing transcription. The functional consequence of binding must be inferred from other data, such as RNA-seq.

Integration Limitations

The integration of RNA-seq and ChIP-seq data identifies candidate regulatory relationships but does not prove them. The assignment of peaks to genes is uncertain, particularly for enhancers that act over long distances. The correlation between binding and expression may be weak for transcription factors that bind permissively across the genome. These limitations should be acknowledged in the interpretation of results.

Safety and Regulatory Context

Data Sharing Requirements

Research involving human subjects or human cell lines is subject to privacy and consent requirements. The National Institutes of Health Genomic Data Sharing Policy describes the expectations for data sharing and the protections required for human genomic data. Researchers should consult their institutional review board and data sharing office before depositing data in public repositories.

Animal Research Oversight

Studies using animal tissues, such as the pig myogenesis work, are subject to animal welfare regulations. The ethical use of animals in research requires approval from an institutional animal care and use committee. Researchers should document the source of animal tissues and the compliance with relevant regulations.

Data Quality Standards

Funding agencies and journals increasingly require adherence to data quality standards. The FAIR Guiding Principles provide a framework for data management and stewardship. Following these principles improves the reproducibility and reusability of sequencing data.

Professional Escalation Criteria

When to Seek Statistical Expertise

Integrated analysis of RNA-seq and ChIP-seq data involves complex statistical methods. Analysts should consult a biostatistician or bioinformatician when the experimental design involves multiple factors, when the number of samples is small, or when the integration requires advanced modeling. Attempting to interpret results from an inadequately powered study can lead to incorrect conclusions.

When to Repeat the Experiment

A ChIP-seq experiment should be repeated when the peak count is unexpectedly low, when the motif enrichment is absent, or when the replicates show poor concordance. An RNA-seq experiment should be repeated when the sequencing depth is insufficient, when the alignment rate is low, or when the quality metrics indicate degradation of the RNA.

When to Consult a Domain Expert

The biological interpretation of integrated data requires knowledge of the gene regulatory mechanisms in the relevant tissue or disease. Researchers should consult a domain expert when the candidate target genes do not fit the expected biology, when the transcription factor has no known binding motif, or when the results conflict with published literature.

Case Examples of Integrated Analysis

Histone Lactylation in Gestational Diabetes

A study in the Journal of Proteome Research combined RNA-seq and ChIP-seq to examine histone lactylation in gestational diabetes mellitus. The authors identified upregulated differentially expressed genes with hyper histone lactylation modification. The integrated analysis identified CACNA2D1 as a key gene with both increased expression and increased lactylation modification. Functional experiments showed that this gene promotes cell vitality and proliferation, suggesting a role in the progression of gestational diabetes.

Transcription Factor Networks in Myogenesis

A study in BMC Biology combined single-cell RNA-seq and ATAC-seq to examine myogenic differentiation in pig. The authors profiled gene expression and chromatin accessibility in developing somites and myotomes, identified myogenic cells, and constructed a differentiation trajectory. The integrated analysis identified EGR1 and RHOB as critical regulators of pig embryonic myogenesis. The study provides a resource for understanding skeletal muscle development in agricultural animals.

Disease-Critical Cell Types in the Brain

A study in Nature Communications integrated genome-wide association study summary statistics with single-cell ATAC-seq and RNA-seq profiles from 83 brain cell types. The authors identified disease-critical cell types for 22 of 28 traits using ATAC-seq and for 8 of 28 traits using RNA-seq. The results showed that chromatin accessibility data can prioritize cell types that are not identified by gene expression data alone.

Osteoclastogenesis Regulation

A study in the Journal of Bone and Mineral Research combined single-cell RNA-seq, single-cell ATAC-seq, bulk RNA-seq, and ChIP-seq to define the regulatory landscape of osteoclastogenesis. The authors identified IRF8 as a master negative regulator that maintains monocyte identity and restricts chromatin accessibility at osteoclastogenic loci. The integrated analysis revealed how transcription factors and chromatin state coordinate to control cell fate decisions.

Nuclear Receptor Binding in Macrophages

A study in the Journal of Biological Chemistry used ChIP-seq to profile retinoid X receptor occupancy in macrophage-like cells treated with agonists of RXR or partner receptors. The results supported a model in which RXR functions as a nonlimiting module, with only a small subset of regulated genes showing decreases in both RXR binding and mRNA levels in response to combined agonist treatment.

Long Noncoding RNA Regulation in Myogenesis

A study in Biology Direct combined transcriptome profiling with p63 ChIP-seq analysis in differentiating mouse myoblasts. The authors identified the long noncoding RNA Airn as a direct transcriptional target of TAp63gamma. The findings positioned Airn as a regulator of myogenic commitment that modulates both MyoD and MyoG expression.

At a Glance

Practical Decision Criteria for Method Selection

Use RNA-seq Alone When

The research question concerns the transcriptome as a whole. Examples include identifying genes that respond to a treatment, comparing expression across cell types, or discovering novel transcripts. RNA-seq alone is sufficient when the goal is descriptive instead of mechanistic.

Use ChIP-seq Alone When

The research question concerns the genomic distribution of a specific protein. Examples include mapping the binding sites of a transcription factor, characterizing histone modification patterns across the genome, or comparing binding between conditions. ChIP-seq alone is sufficient when the goal is to describe where a protein binds.

Use Both Methods When

The research question concerns the regulatory relationship between protein binding and gene expression. Examples include identifying direct target genes of a transcription factor, determining whether a histone modification is associated with activation or repression, or testing whether a binding event leads to a functional outcome. The integration of both methods provides evidence that connects the regulatory event to the transcriptional response.

Use ATAC-seq Instead of ChIP-seq When

The research question concerns chromatin accessibility instead of the binding of a specific protein. ATAC-seq requires less starting material than ChIP-seq and does not require an antibody. The integration of ATAC-seq and RNA-seq identifies open chromatin regions associated with differentially expressed genes, providing a genome wide view of regulatory potential.

Statistical Approaches for Integration

Overlap Analysis

The simplest statistical approach tests whether ChIP-seq peaks overlap the regulatory regions of differentially expressed genes more often than expected by chance. This approach requires a definition of the regulatory region, such as the promoter or the gene body, and a background model that accounts for the genomic distribution of peaks and genes.

Correlation Analysis

A more quantitative approach correlates the strength of protein binding with the magnitude of expression change across genes. This approach can identify genes where binding intensity predicts expression level. The correlation can be calculated at the level of individual peaks or aggregated across all peaks assigned to a gene.

Regression Models

Regression models can incorporate multiple data types to predict gene expression from binding data, chromatin accessibility, and other features. These models can estimate the contribution of each regulatory feature to the expression outcome. The models require careful validation to avoid overfitting.

Machine Learning Approaches

Machine learning methods can integrate RNA-seq and ChIP-seq data with other genomic features to predict regulatory relationships. These approaches can capture nonlinear relationships but require large training data sets and careful evaluation. The results should be interpreted with caution when the model is applied to new cell types or conditions.

Quality Control Metrics

RNA-seq Quality Metrics

The alignment rate indicates the proportion of reads that map to the reference genome. A low alignment rate suggests contamination or poor library quality. The number of genes detected reflects the sequencing depth and the complexity of the library. The distribution of reads across the gene body can reveal 3 prime bias or RNA degradation.

ChIP-seq Quality Metrics

The fraction of reads in peaks indicates the signal to noise ratio. A low fraction suggests weak enrichment or high background. The number of peaks and their genomic distribution provide a first check on the expected biology. The enrichment of the known motif for the transcription factor confirms that the antibody captured the intended protein.

Reproducibility Metrics

Replicate concordance measures the consistency of results across biological replicates. For RNA-seq, the correlation of expression values across replicates should be high. For ChIP-seq, the overlap of peaks across replicates should be substantial. Low reproducibility indicates technical variation or biological heterogeneity that should be investigated.

Common Analysis Pitfalls

Using the Wrong Reference Genome

The reference genome version affects alignment and annotation. Different versions of the same genome can produce different results, particularly in regions of structural variation. The reference genome version should be recorded and reported with the analysis.

Ignoring Strand Specificity

RNA-seq libraries can be strand specific or non strand specific. The choice affects the interpretation of reads that overlap genes on opposite strands. The library preparation protocol determines whether strand information is preserved, and the analysis should account for this.

Applying Inappropriate Normalization

RNA-seq data require normalization to account for differences in sequencing depth and library composition. The choice of normalization method affects the results of differential expression analysis. Methods that assume most genes are unchanged between conditions may be inappropriate when large fractions of the genome are differentially expressed.

Calling Peaks Without an Input Control

ChIP-seq peak calling requires an input control to distinguish true binding from background. Without a control, the analysis will produce many false peaks in regions of high chromatin accessibility. The input control should be sequenced to a depth comparable to the immunoprecipitated sample.

Reporting Standards

Methods Section

The methods section should describe the experimental and computational steps in sufficient detail for another researcher to reproduce the work. This includes the cell culture conditions, treatment protocols, library preparation kits, sequencing instrument, alignment software and parameters, peak caller and thresholds, and integration criteria.

Data Availability Statement

The data availability statement should describe where the raw and processed data are deposited. The accession numbers for the Sequence Read Archive, Gene Expression Omnibus, or other repositories should be provided. The analysis scripts should be available in a public repository.

Figure Presentation

Integrated results are often presented as genome browser tracks showing ChIP-seq peaks and RNA-seq signal at a locus of interest. The tracks should be scaled consistently across conditions. The figure should include the genomic coordinates, the gene annotation, and the scale of the signal.

Frequently Asked Questions

What is the main difference between RNA-seq and ChIP-seq?

RNA-seq measures the abundance of RNA transcripts in a sample, providing a genome wide view of gene expression. ChIP-seq identifies genomic regions where a specific protein, such as a transcription factor or a modified histone, is bound. RNA-seq answers the question of what genes are expressed, while ChIP-seq answers the question of where regulatory proteins are located on the DNA.

Can RNA-seq replace ChIP-seq for studying gene regulation?

RNA-seq cannot replace ChIP-seq because the two methods measure different molecular layers. RNA-seq shows the outcome of gene regulation but does not reveal which proteins caused the change. ChIP-seq shows where a protein binds but does not reveal whether that binding changed transcription. Both methods are needed to connect regulatory events to transcriptional outcomes.

How do I decide which method to use first?

The decision depends on the biological question. If the goal is to identify genes that respond to a treatment, RNA-seq is the appropriate first step. If the goal is to map the binding sites of a specific transcription factor, ChIP-seq is the appropriate first step. For a mechanistic study, many researchers start with RNA-seq to identify candidate genes and then use ChIP-seq to test whether a specific protein binds near those genes.

What is the difference between ChIP-seq and ATAC-seq?

ChIP-seq identifies binding sites of a specific protein using an antibody to immunoprecipitate the protein of interest. ATAC-seq identifies regions of open chromatin using a transposase that inserts sequencing adapters into accessible DNA. ATAC-seq does not require an antibody and provides a genome wide view of regulatory potential, while ChIP-seq provides information about a specific protein.

How do I integrate RNA-seq and ChIP-seq data?

The first step is to process each data type independently with appropriate quality control. The second step is to assign ChIP-seq peaks to genes, either by nearest gene or by a more sophisticated method. The third step is to overlap the assigned peaks with differentially expressed genes from the RNA-seq analysis. Genes with both a nearby binding event and a significant expression change are candidate direct targets.

What are the common reasons for failed ChIP-seq experiments?

The most common cause of failure is an antibody that does not specifically enrich the target protein. Other causes include insufficient crosslinking, poor chromatin fragmentation, low cell numbers, and inadequate sequencing depth. Signs of failure include a low number of peaks, no enrichment of the expected motif, and poor signal relative to the input control.

How many replicates do I need for an integrated study?

The number of replicates depends on the expected effect size and the heterogeneity of the sample. For RNA-seq, three to five biological replicates per condition are common. For ChIP-seq, two to three replicates are often used. More replicates may be needed when the expected differences are small or when the samples are highly variable.

What should I do if the ChIP-seq and RNA-seq results do not agree?

Disagreement between binding and expression can have several explanations. The binding may occur at a different time point than the expression change. The protein may bind without affecting transcription. The peak may be assigned to the wrong gene. The expression change may be caused by a different factor. Each of these possibilities should be investigated before concluding that the results are contradictory.

Related Bioinformatics Guides

• Alternative Splicing Analysis from RNA-Seq Data
• RNA Polymerase: Structure, Transcription Mechanisms, and Transcriptional Regulation in Prokaryotes and Eukaryotes
• RNA-Seq Differential Expression: DESeq2, edgeR, and limma-voom Frameworks
• Structural Dynamics of SARS-CoV-2 Spike Protein: Computational Insights into Immune Evasion
• ChIP-Seq Bioinformatics Workflows

References and Further Reading

• EMBL-EBI Training. European Bioinformatics Institute.
• NCBI Data Resources. National Center for Biotechnology Information.
• Genomic Data Sharing Policy. National Institutes of Health.
• The FAIR Guiding Principles. Scientific Data.
• Integrative Single-Cell RNA-Seq and ATAC-Seq Analysis of Human Developmental Hematopoiesis.. Cell stem cell, 2021.
• ChIP-seq and RNA-seq Reveal the Involvement of Histone Lactylation Modification in Gestational Diabetes Mellitus.. Journal of proteome research, 2024.
• Integrative single-cell RNA-seq and ATAC-seq analysis of myogenic differentiation in pig.. BMC biology, 2023.
• Integrated analysis of ATAC-seq and RNA-seq reveals the transcriptional regulation network in SLE.. International immunopharmacology, 2023.
• Integrated analysis of ATAC-seq and RNA-seq unveils the role of ferroptosis in PM2.5-induced asthma exacerbation.. International immunopharmacology, 2023.
• Integrative analysis based on ATAC-seq and RNA-seq reveals a novel oncogene PRPF3 in hepatocellular carcinoma.. Clinical epigenetics, 2024.
• Leveraging single-cell ATAC-seq and RNA-seq to identify disease-critical fetal and adult brain cell types.. Nature communications, 2024.
• Integrative single-cell RNA-seq and ATAC-seq identifies transcriptional and epigenetic blueprint guiding osteoclastogenic trajectory.. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research, 2025.
• Genome-wide chromatin profiling reveals a nonlimiting role for RXR in macrophage-like cells stimulated with multiple nuclear receptor agonists.. 2026.
• A novel TAp63γ-Airn regulatory axis governs early myogenic gene networks.. 2026.
• BRD4 PROTAC degrader enhances fulvestrant sensitivity in ER+ breast cancer via super-enhancer associated GREB1
• RNA-seq and ChIP-seq unveils thyroid hormone receptor α deficiency affects skeletal muscle myoblast proliferation and differentiation via Col6a1 during aging. Journal of Muscle Research and Cell Motility, 2025.
• RNA-seq and ChIP-seq as Complementary Approaches for Comprehension of Plant Transcriptional Regulatory Mechanism. International Journal of Molecular Sciences, 2019.
• Integrative analyses of RNA-seq and ChIP-seq Reveal MITF as a Target Gene of TFPI-2 in MDA231 Cells. Biochemical Genetics, 2023.
• Statistical approaches for the analysis of RNA-Seq and ChIP-seq data and their integration. 2012.
• Integration of ChIP-seq and RNA-seq data in structure-function analysis of cis-regulatory elements. The Thirteenth International Multiconference, 2022.
• Understanding gene regulatory mechanisms by integrating ChIP-seq and RNA-seq data: Statistical solutions to biological problems. Frontiers in Cell and Developmental Biology, 2014.
• Identification of potential target genes of USP22 via ChiP-seq and RNA-seq analysis in HeLa cells. Genetics and Molecular Biology, 2018.
• Integrative rna-seq and h3 trimethylation chip-seq analysis of human lung cancer cells isolated by laser-microdissection. Cancers, 2021.

This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.