Initial Publication Date: October 24, 2026
DOI | Cite this

Testing the Ortholog Conjecture: Evolution of Gene Expression During Mammalian Embryonic Development

SUSANTA BEHURA, University of Missouri-Columbia

Location: Missouri

Abstract

This Course-Based Undergraduate Research Experience (CURE) introduces students to authentic computational research in evolutionary developmental biology by investigating how gene duplication and evolutionary selection shape gene expression during mammalian embryonic development. The central research question is whether orthologous genes exhibit more conserved expression patterns across species than paralogous genes, as predicted by the ortholog conjecture, or whether expression similarity is more strongly influenced by developmental stage and cellular context.

Students first learn fundamental concepts of gene duplication, orthology, paralogy, natural selection, and functional evolution, followed by hands-on training in Unix/Linux-based computational tools and analysis of publicly available transcriptomic datasets. Using embryonic gene expression data from multiple mammalian species and orthology/paralogy annotations from Ensembl, students apply a reproducible computational analysis pipeline to compare evolutionary patterns of gene expression. Different student groups investigate distinct gene families associated with reproduction and embryonic development, formulate hypotheses, analyze expression conservation and divergence, and interpret their findings in the context of evolutionary mechanisms, including functional conservation and neofunctionalization.

This entirely computational CURE integrates classroom learning with original research without requiring specialized laboratory infrastructure. Students develop skills in bioinformatics, comparative genomics, data analysis, hypothesis testing, and scientific communication while generating potentially novel insights into the evolutionary regulation of mammalian embryonic development. The modular design allows instructors to adapt the experience to different undergraduate courses and levels of computational expertise.

Student Goals

To develop a conceptual understanding of evolutionary developmental biology and functional genomics, including gene duplication, orthology, paralogy, natural selection, and gene expression evolution, and apply these concepts to formulate testable hypotheses about the evolutionary regulation of mammalian embryonic development.
To develop practical skills in Unix/Linux-based computational analysis, bioinformatics, and comparative genomics by retrieving publicly available embryonic transcriptomic datasets, identifying orthologous and paralogous genes, and applying reproducible computational workflows and statistical methods to investigate evolutionary patterns of gene expression.
To develop independent scientific inquiry, critical thinking, collaborative research, and scientific communication skills by designing and testing original research hypotheses, interpreting computational results in the context of evolutionary developmental biology, evaluating alternative explanations, and communicating findings through scientific presentations and research reports.

Research Goals

To test the ortholog conjecture by investigating the evolutionary conservation and divergence of gene expression among orthologous and paralogous genes across mammalian species and embryonic developmental stages, and determine whether gene expression similarity is more strongly associated with evolutionary relationships or developmental and cellular contexts.
To investigate how gene duplication and evolutionary selection contribute to the diversification of gene expression during mammalian embryonic development, focusing on developmental gene families to identify patterns of expression conservation, divergence, and potential neofunctionalization among paralogous genes.

Context

This CURE is designed to be integrated into an upper-level undergraduate course in Animal Development, with a recommended enrollment of 15–30 students. It will connect fundamental concepts of animal embryogenesis, developmental gene regulation, and evolutionary developmental biology with authentic computational research. The CURE can be implemented over a full semester (approximately 14–16 weeks), with students working in small groups of 3–5 to investigate different gene families involved in mammalian embryonic development.

During the initial 3–4 weeks, students will learn about gene duplication, orthology, paralogy, evolutionary selection, and the regulation of gene expression during embryogenesis, along with introductory Unix/Linux commands and bioinformatics tools. During the following 7–9 weeks, students will retrieve publicly available embryonic transcriptomic datasets from different mammalian species, identify orthologous and paralogous genes using Ensembl annotations, and apply reproducible computational workflows to investigate evolutionary patterns of gene expression. The final 2–3 weeks will focus on interpreting findings in the context of animal development and evolution, evaluating research hypotheses, and communicating results through scientific presentations and research reports.

Students should have foundational knowledge of genetics, molecular biology, and basic developmental biology. Prior experience in programming, Unix/Linux, or bioinformatics is not required, as introductory training, guided exercises, and computational scripts will be provided. The CURE is entirely computational and requires only computers with internet access and a Unix/Linux environment, making it accessible without specialized laboratory infrastructure. Its modular structure allows instructors to adapt the research activities to different class sizes, course schedules, and levels of student preparation.

Target Audience:Major, Upper Division
CURE Duration:A full term

CURE Design

The central research theme of this CURE is to understand how gene duplication, evolutionary selection, and changes in gene expression regulation contribute to the evolution of mammalian embryonic development. The project focuses on testing the ortholog conjecture, which predicts greater functional conservation among orthologous genes than paralogous genes, while investigating whether similarities in embryonic gene expression are more strongly associated with evolutionary relationships, developmental stages, or cellular contexts. Students will also explore how expression divergence following gene duplication may contribute to the evolution of developmental gene regulation and potentially novel gene functions.

The CURE is integrated into an undergraduate Animal Development course and follows a structured progression from conceptual learning to independent computational research. Initially, students will learn fundamental principles of embryogenesis, gene duplication, orthology, paralogy, natural selection, and evolutionary developmental biology through lectures, discussions of primary research literature, and guided exercises. Students will then receive hands-on training in Unix/Linux-based computational tools, public transcriptomic databases, Ensembl gene annotations, and reproducible bioinformatics workflows.

Students will work in small research groups, with each group investigating a distinct gene family known to influence mammalian reproduction or embryonic development. Using publicly available transcriptomic datasets from multiple mammalian species, students will formulate testable hypotheses, identify orthologous and paralogous genes, extract developmental gene expression profiles, and perform comparative and statistical analyses to evaluate expression conservation and divergence. They will examine whether orthologous genes exhibit greater expression similarity than paralogous genes and assess how developmental stage and cellular context influence these relationships. Students will also investigate patterns of expression divergence among duplicated genes and consider their potential evolutionary significance.

The learning environment will combine instructor-guided training with progressively independent, collaborative research. To ensure that students with different levels of computational experience can participate successfully, the course will provide step-by-step tutorials, example datasets, annotated analysis scripts, and clearly defined research milestones. Regular group discussions, instructor feedback, peer collaboration, and troubleshooting sessions will help students develop confidence in computational analysis and scientific reasoning. As students gain experience, they will take increasing responsibility for selecting research questions, interpreting results, evaluating alternative explanations, and proposing follow-up investigations.

An important feature of this CURE is that students will not simply reproduce predetermined results. Different research groups will investigate different developmental gene families, allowing them to generate original findings that contribute to a broader comparative understanding of mammalian embryonic evolution. Students will integrate their results across groups, discuss similarities and differences among gene families, and communicate their findings through research presentations and written scientific reports. This approach connects classroom learning with authentic scientific discovery while developing students' abilities to formulate hypotheses, conduct reproducible computational research, critically evaluate evidence, and understand how genomic mechanisms contribute to the evolution of animal development.

The primary stakeholders outside the classroom include researchers in evolutionary developmental biology, comparative genomics, reproductive biology, and functional genomics who are interested in understanding how gene duplication and evolutionary changes in gene expression contribute to mammalian embryonic development. The research questions are aligned with broader scientific interests in gene function evolution, developmental gene regulation, and the evolutionary conservation and divergence of gene expression. These interests will help guide the selection of biologically relevant gene families, comparative analyses, and interpretation of student findings.

Students will share their research through scientific presentations, undergraduate research symposia, and written reports. Validated computational workflows, datasets, and analysis results may also be shared through publicly accessible repositories such as GitHub, allowing other researchers and educators to reproduce and extend the analyses. Findings from different student groups will be integrated to identify broader evolutionary patterns and generate new research hypotheses. Results that provide novel scientific insights may contribute to collaborative research projects, conference presentations, or peer-reviewed publications, with appropriate recognition of student contributions. This approach will connect undergraduate classroom research with the broader scientific community while providing students with experience in communicating original research beyond the classroom.

Core Competencies: Analyzing and interpreting data, Asking questions (for science) and defining problems (for engineering), Constructing explanations (for science) and designing solutions (for engineering), Developing and using models, Planning and carrying out investigations
Nature of Research:Basic Research, Informatics/Computational Research

Tasks that Align Student and Research Goals

Research Goals →
Student Goals ↓

Research Goal 1: To test the ortholog conjecture by investigating the evolutionary conservation and divergence of gene expression among orthologous and paralogous genes across mammalian species and embryonic developmental stages, and determine whether gene expression similarity is more strongly associated with evolutionary relationships or developmental and cellular contexts.

Research Goal 2: To investigate how gene duplication and evolutionary selection contribute to the diversification of gene expression during mammalian embryonic development, focusing on developmental gene families to identify patterns of expression conservation, divergence, and potential neofunctionalization among paralogous genes.

Student Goal 1: To develop a conceptual understanding of evolutionary developmental biology and functional genomics, including gene duplication, orthology, paralogy, natural selection, and gene expression evolution, and apply these concepts to formulate testable hypotheses about the evolutionary regulation of mammalian embryonic development.

Students will review primary literature on the ortholog conjecture, gene duplication, and gene expression evolution during mammalian embryonic development. Each student group will select a developmentally relevant gene family, identify orthologous and paralogous relationships using Ensembl annotations, and formulate at least two testable hypotheses about gene expression conservation and divergence across mammalian species and embryonic stages. Students will document their hypotheses, biological rationale, and predicted outcomes in a short research proposal.

Students will investigate the evolutionary history of selected gene families involved in mammalian embryonic development by examining gene duplication events, phylogenetic relationships, and functional annotations. Each group will construct or interpret a gene family phylogeny, identify major duplication and speciation events, and develop testable predictions about how duplicated genes may have undergone expression divergence, functional conservation, or neofunctionalization during evolution. Students will summarize their evolutionary interpretations and predictions in a short written assignment.

Student Goal 2: To develop practical skills in Unix/Linux-based computational analysis, bioinformatics, and comparative genomics by retrieving publicly available embryonic transcriptomic datasets, identifying orthologous and paralogous genes, and applying reproducible computational workflows and statistical methods to investigate evolutionary patterns of gene expression.

Students will retrieve publicly available embryonic transcriptomic datasets from at least three mammalian species, extract expression profiles of selected orthologous and paralogous genes, and organize the data using Unix/Linux-based computational tools. They will apply reproducible analysis scripts to compare expression patterns across species and developmental stages, calculate expression similarity measures, and perform statistical tests to evaluate the ortholog conjecture. Each group will submit processed datasets, documented analysis scripts, and graphical summaries of its findings.

Students will use Unix/Linux-based computational workflows and comparative transcriptomic datasets to quantify expression divergence among paralogous genes across mammalian species and embryonic developmental stages. They will generate expression heatmaps, developmental expression profiles, and statistical comparisons to identify conserved and divergent expression patterns. Where suitable data are available, students will examine associations between expression divergence, gene duplication history, and developmental or cellular context. Each group will submit reproducible analysis scripts, figures, and a summary table of its results.

Student Goal 3: To develop independent scientific inquiry, critical thinking, collaborative research, and scientific communication skills by designing and testing original research hypotheses, interpreting computational results in the context of evolutionary developmental biology, evaluating alternative explanations, and communicating findings through scientific presentations and research reports.

Students will critically evaluate whether their comparative gene expression results support or challenge the predictions of the ortholog conjecture. Each group will interpret statistical results, examine alternative explanations involving developmental stage and cellular context, and compare findings with those of other student groups. Students will communicate their conclusions through a scientific presentation and written research report that includes hypotheses, methods, results, limitations, and proposed follow-up questions.

Students will integrate their comparative expression analyses with evolutionary and developmental evidence to evaluate possible mechanisms underlying the diversification of duplicated genes. Each group will identify at least one gene pair showing potentially meaningful expression divergence, critically assess whether the findings are consistent with neofunctionalization or other evolutionary mechanisms, and propose follow-up analyses or experiments to distinguish among alternative explanations. Students will present their findings in a final research symposium and prepare a written report highlighting their original contributions, limitations, and future research directions.

Instructional Materials

The instructional materials for this CURE will be developed from my published research on the evolutionary regulation of paralogous gene expression and an existing computational pipeline developed in my laboratory. A foundational research article is Islam and Behura (2024), "Role of paralogs in the sex-bias transcriptional and metabolic regulation of the brain-placental axis in mice," published in Placenta (145:143–150; https://doi.org/10.1016/j.placenta.2023.12.019). This study provides an example of how gene expression data, Ensembl homology annotations, gene phylogenies, and Bayesian evolutionary modeling can be integrated to investigate the evolutionary regulation of duplicated genes. Students will study the conceptual framework and computational methodology of this publication before extending the approach to comparative analyses of mammalian embryonic development.

The primary computational resource will be my publicly available GitHub pipeline, "Pipeline to Model Evolution of Paralog Expression," available at https://github.com/sbehura392/FUNCTIONAL_GENOMICS_SCRIPTS_BEHURALAB/blob/main/Pipeline_to_model_evoluion_of_paralog_expression.txt. This pipeline will serve as the foundation for hands-on training in Unix/Linux-based bioinformatics and evolutionary analysis of gene expression. Students will learn to organize genomic datasets, identify paralogous gene families, analyze gene sequence relationships, construct phylogenetic trees, and investigate associations between evolutionary relationships and gene expression variation. The computational exercises will progressively move from instructor-guided examples to independent analyses of student-selected developmental gene families.

Students will obtain publicly available embryonic transcriptomic datasets from resources including EmExplorer, DevOmics, dbEmbryo, TEDD, EMBL-EBI Single Cell Expression Atlas, and NCBI Gene Expression Omnibus (GEO). These databases will provide gene expression information from different mammalian species and embryonic developmental stages. Publicly available gene homology and functional annotation data from Ensembl will be used to identify orthologous and paralogous genes and establish evolutionary relationships among members of selected gene families.

Instructional activities will include lectures introducing gene duplication, the ortholog conjecture, natural selection, and developmental gene regulation; guided tutorials on accessing and processing public genomic datasets; and computational exercises involving gene family identification, expression comparisons, phylogenetic analysis, and statistical modeling. Students will use the published computational framework to test whether gene expression conservation is more strongly associated with orthology/paralogy relationships or developmental and cellular contexts. Different student groups will investigate distinct gene families involved in mammalian embryogenesis, allowing their analyses to contribute to a broader comparative research question.

The CURE will require computers with internet access, a Unix/Linux computational environment, publicly available bioinformatics software, and access to the GitHub analysis pipeline. No specialized wet-laboratory equipment or experimental samples will be required. Instructional materials will include research articles, database access instructions, annotated computational scripts, example datasets, analysis guidelines, and assignments for interpreting and communicating research findings. Together, these resources will provide a reproducible and adaptable framework for integrating authentic computational research into an undergraduate Animal Development course.

Computational_Pipeline_Student_Guide (Acrobat (PDF) 18kB Oct15 26)
Referece publication (Acrobat (PDF) 3.2MB Oct15 26)

Assessment

Student_Assessment (Acrobat (PDF) 79kB Oct15 26)

Instructional Staffing

The CURE will be coordinated by the course instructor, who will provide expertise in evolutionary developmental biology, functional genomics, comparative genomics, and computational biology. The instructor will introduce the scientific concepts underlying the research, provide training in Unix/Linux-based bioinformatics, and guide students in using the Behura Lab computational pipeline to investigate the evolution of gene expression during mammalian embryonic development.

Students will work collaboratively in small research groups, with peer mentoring encouraged throughout the computational exercises and independent research activities. Where available, graduate teaching assistants or advanced undergraduate students with computational biology experience may assist with software setup, troubleshooting, data processing, and interpretation of analytical results. These assistants will become familiar with the CURE through the published research article, GitHub pipeline, instructional tutorials, and example analyses provided by the instructor.

The instructor will oversee the selection of research questions and gene families, ensure the scientific rigor and reproducibility of computational analyses, and provide feedback on student research proposals, presentations, and final reports. Students will progressively assume responsibility for developing hypotheses, conducting analyses, interpreting findings, and communicating their research.

The CURE is designed to be implemented primarily by a single instructor using publicly available genomic resources and computational tools, without requiring dedicated laboratory personnel or specialized experimental facilities. Additional teaching assistants or research collaborators can enhance student support but are not essential for implementing the CURE.

Author Experience

SUSANTA BEHURA, University of Missouri-Columbia

My motivation for developing this CURE is to integrate undergraduate education with authentic research in evolutionary developmental biology and functional genomics. I want students to move beyond learning computational tools as isolated technical skills and instead use them to investigate fundamental biological questions about how gene duplication and evolutionary selection shape gene expression during mammalian embryonic development. By analyzing publicly available transcriptomic datasets, students will gain hands-on experience in computational biology, hypothesis testing, and scientific interpretation while contributing potentially novel findings to ongoing research. This CURE is also intended to make original research accessible to undergraduates without requiring expensive laboratory infrastructure and to encourage students to develop independent research questions at the intersection of genomics, evolution, and developmental biology.

Advice for Implementation

Although this CURE has not yet been formally implemented in an undergraduate classroom, the computational research framework is based on my published research and an existing bioinformatics pipeline developed in my laboratory.

For successful implementation, I recommend introducing students to the biological concepts of gene duplication, orthology, paralogy, and gene expression evolution before beginning computational exercises. Students should initially work with a small, instructor-prepared example dataset to become familiar with Unix/Linux commands, gene annotations, and the analysis workflow before independently investigating selected gene families.

Potential challenges include limited prior programming experience, difficulties matching gene identifiers across databases, differences in developmental-stage annotations among species, and interpreting statistical results in an evolutionary context. These challenges can be addressed through annotated scripts, guided tutorials, standardized input files, regular group discussions, and instructor feedback. Particular attention should be given to distinguishing gene expression divergence from direct evidence of natural selection or neofunctionalization.

The CURE requires computers with internet access, a Unix/Linux environment, publicly available bioinformatics software, and the Behura Lab computational pipeline available on GitHub. Public embryonic transcriptomic datasets and Ensembl gene homology annotations will serve as the primary research resources. No specialized laboratory equipment or experimental supplies are required.

I recommend organizing students into small groups, assigning different developmental gene families, and using regular research milestones to monitor progress. This approach should allow students with different levels of computational experience to participate meaningfully while generating reproducible results that address shared evolutionary developmental biology questions.

Iteration

Iteration, troubleshooting, and critical evaluation of computational predictions will be integral components of this CURE. Students will learn that computational research involves not only analyzing biological data but also evaluating the reliability of the annotations and assumptions underlying their analyses.

An important learning objective will be to understand how uncertainty in orthology and paralogy predictions can confound comparative gene expression analyses and influence conclusions about the ortholog conjecture. Students will independently obtain gene homology annotations from Ensembl and NCBI, compare the predicted orthologous and paralogous relationships, and identify inconsistencies between the two annotation resources. They will then apply stringent confidence criteria to select gene relationships supported by both resources and repeat their expression analyses using this higher-confidence dataset.

By comparing results obtained from the initial and refined gene annotations, students will evaluate how annotation uncertainty affects estimates of expression conservation and divergence, statistical significance, and the strength of evidence supporting or challenging the ortholog conjecture. This exercise will provide an opportunity to understand how improving input data quality can strengthen the reliability of biological inferences.

Students will also troubleshoot computational errors, inconsistent gene identifiers, missing expression values, and differences in developmental-stage annotations across species. If their initial hypotheses are not supported, they will investigate alternative explanations involving developmental stage, tissue specificity, cellular context, or gene duplication history rather than assuming that unexpected findings represent analytical failures.

The course will include regular research discussions, intermediate milestones, and dedicated time for repeating analyses, refining datasets, and evaluating alternative interpretations. Students will document the problems encountered, corrective actions taken, and changes in their conclusions in their final research reports.

Through this iterative process, students will learn that scientific conclusions depend on the quality of underlying data, the reliability of computational predictions, and the robustness of analytical methods. Most importantly, they will gain experience in critically evaluating and improving their own research rather than simply accepting the initial results generated by computational tools.

Using CURE Data

The data and analyses generated by student groups will be integrated into a shared comparative genomics framework to address the two central research questions: whether orthologous genes exhibit greater conservation of expression than paralogous genes during mammalian embryonic development, and how gene duplication contributes to evolutionary diversification of developmental gene expression.

Each student group will investigate a distinct gene family involved in mammalian reproduction or embryonic development using publicly available transcriptomic datasets and the Behura Lab computational pipeline. Students will generate standardized outputs, including ortholog and paralog annotations, gene expression matrices, phylogenetic relationships, expression similarity estimates, statistical results, and graphical summaries. These outputs will be aggregated across student groups to identify broader patterns of expression conservation and divergence across gene families, mammalian species, and developmental stages.

**Data quality and reproducibility** will be ensured through standardized computational workflows, consistent gene identifiers, documented data sources, and systematic quality-control procedures. Importantly, students will independently obtain orthology and paralogy annotations from Ensembl and NCBI and compare their predictions. Gene relationships supported by both resources and meeting predefined high-confidence criteria will be prioritized for the main comparative analyses. Students will repeat selected analyses using the refined annotations to determine how uncertainty in gene homology predictions influences their conclusions about the ortholog conjecture. Additional quality-control measures will include checking expression data completeness, evaluating developmental-stage comparability, reviewing statistical assumptions, and documenting computational parameters and software versions.

Student research progress will be shared through regular group discussions, interim presentations, and a final research symposium. Each group will submit its datasets, documented computational scripts, figures, and research report. These materials will be organized into a cumulative research collection that can be used by subsequent offerings of the CURE to extend previous analyses, investigate additional gene families, and test new evolutionary hypotheses. Reproducible computational workflows and appropriate nonrestricted research outputs may also be shared through GitHub, institutional research presentations, and scientific conferences.

The aggregated findings will be evaluated for their potential contribution to ongoing research in evolutionary developmental biology and functional genomics. Results that provide sufficiently novel and scientifically rigorous insights may be developed into conference presentations or peer-reviewed publications. Students who make substantial intellectual and research contributions will be appropriately recognized as coauthors in accordance with established scientific authorship criteria. Other contributions will be acknowledged as appropriate. Authorship expectations, research documentation, and responsible data-sharing practices will be discussed with students at the beginning of the CURE.

This approach will allow individual student projects to contribute to a larger, cumulative research effort while providing students with authentic experience in collaborative scientific discovery, reproducible computational research, and responsible communication of research findings.

Resources