Genomic Data Types: A Primer for Bioinformatics Beginners

Genomic data encompasses the complete set of DNA sequence information, gene expression measurements, and epigenetic modifications that can be extracted from biological samples using high-throughput sequencing technologies. For students, researchers, analysts, and life-science professionals entering bioinformatics, understanding the distinctions between whole-genome sequencing, exome sequencing, transcriptome profiling, and epigenome mapping is the first step toward selecting appropriate analysis workflows. This primer defines each major genomic data type, explains the biological questions each addresses, and provides practical guidance for choosing analysis pipelines, managing data quality, and interpreting results within the limits of each technology.

The Scope of Genomic Data in Modern Biology

The scale of genomic data production has expanded dramatically over the past two decades. What began with a handful of genetic markers in the 1970s has progressed through restriction fragment length polymorphisms, microsatellites, and single nucleotide polymorphisms to the current capacity for generating thousands of complete genomes [8]. This trajectory means that modern researchers routinely work with data types that did not exist in earlier eras of molecular biology, and each data type carries distinct analytical requirements.

Genomic data can be organized into four primary categories. Whole-genome sequencing captures the complete DNA sequence of an organism. Whole-exome sequencing targets only the protein-coding regions. Transcriptome data measures RNA expression levels across all genes. Epigenome data maps chemical modifications to DNA and chromatin that influence gene activity without changing the underlying sequence. Each category answers different biological questions and requires different computational approaches.

The choice of genomic data type depends on the research question, the available sample material, the budget, and the analytical infrastructure. A researcher studying rare disease variants may choose whole-genome sequencing to capture variants in non-coding regions. A researcher investigating gene expression changes under different conditions will select RNA sequencing. A researcher examining how chromatin structure affects gene regulation will generate epigenomic data such as chromatin immunoprecipitation sequencing or methylation profiling.

Core Principles of Genomic Data Generation

From Biological Sample to Sequence Reads

All genomic data types begin with the extraction of nucleic acids from biological samples. DNA sequencing requires genomic DNA, while transcriptome analysis requires RNA that is typically converted to complementary DNA before sequencing. The quality and quantity of the starting material directly affect the quality of the resulting data.

Sequencing platforms generate short or long reads that represent fragments of the original nucleic acid molecules. These reads must be aligned to a reference genome or assembled de novo before they can be interpreted. The reference sequence database maintained by the National Center for Biotechnology Information provides annotated genomic, transcript, and protein sequence records for more than 55,000 organisms, ranging from single records to complete genomes [9]. This resource serves as the alignment target for most sequencing projects.

Reference Genomes and Annotation

Reference genomes are essential for most genomic data analysis. The RefSeq project at NCBI leverages data submitted to the International Nucleotide Sequence Database Collaboration and combines computation, manual curation, and collaboration to produce a standard set of stable, non-redundant reference sequences [9]. These references are augmented with current knowledge including publications, functional features, and informative nomenclature.

For model organisms and major crop species, community databases provide specialized resources. MaizeGDB, for example, stores genetic, genomic, sequence, gene product, functional characterization, literature reference, and person or organization contact information for Zea mays [10]. These community resources offer standardized custom interfaces that enable researchers to browse data and seek specific information matching explicit search criteria.

Data Formats and Storage

Genomic data files follow standardized formats that support interoperability across analysis tools. Raw sequencing data typically appears in FASTQ format, which contains sequence reads and quality scores. Aligned data uses SAM or BAM formats. Variant calls use VCF format. Gene expression counts may appear in matrix formats or specialized containers.

Storage requirements vary substantially by data type. Whole-genome sequencing of a human sample generates roughly 100 gigabytes of raw data. Exome sequencing generates less data because it targets a smaller portion of the genome. RNA sequencing data volume depends on the depth of coverage required for the experimental design. Epigenomic data volumes vary by assay type.

Whole-Genome Sequencing Data

Biological Questions Addressed by WGS

Whole-genome sequencing provides the most complete view of an organism's genetic material. This data type captures single nucleotide variants, small insertions and deletions, copy number alterations, and structural variants across the entire genome, including non-coding regions that may contain regulatory elements.

The clinical utility of whole-genome sequencing has been demonstrated in large-scale initiatives. The Cancer Programme of the 100,000 Genomes Project analyzed whole-genome sequencing data from 13,880 solid tumors spanning 33 cancer types, integrating genomic data with real-world treatment and outcome data within a secure research environment [7]. This project found that the incidence of somatic mutations in genes recommended for standard-of-care testing varied across cancer types. In glioblastoma multiforme, small variants were present in 94% of cases and copy number aberrations in at least one gene in 58% of cases, while sarcoma demonstrated the highest occurrence of actionable structural variants at 13% [7].

Whole-genome sequencing also supports demographic inference in evolutionary biology. The next generation sequencing revolution has multiplied the amount of genetic data for many organisms by orders of magnitude, providing information on genetic diversity, genome-wide patterns of selection, linkage disequilibrium, and recombination and mutation processes [8]. These data enable researchers to infer the past demographic history of populations.

Analysis Workflow for WGS Data

The standard whole-genome sequencing analysis pipeline begins with quality control of raw reads. Adapter sequences are removed, low-quality bases are trimmed, and reads are aligned to a reference genome. Duplicate reads are marked and removed. Variant calling identifies positions where the sample differs from the reference.

Variant annotation assigns biological meaning to identified variants. This step determines whether a variant falls within a gene, affects protein sequence, or lies in a regulatory region. Pathogenicity classification follows standardized guidelines. The American College of Medical Genetics and Genomics and the Association of Molecular Pathology published guidelines aiming to improve and standardize the pathogenicity classification of genomic variants [5]. The Clinical Sequencing Exploratory Research consortium evaluated the use of these guidelines across nine laboratories and identified the lack of a definition of cosegregation as an obstacle to consistent usage [5]. Cosegregation data differ from many other types of pathogenicity data in being quantitative, and the guidelines did not define quantitative criteria for use of these data [5].

When to Choose WGS Over Other Data Types

Whole-genome sequencing is appropriate when the research question requires complete coverage of the genome, including non-coding regions. This data type is necessary for detecting structural variants, examining regulatory regions, and performing comprehensive variant discovery. The cost per sample is higher than exome sequencing, but the data are more complete.

Whole-genome sequencing is also valuable when integrating genomic data with other data types. The linkage of whole-genome sequencing and longitudinal life course clinical data allowed the assessment of treatment outcomes for patients stratified according to pangenomic markers in the 100,000 Genomes Cancer Programme [7]. This integration demonstrated the utility of linking genomic and real-world clinical data to enable survival analysis and identify cancer genes that affect prognosis [7].

Whole-Exome Sequencing Data

Targeting Protein-Coding Regions

Whole-exome sequencing focuses on the protein-coding regions of the genome, which constitute approximately 1 to 2 percent of the total genome. This targeted approach reduces sequencing cost and data storage requirements while capturing the majority of known disease-causing variants, which tend to cluster in coding regions.

Exome sequencing is widely used in clinical diagnostics and rare disease research. The reduced data volume compared to whole-genome sequencing allows higher sequencing depth, which improves the sensitivity of variant detection. Higher depth is particularly valuable for detecting mosaic variants and variants present at low allele frequencies.

Analysis Considerations for Exome Data

The exome analysis workflow parallels whole-genome sequencing but includes an additional capture step. Sequence reads are aligned to the reference genome, but analysis is typically restricted to the targeted regions. Variant calling within exome data must account for the capture efficiency and the uniformity of coverage across targeted regions.

Coverage uniformity is a critical quality metric for exome sequencing. Regions with insufficient depth may produce false negative variant calls. Regions with excessive depth may indicate capture artifacts or PCR duplication. Quality control should include assessment of the percentage of targeted bases achieving specified depth thresholds.

Limitations of Exome Data

Exome sequencing cannot detect variants in non-coding regions, including promoters, enhancers, and intronic regulatory elements. Structural variants that do not affect coding regions may be missed. Copy number analysis from exome data is possible but less reliable than from whole-genome data due to the non-uniform capture efficiency across targeted regions.

The choice between whole-genome and exome sequencing depends on the research question and the known genetic architecture of the trait under study. For traits where coding variants are expected to explain most of the genetic contribution, exome sequencing offers a cost-effective approach. For traits where regulatory variation may be important, whole-genome sequencing provides more complete information.

Transcriptome Data from RNA Sequencing

Measuring Gene Expression

RNA sequencing measures the abundance of RNA transcripts in a biological sample. This data type provides a snapshot of gene expression at the time of sample collection and can identify differentially expressed genes between conditions, detect alternative splicing events, and quantify allele-specific expression.

The utility of RNA sequencing extends across biological disciplines. In crop research, RNA sequencing has been used to identify differentially expressed genes during testis development in the invasive pest Zeugodacus tau, leading to the identification of genes related to testis development that are conserved in tephritid species [18]. In plant science, RNA sequencing has been used to preselect novel reference genes for normalizing real-time quantitative PCR data in lettuce and wild relatives [21].

RNA-Seq Analysis Workflow

The RNA sequencing analysis pipeline begins with quality control of raw reads, followed by alignment to a reference genome or transcriptome. Read counts are generated for each gene or transcript. Normalization accounts for differences in sequencing depth and library composition across samples.

Differential expression analysis identifies genes whose expression levels differ significantly between experimental conditions. This analysis requires biological replicates to account for biological variability. The choice of statistical method depends on the experimental design and the distribution of the count data.

Reference gene selection is a critical step in validating RNA sequencing results. RNA sequencing is useful to preselect novel reference genes, although validation by quantitative PCR is still advisable [21]. In a study of anthocyanin-related genes in lettuce, six candidate reference genes with low variation in expression according to RNA sequencing analyses were selected for future expression studies, and the stability of these candidates was assessed by four different analytical tools [21].

Single-Cell RNA Sequencing

Single-cell RNA sequencing extends transcriptome analysis to individual cells, revealing heterogeneity within cell populations that is masked by bulk RNA sequencing. This data type requires specialized analysis methods for cell clustering, cell type annotation, and trajectory inference.

Cell type annotation remains a challenge in single-cell RNA sequencing analysis. Manual annotation of clusters of cells is a challenge yet to be addressed more scalably and faster [20]. Supervised techniques have obtained more robust results compared to clustering methods, and a complementary step of feature selection helped support vector machines outperform other classifiers in different scenarios [20]. XGBoost has been shown to automatically annotate cell types in a simpler and more scalable framework [20].

Integrating RNA-Seq with Other Data Types

Transcriptome data gains additional value when integrated with other genomic data types. Radiogenomics approaches combine imaging data with RNA sequencing data to improve diagnostic accuracy. In muscle-invasive bladder cancer, a model combining magnetic resonance imaging features with bulk RNA sequencing data achieved mean sensitivity, specificity, and accuracy of 94%, 88%, and 92% in differentiating intra-versus extra-bladder cancer [19]. This combined model demonstrated improvement compared to genetic-based and radiomic-based models alone [19].

Epigenome Data

Types of Epigenomic Assays

Epigenome data captures chemical modifications to DNA and chromatin that regulate gene expression without altering the DNA sequence. Common epigenomic assays include chromatin immunoprecipitation sequencing for mapping protein-DNA interactions, methylation sequencing for identifying methylated cytosines, and chromatin conformation capture for mapping three-dimensional genome architecture.

Histone modifications play essential roles in three-dimensional genome organization. H3 lysine 9 dimethylation, one of the abundant histone modifications, has been investigated for its functional roles in genome organization [11]. Inhibition of methyltransferases G9a and GLP in differentiated cells eliminated H3K9me2 predominantly at active genomic compartments, while the level of residual H3K9me2 modifications was strongly associated with inactive genomic compartments [11].

Analysis of Epigenomic Data

Epigenomic analysis workflows vary by assay type. Chromatin immunoprecipitation sequencing analysis involves peak calling to identify regions of protein-DNA interaction. Methylation analysis involves comparing methylation levels at individual cytosines or across genomic regions. Chromatin conformation analysis involves constructing contact maps that reveal the spatial organization of the genome.

The interpretation of epigenomic data requires integration with other data types. Chemical inhibition of G9a and GLP in mouse hepatocytes led to decreased chromatin-nuclear lamina interactions mainly at G9a and GLP sensitive regions, increased degree of genomic compartmentalization, and up-regulation of hundreds of genes that were associated with alterations of the three-dimensional chromatin [11]. These findings demonstrate that epigenomic data must be interpreted in the context of gene expression and genome architecture.

Integrative Epigenomic Analysis

Epigenomic data are often combined with transcriptome data to understand how chromatin state influences gene expression. Integrative methods using genomic data sets have become a major area of interest in cancer research [12]. Confidence Based Integration is an integration method that extracts similar as well as complementing information from genomic data sets, using expression data sets of gene, miRNA, and DNA methylation in fusion experiments on five different cancer types [12].

At a Glance: Genomic Data Types and Analysis Pipelines

Practical Implementation Steps for Genomic Data Analysis

Step 1: Define the Biological Question

The first step in any genomic data analysis project is to define the biological question precisely. This definition determines the choice of data type, the experimental design, and the analysis workflow. A question about genetic variants associated with a disease requires DNA sequencing data. A question about gene expression changes under treatment requires RNA sequencing data. A question about regulatory mechanisms requires epigenomic data.

Step 2: Select the Appropriate Data Type

The choice of data type should follow directly from the biological question. Whole-genome sequencing provides complete variant information but at higher cost. Exome sequencing targets coding regions at lower cost. RNA sequencing measures expression. Epigenomic assays measure regulatory modifications. Consider the tradeoffs between data completeness, cost, and analytical complexity.

Step 3: Design the Experiment with Replicates

Biological replicates are essential for most genomic analyses. Replicates capture biological variability and enable statistical inference. Technical replicates, which involve sequencing the same sample multiple times, assess technical variability but do not substitute for biological replicates. The number of replicates required depends on the expected effect size and the variability of the system.

Step 4: Perform Quality Control

Quality control should be performed at multiple stages of the analysis pipeline. Raw sequencing reads should be assessed for quality scores, adapter contamination, and GC content. Aligned data should be assessed for mapping rates, coverage uniformity, and duplication rates. The results of quality control should be documented and used to make decisions about data inclusion or exclusion.

Step 5: Document the Analysis Pipeline

Reproducibility requires complete documentation of the analysis pipeline, including software versions, parameter settings, and reference genome versions. The FAIR Guiding Principles provide a framework for making data findable, accessible, interoperable, and reusable [4]. These principles apply to both the raw data and the analysis outputs.

Step 6: Interpret Results Within Biological Context

Genomic data analysis produces lists of variants, genes, or regions that must be interpreted within biological context. This interpretation requires knowledge of the biological system under study, the limitations of the data type, and the statistical methods used. Results should be validated using orthogonal methods where possible.

Records and Measurements for Genomic Projects

Essential Records to Maintain

Genomic data projects generate large volumes of metadata that must be recorded systematically. Sample identifiers should be linked to source material, collection date, and processing history. Sequencing run information should include platform, flow cell identifier, and run date. Analysis records should document software versions, reference genome builds, and parameter settings.

The National Institutes of Health Genomic Data Sharing Policy establishes expectations for the sharing of genomic data generated from NIH-funded research [3]. This policy addresses data submission, data sharing timelines, and the protection of research participant privacy. Researchers should review the policy requirements applicable to their funding sources and research context.

Quality Metrics to Track

Quality metrics provide objective measures of data quality that support decisions about data inclusion and interpretation. For sequencing data, key metrics include read depth, coverage uniformity, mapping rate, and duplication rate. For expression data, key metrics include library size, gene detection rate, and correlation between replicates.

The RefSeq database at NCBI provides a standard set of stable, non-redundant reference sequences that support quality assessment across projects [9]. These references are augmented with current knowledge including publications, functional features, and informative nomenclature [9].

Common Failure Patterns in Genomic Data Analysis

Insufficient Sequencing Depth

Insufficient sequencing depth is a common cause of failed variant detection. Low depth regions produce false negative calls, particularly for heterozygous variants. The required depth depends on the data type and the application. Clinical applications typically require higher depth than discovery research.

Batch Effects and Technical Artifacts

Batch effects arise when samples are processed in different batches, introducing technical variation that can be mistaken for biological variation. These effects are particularly problematic in large studies where samples are processed over extended periods. Experimental design should include randomization of samples across batches and the inclusion of technical controls.

Misalignment and Mapping Errors

Alignment errors occur when reads map to incorrect genomic locations. These errors are more common in repetitive regions and in samples with structural variants. The choice of alignment tool and parameters affects mapping accuracy. The reference genome version should be consistent across all samples in a study.

Overinterpretation of Results

Genomic data analysis can produce large numbers of candidate variants or genes, and the risk of overinterpretation increases with the number of tests performed. Multiple testing correction is essential for most genomic analyses. Results should be interpreted with appropriate statistical rigor and validated in independent cohorts where possible.

Limitations of Genomic Data Types

Technical Limitations

Each genomic data type has technical limitations that affect the interpretation of results. Whole-genome sequencing may miss variants in repetitive regions that cannot be uniquely aligned. Exome sequencing may miss variants in poorly captured regions. RNA sequencing measures steady-state RNA levels, which reflect both transcription and degradation. Epigenomic assays measure average modifications across cell populations.

Biological Limitations

Genomic data provide a static snapshot of the biological state at the time of sample collection. Gene expression levels fluctuate over time and in response to environmental conditions. Epigenetic modifications are dynamic and cell-type specific. A single time point may not capture the full range of biological variation.

Interpretation Limitations

The interpretation of genomic data requires knowledge of the biological context and the limitations of the analytical methods. Variant pathogenicity classification is particularly challenging. The American College of Medical Genetics and Genomics and the Association of Molecular Pathology guidelines aim to improve and standardize the pathogenicity classification of genomic variants, but the lack of a definition of cosegregation was identified as an obstacle to consistent usage [5].

Safety and Regulatory Context

Data Privacy and Security

Genomic data can identify individuals and reveal sensitive information about health status and family relationships. Researchers must implement appropriate data security measures to protect participant privacy. The National Institutes of Health Genomic Data Sharing Policy addresses the protection of research participant privacy and the responsible sharing of genomic data [3].

Data Sharing Requirements

Many funding agencies require genomic data to be shared in public repositories. The NCBI Data Resources provide access to a wide range of genomic data and analysis tools [2]. The EMBL-EBI Training program offers educational resources for researchers working with genomic data [1]. Researchers should identify the appropriate repositories for their data types and comply with data sharing requirements.

Ethical Considerations

Genomic data analysis raises ethical considerations related to informed consent, return of results, and the potential for discrimination. Researchers should consider these issues when designing studies and interpreting results. The integration of genomic data with clinical data requires attention to the potential implications for research participants.

Professional Escalation Criteria

When to Seek Specialized Expertise

Genomic data analysis often requires specialized expertise that goes beyond the capabilities of general bioinformatics training. Consider seeking specialized expertise when the analysis requires advanced statistical methods, when the data type is unfamiliar, or when the results will be used for clinical decision-making.

When to Consult Clinical Genetics

Variant interpretation for clinical applications requires expertise in clinical genetics. The pathogenicity classification of genomic variants should follow standardized guidelines, and the quantitative criteria for cosegregation data should be applied appropriately [5]. Clinicians should be consulted when variant interpretation will affect patient care.

When to Escalate Data Quality Issues

Data quality issues that cannot be resolved through standard quality control procedures should be escalated to the sequencing facility or the appropriate technical experts. Persistent quality problems may indicate issues with sample preparation, sequencing reagents, or instrument performance.

Frequently Asked Questions

What is the difference between whole-genome sequencing and whole-exome sequencing?

Whole-genome sequencing captures the complete DNA sequence of an organism, including both coding and non-coding regions. Whole-exome sequencing targets only the protein-coding regions, which constitute approximately 1 to 2 percent of the genome. Whole-genome sequencing provides more complete information but at higher cost and with greater data storage requirements. Exome sequencing is more cost-effective for studies where coding variants are the primary focus.

How do I choose between RNA sequencing and microarray for gene expression analysis?

RNA sequencing provides a more complete view of the transcriptome, including the ability to detect novel transcripts, alternative splicing events, and allele-specific expression. Microarrays are limited to known probes and have a narrower dynamic range. RNA sequencing is generally preferred for discovery research, while microarrays may be appropriate for well-characterized gene sets where the additional information from RNA sequencing is not needed.

What is the minimum sequencing depth required for reliable variant calling?

The required sequencing depth depends on the data type, the application, and the variant type being detected. Clinical applications typically require higher depth than discovery research. Germline variant calling from whole-genome sequencing typically requires 30-fold coverage, while somatic variant calling in cancer samples may require higher depth to detect variants present at low allele frequencies. The specific requirements should be determined based on the research question and the expected variant allele frequencies.

How many biological replicates are needed for differential expression analysis?

The number of biological replicates required depends on the expected effect size, the biological variability of the system, and the statistical power desired. Three replicates per condition is a common minimum, but more replicates may be needed for subtle effects or highly variable systems. Power analysis can help determine the appropriate number of replicates for a specific experimental design.

What is the difference between bulk RNA sequencing and single-cell RNA sequencing?

Bulk RNA sequencing measures the average gene expression across all cells in a sample, masking cellular heterogeneity. Single-cell RNA sequencing measures gene expression in individual cells, revealing differences between cell types and states within a population. Single-cell RNA sequencing requires specialized analysis methods for cell clustering and cell type annotation [20].

How do I integrate multiple types of genomic data in a single analysis?

Integrative analysis combines information from multiple genomic data types to gain insights that would not be possible from any single data type. Methods such as Confidence Based Integration extract similar as well as complementing information from genomic data sets [12]. The choice of integration method depends on the data types being combined and the biological question being addressed.

What are the main challenges in analyzing epigenomic data?

Epigenomic data analysis faces challenges related to the dynamic nature of epigenetic modifications, the cell-type specificity of these modifications, and the technical variability of epigenomic assays. The interpretation of epigenomic data requires integration with gene expression data and knowledge of the three-dimensional genome organization [11].

How should I handle missing genomic data in my analysis?

Missing genomic data can arise from failed sequencing reactions, poor sample quality, or technical artifacts. The approach to handling missing data depends on the amount and pattern of missingness. Methods for imputation should ensure the biological plausibility of the imputed features [14]. When genomic data is incomplete, most methods perform feature imputation solely by learning from data distributions, without ensuring the biological plausibility of the imputed features [14].

Related Bioinformatics Guides

• Transcriptome Assembly Without a Reference Genome
• Structure-Based Drug Design in Bioinformatics: Computational Pipelines, Active Site Grid Mapping, and Virtual Screening Workflows
• ChIP-Seq Bioinformatics Workflows
• Predicting AMR from Genomic Data
• Data Sharing and Privacy in Genomic Research

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.
• Consideration of Cosegregation in the Pathogenicity Classification of Genomic Variants.. American journal of human genetics, 2016.
• Combined visualization of genomic and epidemiological data for outbreaks.. Epidemiology and infection, 2024.
• Insights for precision oncology from the integration of genomic and clinical data of 13,880 tumors from the 100,000 Genomes Cancer Programme.. Nature medicine, 2024.
• Demographic inference.. Current biology : CB, 2021.
• Reference sequence (RefSeq) database at NCBI: current status, taxonomic expansion, and functional annotation.. Nucleic acids research, 2016.
• MaizeGDB.. Methods in molecular biology (Clifton, N.J.), 2007.
• G9a/GLP-sensitivity of H3K9me2 Demarcates Two Types of Genomic Compartments.. Genomics, proteomics & bioinformatics, 2020.
• Identifying cancer sub-types from genomic scale data sets using confidence based integration (CBI).. Journal of biomedical informatics, 2022.
• Beyond traditional outbreak investigation: Using genomic data for enhanced detection of COVID-19 disease clusters in Utah.. 2026.
• Causality-Guided Diffusion and Fusion of incomplete multi-modal data for robust survival prognosis.. 2026.
• GREAC: An Open-source software for Genome Region Extraction and Classification. 2026.
• Can Artificial Intelligence Transform Early Warning for Antimicrobial-Resistant Outbreak Clones? Approaches, Gaps, and Opportunities: A Scoping Review. 2026.
• Deep learning for genomic insights into athletic performance in sports education.. 2026.
• Identification of testis development-related genes by combining Iso-Seq and RNA-Seq in Zeugodacus tau. Frontiers in Cell and Developmental Biology, 2024.
• MRI/RNA-Seq-Based Radiogenomics and Artificial Intelligence for More Accurate Staging of Muscle-Invasive Bladder Cancer. International Journal of Molecular Sciences, 2023.
• Cell Type Annotation Model Selection: General-Purpose vs. Pattern-Aware Feature Gene Selection in Single-Cell RNA-Seq Data. Genes, 2023.
• Selection of Novel Reference Genes by RNA-Seq and Their Evaluation for Normalising Real-Time qPCR Expression Data of Anthocyanin-Related Genes in Lettuce and Wild Relatives. International Journal of Molecular Sciences, 2023.
• Integrative analysis of multiple types of genomic data using an accelerated failure time frailty model. Computational Statistics, 2021.
• A stochastic carcinogenesis model incorporating multiple types of genomic instability fitted to colon cancer data. Journal of Theoretical Biology, 2008.
• Integrative Analysis of Genomic Data Types and AI Methodologies in Healthcare Applications. 2nd International Conference on Cyber Resilience Iccr 2024, 2024.
• Specific genomic regions are differentially affected by copy number alterations across distinct cancer types, in aggregated cytogenetic data. Plos One, 2012.

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