1. Working with Agentic AI: Bioinformatics Scientist

How human skills and agentic skills are dynamically integrated and coordinated across bioinformatics responsibilities

1.1. Bioinformatics Scientist as an Occupation

A bioinformatics scientist develops and applies computational methods, algorithms, and software tools to analyze and interpret complex biological data. This occupation connects molecular biology, computer science, and statistics by transforming high-throughput sequence data, structural molecular information, and multi-omics datasets into biological insights, predictive models, and reproducible analytical pipelines. To ensure that this computational work accurately reflects underlying biological realities, the scientist must understand how molecular mechanisms, experimental designs, data structures, algorithmic limitations, and research objectives collectively determine the validity and usefulness of computational results.

Typical responsibilities of a bioinformatics scientist include:

• bioinformatics pipeline development and workflow automation;
• high-throughput biological data analysis and interpretation;
• bioinformatics algorithm and software tool development;
• biological database architecture and knowledge management;
• multi-omics integration and systems biology modeling; and
• computational research translation and cross-disciplinary collaboration.

These responsibilities are fulfilled through distinct tasks and subtasks, each requiring one or more actions with specific skill requirements. Human skills and agentic skills both contribute to these responsibilities, but their levels and forms of involvement differ according to the responsibility, task, action, data, biological context, technical environment, and expected outcome.

1.2. Human Skills and Agentic Skills in Bioinformatics

Bioinformatics increasingly integrates both human skills and agentic skills.

Human skills are reusable capabilities possessed and developed by bioinformatics scientists. They include molecular biology understanding, algorithmic reasoning, experimental design interpretation, structural biology intuition, statistical judgment, critical evaluation of biological significance, ontology design, and cross-disciplinary communication.

Agentic skills are reusable capabilities available to AI agents through their underlying models, contextual information, instructions, constraints, tools, and computational resources. In bioinformatics, these skills include code generation, sequence alignment automation, literature retrieval, API querying, pipeline scripting, structural prediction execution, pattern recognition in high-dimensional omics data, and documentation support.

Both skill types contribute to the same occupational responsibilities, but their involvement varies by action. Understanding the biological implications of a novel genetic variant relies primarily on human knowledge of molecular pathways and evolutionary conservation; an AI agent may then use agentic skills to retrieve structural predictions, query population frequency databases, and summarize relevant literature. In pipeline development, the scientist first determines the appropriate computational tools and biological filters, and an AI agent may subsequently generate the workflow scripts, resolve software dependencies, and automate execution. For multi-omics integration, an AI agent may perform the high-dimensional matrix factorization, while the scientist interprets the output to ensure it maps correctly to known metabolic or regulatory networks. Across these actions, human and agentic skills are activated sequentially or iteratively based on task requirements, with the human professional maintaining oversight and final judgment.

The relationship between these skills is not a fixed division where certain responsibilities permanently belong to humans and others to AI. Instead, human and agentic involvement shifts across tasks and between different actions within the same task.

1.3. Dynamic Human–Agentic Skill Integration and Collaboration in Bioinformatics

Bioinformatics functions can be understood as an occupation where responsibilities are fulfilled through tasks and subtasks, tasks and subtasks are executed through actions, and actions require appropriate configurations of human and agentic skills.

An occupation provides the overall professional context. For bioinformatics, this context encompasses developing computational tools, analyzing molecular data, managing biological databases, and translating computational findings into testable biological hypotheses.

A responsibility represents an expected area of professional work or an outcome that the bioinformatics scientist accomplishes or maintains. Responsibilities such as pipeline automation, sequence analysis, algorithm development, and multi-omics integration each contain multiple tasks.

A task is a context-dependent unit of work used to fulfill part or all of a responsibility. When a task is too broad or complex for reliable execution, it may be decomposed into smaller subtasks.

Tasks and subtasks are executed through one or more actions. Actions represent reusable and configurable operations such as aligning a DNA sequence, querying a protein database, writing a parsing script, predicting a molecular structure, integrating transcriptomic and proteomic datasets, or generating a phylogenetic tree.

Each action has skill requirements. These requirements describe the capabilities needed for effective and reliable execution. Depending on the action and its context, human skills, agentic skills, or an integration of both may satisfy these requirements.

The skills available for bioinformatics reside in a skill space containing both human skills and agentic skills. Relevant skills in this space can serve as candidate skills for executing the actions required by a particular task. A human skill in structural biology intuition, for example, may serve as a candidate for evaluating the physical plausibility of a predicted protein fold, while an agentic skill in structural prediction execution may serve as a candidate for generating the initial 3D coordinates.

Candidate skills are evaluated relative to the requirements of the action and its occupational and task context. Appropriate skills are then selected and organized into a skill configuration for execution. Depending on these requirements, a configuration may rely mainly on human skills, mainly on agentic skills, or on an integrated arrangement of both.

1.3.1. Skills Competition and Skills Collaboration

Two important relationships can occur between human skills and agentic skills when constructing a skill configuration: skills competition and skills collaboration.

Skills competition occurs when both a human skill and an agentic skill can satisfy the same or similar requirements for an action, but simultaneous application of both is unnecessary. Under a particular occupational and task context, one skill type may provide a better fit for the action than the other.

For example, both a bioinformatics scientist and an AI agent can generate a standard summary of quality control metrics for a new batch of RNA sequencing data. When the sequencing platform, quality thresholds, and reporting templates are clearly specified, the agentic skill provides fast and scalable execution. When the summary requires interpretation of unusual adapter contamination patterns, ambiguous library preparation artifacts, or platform-specific systematic errors, the scientist’s human skills provide the necessary contextual understanding.

The purpose of skills competition is not to establish whether human or agentic skills are generally superior. Rather, it identifies which available skill provides the better fit for a particular action under particular conditions.

Skills collaboration occurs when one skill type alone cannot effectively satisfy the requirements of a task or its actions, or when combining complementary human and agentic skills produces stronger task performance.

Collaboration does not require human and agentic skills to operate simultaneously. Instead, they may be activated at different stages of action execution in a sequential, iterative, or recursive manner according to evolving task requirements.

For example, a scientist first identifies an unexpected cluster of gene expression changes in a transcriptomic dataset and determines which biological pathways require investigation. An AI agent then retrieves relevant gene ontology annotations, compares the expression profiles against public reference databases, and identifies initial regulatory network patterns. The scientist subsequently interprets these patterns within the specific experimental context and formulates possible molecular mechanisms. Based on the scientist’s direction, the agent next performs additional motif enrichment analysis to examine those mechanisms. Finally, the scientist evaluates the resulting evidence and determines which findings are sufficiently supported to propose for wet-lab validation. Throughout this sequence, human and agentic skills are activated iteratively as task requirements evolve, while the scientist directs the investigation and evaluates its outcomes.

This dynamic relationship allows bioinformatics scientists to leverage the complementary strengths of both skill types rather than relying on a permanent allocation of work.

1.3.2. Skill Configuration and Task Execution

A skill configuration specifies which human skills and agentic skills are used to execute the actions associated with a task or subtask and how those skills are coordinated.

The configuration depends on factors such as action requirements, data volume, biological complexity, algorithmic constraints, expected reliability, available computational resources, information standardization levels, research consequences, and intermediate results.

For one action, an agentic skill may independently satisfy the requirements. For another, a human skill may provide the better fit. A complex task may require several configurations across multiple actions, with human and agentic skills repeatedly interacting as the task progresses.

The configuration is therefore dynamic rather than permanent. It may change when the action changes, new biological annotations become available, an agentic result fails to satisfy expected computational constraints, or human interpretation identifies a new biological anomaly.

Throughout this process, the bioinformatics scientist remains the professional responsible for directing and controlling the work associated with the occupation and its responsibilities. Agentic AI provides powerful complementary capabilities that expand what the professional can align, predict, calculate, integrate, visualize, summarize, and automate.

1.3.3. Task Performance Evaluation and Skill Space Adaptation

The performance produced through a selected skill configuration is evaluated against predefined evaluation metrics. These metrics determine whether the human skills, agentic skills, or their integrated configuration are sufficiently capable of performing the relevant task and its required actions.

Evaluation criteria depend on the work being performed. In bioinformatics, they may include algorithmic accuracy, biological validity, computational efficiency, reproducibility, pipeline stability, data integration coherence, structural plausibility, and other task-specific expectations.

When a selected skill or skill configuration produces performance above the defined threshold, it passes the performance evaluation and is retained for future applications under appropriate conditions. Repeated successful use provides additional evidence about where that skill or configuration performs effectively.

When performance falls below the required threshold, the corresponding skill or configuration is not reused without modification. The relevant human or agentic skills may be refined or reconfigured, existing skills can be combined differently, or new human and agentic skills may be developed or constructed from scratch when the existing skill space lacks adequate capability.

For example, an AI agent first generates a script to parse and annotate a novel metagenomic dataset, but repeatedly misclassifies sequences from a recently discovered microbial clade due to outdated reference databases. Task performance evaluation identifies this biological misclassification as falling below the required threshold. The agentic skill is then refined through updated database retrieval instructions, custom taxonomic mapping logic, or integration with specialized biological APIs. The revised skill is evaluated again, and once it satisfies the required performance criteria, it is retained for appropriate future metagenomic tasks. Human skills follow a corresponding development process: when a scientist lacks the mathematical background required to evaluate a newly introduced tensor decomposition method for multi-omics data, the relevant capability may be developed through learning, collaboration with computational mathematicians, and practical experimentation before its performance is evaluated through subsequent work.

Refined and newly constructed skills that satisfy the required performance criteria are added to the skill space. In this way, task performance evaluation provides feedback not only for completing current work but also for continuously optimizing and expanding the available skill space.

The skill space is not constructed as an isolated collection of occupation-specific representations. Many human and agentic skills used in bioinformatics are reusable across occupations. Code generation, statistical reasoning, database querying, literature retrieval, pattern recognition, visualization, and documentation contribute to work in computational chemistry, systems biology, epidemiology, data science, and many other fields.

Reusable skills can therefore be represented once and configured differently according to the responsibilities, tasks, actions, and contexts of different occupations. This allows the skill space to serve a growing range of occupations without unnecessarily duplicating or exploding the number of skill representations. Occupational specialization is introduced through task context, action requirements, parameters, biological data sources, tools, constraints, and skill configurations rather than by creating a completely separate skill representation whenever the same underlying capability appears in another occupation.

The resulting process is adaptive:
occupational responsibility → task and subtask → actions → skill requirements → candidate human and agentic skills → skill configuration → task execution → task performance evaluation → skill refinement, reconfiguration, or construction → optimized and expanded skill space

This cycle allows human and agentic capabilities to evolve alongside the requirements of bioinformatics work.

1.4. Essential Responsibilities of Bioinformatics Scientist

1.4.1. Bioinformatics Pipeline Development and Workflow Automation

Bioinformatics pipeline development and workflow automation establishes the reproducible computational infrastructure required to process raw biological data into analyzable formats. This responsibility includes designing data processing workflows, selecting appropriate computational tools, managing software dependencies, automating execution across high-performance computing clusters or cloud environments, and ensuring pipeline scalability for large-scale sequencing or mass spectrometry datasets.

Agentic skills contribute extensively to workflow scripting, dependency resolution, boilerplate code generation, containerization support, and automated testing of pipeline logic. Human skills remain essential for defining the biological constraints of the workflow, selecting appropriate algorithmic parameters based on experimental design, resolving edge cases in biological data formatting, and deciding whether a proposed pipeline architecture appropriately balances computational efficiency with analytical sensitivity.

For example, a research team first acquires a massive dataset of long-read genomic sequences requiring specialized error correction. The bioinformatics scientist clarifies the biological requirements for variant detection and defines the acceptable computational resource limits. An AI agent then generates draft workflow scripts, proposes containerized environments for the required alignment tools, and writes automated testing routines. The scientist reviews these proposals and identifies that the default memory allocation in the proposed script will fail during the assembly of highly repetitive genomic regions. After the scientist revises the resource constraints, the agent updates the configuration and executes validation tests on a subset of the data. Finally, the scientist verifies that the automated pipeline accurately reflects the approved analytical strategy.

1.4.2. High-Throughput Biological Data Analysis and Interpretation

High-throughput biological data analysis and interpretation transforms processed molecular data into evidence about gene expression, genetic variation, epigenetic modifications, or microbial community composition. This responsibility includes executing analytical methods, evaluating statistical significance, filtering technical artifacts, identifying biologically relevant signals, and formulating hypotheses based on molecular evidence.

Agentic skills are well suited to executing standard analytical commands, performing large-scale statistical calculations, automating quality control filtering, querying public reference databases, and generating initial visualizations of high-dimensional data. Human skills contribute to determining the biological meaning of statistical outliers, distinguishing true molecular signals from platform-specific sequencing artifacts, interpreting results within the context of cellular mechanisms, and deciding which findings warrant deeper functional investigation.

For example, an AI agent continuously processes incoming RNA-sequencing data and flags a series of genes showing extreme differential expression under a specific environmental stressor. The agent calculates the statistical significance and generates standard volcano plots. The bioinformatics scientist reviews the flagged genes and recognizes that several of them share a novel, unannotated splice variant that the standard alignment algorithm may have misquantified. The scientist instructs the agent to re-run the quantification using a specialized transcriptome assembly parameter and extracts the specific sequence reads for manual inspection. Human and agentic skills interact iteratively as the process moves from broad statistical detection to nuanced molecular investigation.

1.4.3. Bioinformatics Algorithm and Software Tool Development

Bioinformatics algorithm and software tool development creates novel computational methods to address unresolved challenges in sequence analysis, structural prediction, phylogenetic reconstruction, or network biology. This responsibility includes designing mathematical models, writing optimized code, benchmarking new methods against existing standards, and publishing software tools for the broader scientific community.

Agentic skills assist with code scaffolding, mathematical optimization support, unit test generation, benchmarking automation, and documentation drafting. Human skills remain important for conceptualizing the underlying biological or evolutionary principles, designing the mathematical logic of the algorithm, evaluating the computational complexity, and ensuring that the new tool solves a genuine biological problem rather than a purely theoretical one.

For example, an organization requires a new heuristic method to rapidly bin metagenomic sequences from highly complex environmental samples. The scientist defines the evolutionary constraints and designs the core mathematical logic for sequence clustering. An AI agent then generates the initial codebase, writes unit tests to verify edge cases, and automates the benchmarking process against standard reference datasets. The scientist reviews the benchmarking results and identifies that the algorithm struggles with sequences exhibiting high horizontal gene transfer rates. The scientist revises the mathematical weighting, and the agent updates the code and re-runs the benchmarks. Finally, the scientist validates the tool’s biological accuracy and authorizes its release.

1.4.4. Biological Database Architecture and Knowledge Management

Biological database architecture and knowledge management structures, curates, and maintains the digital repositories that store genomic, proteomic, and molecular interaction data. This responsibility includes designing biological ontologies, integrating disparate public and private datasets, managing data schemas, ensuring nomenclature standardization, and optimizing database queries for complex biological searches.

Agentic skills contribute to schema generation, API data harvesting, automated entity resolution, ontology mapping, and continuous data synchronization. Human skills define the biological relationships within the ontology, resolve conflicting gene or protein nomenclatures across different species, govern data curation rules, and ensure that the database structure accurately reflects current molecular biology paradigms.

For example, a research consortium first aggregates multi-species protein interaction data from various localized databases. The scientist defines the overarching biological ontology and identifies high-risk interaction types requiring strict manual curation. An AI agent then processes millions of records, mapping local protein identifiers to standardized accession identifiers and extracting interaction evidence from unstructured literature abstracts. Because abstract-level evidence may be incomplete, the scientist reviews the agent’s confidence scores for complex, multi-protein complexes and manually resolves conflicting biological evidence. The agent then applies these human-verified rules to the remaining dataset and generates a curation validation report. Finally, the scientist approves the unified database for network analysis.

1.4.5. Multi-Omics Integration and Systems Biology Modeling

Multi-omics integration and systems biology modeling combines diverse molecular data types—such as genomics, transcriptomics, proteomics, and metabolomics—to construct comprehensive models of biological systems. This responsibility includes aligning heterogeneous datasets, performing multi-block statistical analyses, mapping molecular interactions, identifying regulatory hubs, and simulating system-level responses to perturbations.

Agentic skills retrieve relevant pathway databases, perform high-dimensional matrix factorization, generate network visualizations, and automate the mapping of metabolites to enzymatic reactions. Human skills evaluate the systems-level biological coherence, distinguish correlative statistical links from true causal regulatory mechanisms, interpret the physiological consequences of network perturbations, and determine which integrated models are biologically plausible.

For example, a research team seeks to understand the systemic metabolic shifts occurring during cellular differentiation. The scientist defines the integration strategy and specifies the required temporal alignment between the transcriptomic and metabolomic datasets. An AI agent then executes the multi-omics factorization algorithms, identifies correlated gene-metabolite modules, and maps them onto known metabolic pathways. The scientist reviews the resulting network models and identifies that a key regulatory hub proposed by the algorithm conflicts with established biological or biochemical constraints of the cell. The scientist provides corrective biological constraints, and the agent re-runs the integration under the revised parameters. Finally, the scientist evaluates the refined systems model for physiological validity.

1.4.6. Computational Research Translation and Cross-Disciplinary Collaboration

Computational research translation and cross-disciplinary collaboration converts complex computational findings into actionable biological hypotheses, experimental designs, and accessible knowledge for wet-lab biologists, clinicians, or agricultural researchers. This responsibility involves synthesizing computational evidence, designing validation experiments, communicating methodological limitations, and bridging the vocabulary gap between computational and experimental sciences.

Agentic skills summarize vast amounts of literature, generate interactive molecular visualizations, draft methodology sections, and organize evidence for experimental planning. Human skills evaluate the translatability of computational predictions to physical experiments, design the actual wet-lab validation protocols, communicate uncertainty to non-computational stakeholders, and determine which computational insights are robust enough to drive expensive laboratory work.

For example, a computational analysis predicts a novel set of targetable protein interactions in a specific cellular pathway. The scientist defines the biological context and the experimental constraints required for physical validation. An AI agent then organizes the structural evidence, summarizes existing literature regarding the target proteins, and generates 3D interaction visualizations. The scientist reviews the compiled evidence and identifies that one predicted interaction relies on a transient conformational state that may be difficult to capture with standard crystallography. The scientist revises the proposed validation strategy to suggest an alternative structural or biophysical approach, such as cryo-electron microscopy, and the agent updates the experimental protocol documentation. Finally, the scientist presents the refined, computationally supported hypotheses to the experimental biology team.

1.5. Changing Human and Agentic Involvement Across Bioinformatics Responsibilities

Human skills and agentic skills are involved throughout bioinformatics, but their relative involvement differs across responsibilities and changes during task execution.

These involvement levels are not permanent allocations. A single responsibility may contain actions where human and agentic skills compete as alternative capabilities and other actions where they collaborate sequentially, iteratively, or recursively.

The practical objective is to construct skill configurations that use the most suitable available capabilities for each part of the work and revise those configurations when task performance evaluation shows that another arrangement performs more effectively or reliably.

1.6. Working as a Bioinformatics Scientist with Agentic AI

Agentic AI is becoming an increasingly powerful component of bioinformatics. Bioinformatics scientists should therefore adopt and integrate AI agents as complementary participants in professional workflows rather than treating them only as occasional tools for generating scripts, summarizing literature, or formatting data tables.

This shift does not reduce the importance of molecular biology and computational expertise. Human professionals remain responsible for the occupation and its corresponding responsibilities. They understand biological mechanisms, define meaningful molecular questions, interpret experimental contexts, determine appropriate algorithmic methods, evaluate structural and statistical validity, communicate with experimental biologists, control consequential analytical decisions, evaluate task performance, and decide how human and agentic skills are configured and improved.

At the same time, agentic skills greatly expand the volume and variety of biological information that can be processed. AI agents retrieve, organize, and analyze large volumes of sequence data, generate and test pipeline scripts, map heterogeneous omics datasets to standardized representations, calculate complex statistical metrics, predict molecular structures, draft documentation, and synthesize extensive molecular evidence at a scale difficult to achieve through direct human effort alone.

Strong bioinformatics practice is therefore not based on maximizing human work or agentic work independently. It is based on finding appropriate configurations of human skills and agentic skills for the responsibilities, tasks, subtasks, and actions being performed.

Where a human skill and an agentic skill can independently perform the same action, skills competition identifies which is better suited to the current context. Where work requires complementary capabilities, skills collaboration results in human and agentic skills being activated sequentially, iteratively, or recursively to achieve stronger task performance than either skill type achieves alone.

Task performance evaluation then determines whether selected skills and configurations satisfy predefined expectations. Successful skills are retained and reused, while inadequate skills are refined, reconfigured, or newly developed. The resulting capabilities return to an increasingly useful skill space that supports bioinformatics and, where capabilities are reusable, other occupations as well.

For a bioinformatics scientist, working effectively with agentic AI means treating AI as a powerful complementary capability integrated into everyday professional routines while maintaining human control of occupational responsibilities. Through appropriate skill selection, competition, collaboration, evaluation, and continuous improvement, human professionals and agentic AI work together to achieve bioinformatics research that is more efficient, scalable, adaptable, timely, reproducible, and biologically meaningful.