H³NGST: a fully automated, web-based platform for end-to-end ChIP-seq analysis

Abstract

Background

Chromatin immunoprecipitation followed by sequencing (ChIP-seq) is a widely used technique for mapping protein-DNA interactions and histone modifications across the genome. Despite its utility, current analysis tools often require manual file processing, rigid input formats, and a significant level of bioinformatics expertise, posing a challenge for many experimental researchers.

Results

We present H³NGST (Hybrid, High-throughput, and High-resolution NGS Toolkit), a fully automated and user-friendly web-based platform for ChIP-seq analysis. H³NGST streamlines the entire analysis workflow, including raw data retrieval via BioProject ID, quality control, adapter trimming, reference genome alignment, peak calling, and genomic annotation. The system also categorizes peaks by genomic region, such as promoters. With minimal user input, H³NGST provides high resolution and reproducible results for both transcription factor binding and histone modification studies.

Conclusions

H³NGST significantly reduces the technical barriers to ChIP-seq analysis by eliminating the need for local installations, programming skills, or large file uploads. Its intuitive web interface and mobile accessibility extend its usability to researchers with different computational backgrounds. The platform is scalable for high-throughput studies and is freely available at https://ngschiphhh.duckdns.org.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-025-06247-5.

Background

Next-generation sequencing (NGS) technologies have transformed genomic and epigenomic research by enabling high-throughput profiling of gene regulation, chromatin states, and epigenetic modifications [1–4]. In particular, the development of epigenetic therapeutics, such as histone deacetylase (HDAC) inhibitors (e.g., TSA) and EZH2 inhibitors (e.g., GSK343), has led to an increased demand for efficient, scalable, and accurate genome-wide profiling tools [5, 6].

Among NGS-based approaches, chromatin immunoprecipitation followed by sequencing (ChIP-seq) has become a widely used method for mapping protein-DNA interactions and histone modifications across the genome [7–9]. A typical ChIP-seq workflow involves several steps: (i) acquisition of raw sequencing data, (ii) alignment to a reference genome, (iii) peak calling, and (iv) downstream interpretation and annotation [10]. Public repositories such as the Sequence Read Archive (SRA) serve as common sources of raw data, and tools are available for automated data retrieval [11]. Read alignment is a critical step in the workflow, as improper reference selection can lead to inaccurate peak detection. Aligners such as BWA-MEM are commonly used due to their speed, support for paired-end reads, and flexibility for variable read lengths [8, 12, 13]. Peak detection is typically performed using established algorithms such as HOMER [14], MACS2 [15], and SICER [16], with HOMER offering histogram-based peak modeling to reduce false positives. Final analysis steps often include genomic annotation, motif discovery, and functional enrichment to interpret regulatory regions and associated gene networks [17, 18].

While many computational tools and platforms exist for ChIP-seq analysis, few offer a fully automated, upload-free, and end-to-end solution that is simultaneously secure, scalable, and user-friendly. Existing platforms such as Galaxy [19], GenePattern [20], Cistrome Galaxy [21], CSA [22], and commercial services such as Basepair offer varying degrees of web-based functionality, but often require manual data uploads, user registration, or local software installation. In addition, these platforms often lack integrated support for automated data retrieval and full downstream annotation, which poses accessibility challenges for non-specialist users.

To overcome these limitations, we have developed H³NGST (Hybrid, High-throughput, and High-resolution NGS Toolkit), a web-based platform that enables fully automated ChIP-seq analysis from start to finish. By simply entering a BioProject ID, users can initiate a pipeline that performs data retrieval from the SRA, quality control, adapter trimming, genome alignment, peak calling, and genomic annotation. All steps are performed on the server side. No file upload or user authentication is required. All data transmissions are encrypted using SSL/TLS protocols to ensure secure processing and data integrity. H³NGST significantly reduces the technical burden associated with ChIP-seq data analysis, enabling broad access to high-resolution, reproducible results regardless of bioinformatics expertise.

Implementation

Overview of the H³NGST pipeline

H³NGST is a fully automated, web-based platform designed to streamline the analysis of ChIP-seq data from raw data acquisition to peak annotation and visualization. The pipeline consists of four main steps: (1) raw data acquisition, (2) preprocessing and quality control, (3) sequence alignment and file conversion, and (4) peak retrieval with functional annotation (Fig. 1). All processes are performed server-side, without the need for local installation or file uploads.

Fig. 1.

Overview of the H³NGST workflow. Users initiate ChIP-seq analysis by providing a BioProject accession and selecting analysis parameters via a web interface. The server-side pipeline automatically performs all computational steps using FastQC, Trimmomatic, BWA-MEM, Samtools, Bedtools, DeepTools, and HOMER, generating comprehensive results without user intervention

Retrieval of raw data

To initiate analysis, users enter a valid accession number such as a BioProject (PRJNA), SRA experiment (SRX), GEO sample (GSM), or GEO series (GSE) on the H³NGST interface. The system queries the NCBI Entrez system to resolve these accessions into corresponding SRR identifiers, which are then downloaded using the prefetch utility [11]. The retrieved sra files are converted to fastq format using fasterq-dump. During this step, the system automatically determines whether each dataset is single-end or paired-end based on the metadata in the SRA RunInfo table. This classification is critical, as paired-end data provide enhanced alignment accuracy and resolution for peak detection [23, 24]. All downstream parameters, including those for trimming, alignment, and peak calling, are automatically adjusted based on the library type.

Quality control and pre-processing

Following data retrieval, the raw FASTQ files are subjected to quality assessment using FastQC [25] to detect adapter contamination and low-quality reads. Subsequently, Trimmomatic [26] is then used to remove adapter sequences and trim low-quality bases using a sliding window approach. After trimming, FastQC [25] is run again to evaluate the quality of the processed reads. The resulting high-quality reads are retained for alignment.

Sequence alignment and file conversion

Cleaned reads are aligned to a user-specified reference genome (e.g., hg38, mm10) using BWA-MEM [12], generating SAM files. These are sorted and converted to BAM format using Samtools [27]. The pipeline then uses Bedtools [28] to convert BAM files into BED format for downstream analyses. For genome browser visualization, DeepTools [29] is used to generate BigWig signal tracks from BAM files, providing normalized coverage profiles.

Peak calling, annotation, and motif analysis

Peak calling is performed using HOMER, which supports both narrow (e.g., transcription factor binding) and broad (e.g., histone modification) peak profiles [14]. HOMER also performs motif enrichment analysis to aid in the identification of transcriptional regulators. The resulting peaks are annotated with genomic features such as gene names, proximity to transcription start sites (TSS), and functional categories [14, 23, 24]. Output includes peak coordinates, motif occurrences, and summary statistics. A complete list of all command-line tools, their representative commands, and user-defined parameters used in the H³NGST pipeline is summarized in Table 1.

Table 1.

Tools, commands, and user-defined parameters in the H³NGST pipeline

Tool	Full Command	User-defined Parameters
Prefetch	prefetch SRR_ID	Default
Fasterq-dump	fasterq-dump –split-files -O $FASTQ_DIR $SRA_FILE	Default
1st FastQC	fastqc -o $QC_DIR $FASTQ_DIR/${SRA_FILE}.fastq	Default
Trimmomatic	trimmomatic [SE/PE] -phred33 … ILLUMINACLIP:adapters.fa:2:30:10 SLIDINGWINDOW:4:10 MINLEN:20	Default
2nd FastQC	fastqc -o $QC_DIR TRIMMED.fastq	Default
Bwa mem	bwa mem $REFERENCE_GENOME TRIMMED.fastq [or paired files] > aligned.sam	Reference genome (user-defined)
Samtools	samtools view -Sb aligned.sam | samtools sort -o aligned_sorted.bam | samtools index aligned_sorted.bam	Default
Bedtools	bedtools bamtobed -i aligned_sorted.bam > aligned.bed	Default
Deeptools	bamCoverage -b aligned_sorted.bam -o output.bw –extendReads 200 –binSize 5 –normalizeUsing None	Default
MakeTagDirectory	makeTagDirectory $TAG_DIR aligned.bed -format bed	Default
FindPeaks	findPeaks $TAG_DIR -style STYLE -o auto -fdr FDR [-i CONTROL_TAG_DIR]	STYLE, FDR threshold (user-defined)
FindMotifsGenome.pl	findMotifsGenome.pl peaks.txt GENOME motif_dir -size 200 -len 8,10,12	Reference genome (user-defined)
AnnotatePeaks.pl	annotatePeaks.pl peaks.txt GENOME > annotated.txt	Reference genome, Promoter region (user-defined)

Accessing and downloading results

Upon completion, users can retrieve analysis results from the H³NGST homepage using their assigned nickname. All output files, including SAM, BAM, BED, BigWig, annotated peak tables, motif discovery results, and quality control reports, are available for direct download in standardized formats for further interpretation or publication.

Results and discussion

Functional overview of the H³NGST web platform

H³NGST is a browser-based platform that supports fully automated ChIP-seq analysis, requiring no local installation or command-line interaction. The system is designed for ease of use, allowing users, regardless of bioinformatics expertise, to perform complete analyses by simply entering a public BioProject accession number (e.g., PRJNA, SRX, GSM, GEO), assigning a nickname, and configuring minimal parameters via a guided four-step interface (Fig. 2A–D). The platform stores previously used nicknames locally in the browser and allows users to quickly reuse them via a selectable nickname history panel. Upon submission, the backend pipeline retrieves sample metadata from the NCBI SRA, automatically detects the library layout (single-end or paired-end), and performs a comprehensive analysis using a suite of open-source tools: prefetch, fasterq-dump, FastQC, Trimmomatic, BWA-MEM, Samtools, Bedtools, DeepTools, and HOMER. Parameters are dynamically adjusted based on dataset characteristics such as sequencing layout and selected peak type (narrow or broad), while users can directly specify key parameters including reference genome, peak type, promoter region, and FDR threshold to customize the analysis according to their experimental conditions. The entire workflow is executed server-side, without the need for file uploads or manual software configuration.

Fig. 2.

H³NGST web interface for ChIP-seq submission. A Users enter a BioProject ID and assign a unique analysis nickname. B The system retrieves metadata from the NCBI SRA and presents associated SRR entries for sample selection, including optional control datasets. C Users configure analysis parameters such as reference genome, peak type (narrow or broad), FDR threshold, and promoter range. D A summary page provides a final review of the selections before starting the analysis

Result retrieval and visualization

Upon completion, users can retrieve results by entering their assigned nickname on the results access page (Supplementary Fig. 1A). The results page (Supplementary Fig. 1B) includes an analysis summary table with the analysis date, scheduled deletion date, and associated SRR identifiers, along with output file icons featuring tooltips and clickable links that provide direct access to format descriptions summarized in Table 2. In addition, for the trimming step, a dedicated summary table is provided, reporting key quality metrics such as the number of input reads, surviving reads, dropped reads, and survival percentage (Supplementary Fig. 1C). The platform provides real-time updates on the status of each submitted analysis. It also includes a per-sample analysis status table that visualizes the progress of each processing step in the actual pipeline order: QC → SAM → BAM → BED → BigWig → Motif Finding → Peak Finding. Each output type is marked with a checkmark icon to indicate successful generation, and users can click these icons to directly download the corresponding result files (Supplementary Fig. 1D, top). Putative target genes linked to the identified peaks are explicitly listed in the per-sample analysis status table, enabling direct and easy access to the top candidate genes associated with each dataset (Supplementary Fig. 1D, middle). For other steps such as peak finding, numerical results like the number of peaks can be retrieved from the “findPeaks.log” file located in the logs directory (Supplementary Fig. 1D, bottom). BigWig files can be directly visualized using the UCSC Genome Browser [30] for locus-specific signal inspection (Supplementary Fig. 1E), while the Integrative Genomics Viewer (IGV) [31] allows more detailed exploration of read alignments and enrichment profiles in a local environment (Supplementary Fig. 1F). Annotated peak tables include genomic coordinates, associated genes, distances to TSS, peak types, and enrichment scores, providing essential context for downstream interpretation (Supplementary Fig. 1G). User support features include real-time feedback submission, update notifications (Supplementary Fig. 1H), and error reporting forms (Supplementary Fig. 1I). All web traffic is encrypted using SSL/TLS, and the backend is securely deployed using Gunicorn behind a reverse proxy on DuckDNS.

Table 2.

Output file types and descriptions provided by the H³NGST platform

Output File Type	File Extension	Description
FastQC Report	.html,.zip	Summarizes the quality assessment of trimmed sequencing data
SAM File	.sam	Contains reads aligned to the reference genome
BAM File	.bam	Binary format file containing reads aligned to the reference genome
BED File	.bed	Contains genomic coordinates of identified peaks
BigWig File	.bw	Efficient data format storing genome-wide signal distribution for fast visualization
Motif Results File	.html,.motif	Identifies transcription factor binding sequences and significant motifs
Annotated Peak File	.csv	Includes gene names, distances to TSS, and functional annotations for each peak

Comparison with existing platforms

Several platforms offer web-based ChIP-seq analysis [19–22]. However, these tools typically require user login, manual file uploads, or partial workflows that depend on external software. Table 3 provides a comparative overview of the available platforms. Unlike most alternatives, H³NGST uniquely supports fully automated analysis starting from a public BioProject accession, without login credentials, data upload, or manual parameter tuning, significantly lowering the technical barrier for experimental researchers.

Table 3.

Comparison of the H³NGST with existing web-based platforms for the analysis of ChIP-seq

Platform Name	Currently Usable	Free of charge	Usable without login	Usable without file upload	Fully automated	References
H3NGST	Yes	Free	Yes	Yes	Yes	This paper
Galaxy	Yes	Free	No	Yes	No	[19]
Basepair	Yes	Paid	No	No	Yes	Commercial service
Gene Pattern	Yes	Free	Guest mode	No	No	[20]
Cistrome Galaxy	Yes	Free	No	No	No	[21]
CSA	No	Free	No	No	Yes	[22]

Current capabilities and future directions

H³NGST currently supports major human (hg18, hg19, hg38) and mouse (mm9, mm10, mm39) reference genomes, and is optimized for transcription factor and histone modification ChIP-seq analysis. To optimize performance, each analysis session currently supports up to four samples. Concurrent submissions are handled reliably through a queue-based backend system. Ongoing development efforts aim to extend the platform’s capabilities to include support for plant, insect, and other animal genomes, and to integrate additional analysis modules for: RNA-seq [32, 33], single-cell RNA-seq [34], and ATAC-seq [35, 36]. These additions will enable multi-omics integration and advance the platform toward comprehensive regulatory genomics studies. By unifying multiple high-throughput sequencing pipelines within a single automated framework, H³NGST is poised to support a broader range of biological applications, including epigenomic drug screening, chromatin state profiling, and transcriptional regulatory network inference [37].

Conclusion

H³NGST represents a significant advancement in ChIP-seq data analysis by fully automating the end-to-end workflow in an intuitive, web-based environment. By integrating raw data retrieval, quality control, alignment, peak calling, and annotation into a seamless pipeline, the platform removes many of the technical barriers traditionally associated with bioinformatics tools. Users can perform comprehensive analyses without command-line expertise, file uploads, or software installation, and all results are securely delivered through an installation-free interface. The system’s support for dynamic parameter configuration, such as reference genome selection and promoter window size, further enhances its flexibility in different experimental contexts. In addition, its compatibility with mobile devices and lack of user authentication requirements improve accessibility for researchers with varying levels of computational expertise. By promoting reproducibility, scalability, and ease of use, H³NGST serves as a practical solution for large-scale epigenomic studies and broadens the access to high-throughput analysis of the regulatory genome.

Requirements

Project name: H³NGST (Hybrid, High-throughput, and High-resolution NGS Toolkit).

Project home page: https://ngschiphhh.duckdns.org

Operating system: Platform independent (web-based).

Programming language: Python, Bash.

Other requirements: Modern web browser (e.g., Google Chrome); backend server runs on Gunicorn; SSL/TLS secured via DuckDNS for free domain and certificate management.

License: Restricted academic use only.

Any restrictions to use by non-academics: Yes (academic use only).

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

HHH carried out the integration studies of the ChIP-seq analysis and participated in the design of the platform. SJU is the professor who conceived, directed and supervised the research. Both HHH and SJU drafted the manuscript.

Funding

This work was supported in part by the National Research Foundation of Korea grant funded by the Korea government (MSIT) (RS-2025–00562288) and by Korea Basic Science Institute (National Research Facilities and Equipment Center) grant funded by the Ministry of Education (RS-2023-NF001356).

Data availability

The datasets used in this study are available under GEO accession GSE26439 (BioProject PRJNA136821), including runs SRR090229 and SRR090230.

Declarations

Conflict of interest

The authors declare that they have no competing interests.