BSC4434C: Introduction to Bioinformatics

Course Description

BSC4434C, Introduction to Bioinformatics, teaches biology students to work computationally with sequence and structural data. It is the course where a biology major stops treating a genome as a fact in a textbook and starts treating it as a dataset that can be queried, aligned, annotated and interpreted. The University of Florida's catalog describes it precisely: "Lecture and hands-on computer laboratories emphasize data-mining tools freely available in web-based resources that predict gene function from DNA, RNA, and protein sequences."

The C suffix is meaningful. The course is an integrated lecture-and-laboratory course, and the laboratory is a computer laboratory rather than a wet lab: students spend contact hours actually running BLAST searches, building alignments, constructing phylogenies and inspecting protein structures, not watching someone else do it. At the University of Florida the course carries 3 credits and meets twice weekly in extended sessions, consistent with the roughly 60 contact hours typical of a 3-credit integrated course in the Statewide Course Numbering System.

Course-level learning objectives published for the University of Florida section state the target skills directly: on completion students can "retrieve information on genes and proteins from biological and genomic databases," "predict genes from DNA sequences," "identify promoters and regulatory elements in DNA sequences," "analyze protein sequences," "compare protein and DNA sequences," "visualize and analyze protein structures," and "construct and interpret simple phylogenies."

The course appears at approximately 13 Florida institutions. It is a 4000-level course taken in the junior or senior year, usually after a molecular or cell biology foundation.

⚠ Suffix and emphasis vary considerably across Florida — this matters more here than for most courses. The C-suffix, laboratory-integrated version is the University of Florida's. Several institutions teach the subject under the bare number BSC4434, and the emphasis shifts noticeably with the department that owns it:

• University of Florida — BSC4434C, Introduction to Bioinformatics, 3 credits, taught out of Microbiology and Cell Science. Tool-centred and sequence-centred: web-based data-mining resources applied to gene function prediction.
• Florida A&M University — BSC4434, Intro to Bioinformatics. Framed as "an introduction to the theory and practice of bioinformatics and computational biology; emphasizing the use of computer databases to store, retrieve and assist in understanding biological information," covering sequence assembly and pattern recognition, protein modelling and alignment, genomics and proteomics, expression array analysis, and phylogenetics and systematics.
• Florida International University — BSC4434, Bioinformatics for Biologists — a title that signals the audience rather than a different subject.
• University of West Florida — BSC4434, Bioinformatics and Data Science, 3 credits, taught out of the Department of Public Health and offered concurrently with the graduate BSC5459. UWF's version leans toward data science: "data organization, representing and reasoning about sequence data, simple data mining strategies, and ethical protocols for data collection," applying "data science principles to biological and clinical problems." The clinical and ethical framing is genuinely different in emphasis from UF's molecular framing.

These are variants of one subject rather than different subjects, but a student choosing a section — or a transfer evaluator comparing two transcripts — should read the description rather than the number. The presence or absence of the C, and the owning department, both predict what the course will actually contain.

Learning Outcomes

Required Outcomes

• Retrieve and interpret information on genes, transcripts and proteins from the major biological databases, including GenBank, RefSeq, Ensembl, UniProt and the PDB.
• Explain how biological sequence data are represented, stored and versioned, and describe the standard file formats (FASTA, FASTQ, GenBank flat file, GFF/GTF, PDB, VCF).
• Perform and interpret sequence similarity searches using BLAST, selecting the appropriate variant (blastn, blastp, blastx, tblastn, PSI-BLAST) for a given question.
• Interpret alignment statistics correctly — bit score, raw score, E-value, percent identity, query coverage — and explain why a low E-value is not by itself evidence of function.
• Construct and evaluate pairwise and multiple sequence alignments, and explain the role of substitution matrices (BLOSUM, PAM) and gap penalties.
• Explain and apply the dynamic programming algorithms underlying alignment (Needleman-Wunsch for global, Smith-Waterman for local) at least conceptually, and explain why heuristic methods such as BLAST are used in practice.
• Distinguish homology, orthology and paralogy, and explain why inferring function from sequence similarity requires this distinction.
• Predict genes from genomic DNA sequence, identifying open reading frames, splice sites, and coding versus non-coding regions.
• Identify promoters, regulatory elements and sequence motifs, and interpret motif-discovery output.
• Analyse protein sequences: identify domains and families using Pfam and InterPro, and predict signal peptides, transmembrane regions and secondary structure.
• Visualise and analyse three-dimensional protein structures, relating structural features to sequence and to function.
• Construct, root and interpret a simple phylogenetic tree, distinguishing distance-based from character-based methods and interpreting bootstrap support.
• Explain the principles of high-throughput sequencing and describe at a conceptual level how raw reads become an assembled or aligned dataset.
• Apply functional annotation and enrichment analysis (Gene Ontology, KEGG) to a gene list and interpret the result critically.
• Document an analysis so that another person can reproduce it, recording tool versions, database versions, parameters and dates.
• Discuss the ethical handling of biological and clinical sequence data, including privacy, consent and the limits of genomic de-identification.

Optional Outcomes

• Write and adapt scripts in Python (with Biopython) or R (with Bioconductor) to automate a routine analysis — emphasised where the course is programming-oriented, omitted where it is web-tool-oriented.
• Work at the Unix command line and run command-line bioinformatics tools on a server or cluster.
• Analyse RNA-seq or microarray expression data through a differential-expression workflow.
• Perform variant calling and interpretation, including clinical variant classification against ClinVar and gnomAD.
• Carry out genome assembly and annotation for a small genome.
• Analyse metagenomic or microbiome data, including 16S rRNA community profiling.
• Apply structural bioinformatics beyond visualisation: homology modelling, molecular docking, and interpretation of AlphaFold predictions with their confidence metrics.
• Apply machine learning methods to biological classification and prediction problems.
• Build and query relational databases of biological data, and use public APIs (NCBI E-utilities, Ensembl REST) programmatically.
• Complete an independent project analysing a real dataset and present the findings.

Major Topics

Required Topics

• Introduction to bioinformatics and computational biology: what questions the field answers, and how it came to exist alongside sequencing technology
• Biological databases: GenBank/EMBL/DDBJ, RefSeq, Ensembl, UniProt/Swiss-Prot, PDB, and the distinction between primary, curated and derived databases
• Data organisation and file formats: FASTA, FASTQ, GenBank, GFF/GTF, SAM/BAM, VCF, PDB
• Database searching with BLAST: algorithm concept, variants, parameters, and the interpretation of output
• Alignment theory: scoring matrices, gap models, global and local alignment, dynamic programming
• Multiple sequence alignment: Clustal Omega, MUSCLE, MAFFT; alignment quality and manual curation
• Homology, orthology, paralogy, and the inference of function from sequence
• Gene prediction and genome annotation: open reading frames, codon usage, splice sites, ab initio versus evidence-based prediction
• Regulatory sequence analysis: promoters, transcription factor binding sites, motif discovery, genome browsers
• Protein sequence analysis: domains and families (Pfam, InterPro, PROSITE), physicochemical property prediction, signal peptides, transmembrane topology, secondary structure prediction
• Protein structure: the PDB, structure determination methods in outline, visualisation, and the sequence-structure-function relationship
• Molecular phylogenetics: distance and character-based methods, tree building, rooting, bootstrap support, and the difference between a gene tree and a species tree
• Genomics and high-throughput sequencing: platforms, read data, assembly and mapping at a conceptual level
• Functional genomics: expression analysis, Gene Ontology, pathway and enrichment analysis (KEGG, Reactome)
• Reproducibility and documentation: version tracking of tools and databases, parameter recording, and why a bioinformatics result is not reproducible without them
• Ethics and data governance: genomic privacy, informed consent, HIPAA where clinical data are involved, controlled-access repositories such as dbGaP

Optional Topics

• Programming for biologists: Python with Biopython, or R with Bioconductor
• Unix command line, shell scripting, and high-performance computing environments
• RNA-seq analysis and differential expression
• Variant calling, annotation and clinical interpretation (ACMG criteria, ClinVar, gnomAD)
• De novo genome assembly and annotation pipelines
• Metagenomics and microbiome analysis
• Structural bioinformatics: homology modelling, docking, molecular dynamics in outline, AlphaFold and its confidence metrics
• Machine learning in biological data analysis
• Systems biology and network analysis
• Single-cell sequencing analysis
• Workflow management and containerisation (Nextflow, Snakemake, Docker) for reproducible pipelines
• An independent capstone analysis project

Resources & Tools

• Bioinformatics and Functional Genomics (Pevsner) is the most widely adopted comprehensive text for a course at this level; Understanding Bioinformatics (Zvelebil & Baum) and Bioinformatics: Sequence and Genome Analysis (Mount) are common alternatives.
• Bioinformatics Data Skills (Buffalo) and Practical Computing for Biologists (Haddock & Dunn) are used where the course includes command-line and scripting work; Bioinformatics with Python Cookbook and the Biopython Tutorial support programming-oriented sections.
• Many sections use no textbook at all, relying instead on the documentation and tutorials published by the NCBI, EMBL-EBI and UniProt — all free, all authoritative, and all more current than any printed text.
• NCBI (ncbi.nlm.nih.gov): BLAST, GenBank, RefSeq, PubMed, Gene, dbSNP, SRA, and the E-utilities API. The NCBI Bookshelf's BLAST Handbook and the NCBI Insights tutorials are standard course reading.
• EMBL-EBI (ebi.ac.uk): Ensembl, UniProt, InterPro, Pfam, Clustal Omega, and the EBI's free online training courses, which are frequently assigned directly.
• Alignment and phylogenetics: Clustal Omega, MUSCLE, MAFFT, MEGA (free, and the most common teaching tool for phylogenetics), Jalview, IQ-TREE, RAxML.
• Structure: the RCSB Protein Data Bank, PyMOL (free educational licence), UCSF ChimeraX, Mol*, and the AlphaFold Protein Structure Database.
• Genome browsers: the UCSC Genome Browser, Ensembl, and IGV for read-level inspection.
• Computing environments: Galaxy (usegalaxy.org) is the standard choice for teaching pipeline work without requiring command-line skills; Jupyter notebooks, Google Colab and RStudio/Posit Cloud are used where the course involves scripting.
• Florida research infrastructure worth knowing: the University of Florida Interdisciplinary Center for Biotechnology Research (ICBR) and HiPerGator, one of the largest university supercomputers in the United States; the UF Genetics Institute; FSU's Center for Genomics and Personalized Medicine; and the Moffitt Cancer Center bioinformatics core in Tampa.

Career Pathways

• Bioinformatics Scientist and Computational Biologist — SOC 19-1029 (Biological Scientists, All Other) and 15-2051 (Data Scientists). Most independent positions require a master's or doctorate; this course is the undergraduate entry point.
• Bioinformatics Analyst / Research Technician — SOC 19-4021 (Biological Technicians) and 15-2051. Reachable at the bachelor's level, particularly in academic core facilities and sequencing labs.
• Clinical Laboratory Scientist with molecular specialisation — SOC 29-2011. Note that clinical laboratory personnel in Florida require state licensure through the Department of Health and normally a NAACLS-accredited programme; a bioinformatics course supports but does not substitute for that pathway. Molecular diagnostics certification is available through ASCP.
• Genetic Counselor — SOC 29-9092, requiring an ACGC-accredited master's programme and ABGC certification. Sequence interpretation literacy is directly relevant.
• Biostatistician and Health Data Analyst — SOC 15-2041 and 15-2051, typically with graduate study in biostatistics or public health.
• Pharmaceutical and Biotechnology R&D — target identification, genomics and computational chemistry roles; SOC 19-1042 (Medical Scientists) and 19-1029.
• Public Health Genomics and Epidemiology — SOC 19-1041. Pathogen genomic surveillance became a routine public health function after 2020 and is now a standing capability at state health departments, including Florida's.
• Agricultural and Environmental Genomics — SOC 19-1013 (Soil and Plant Scientists) and 19-1023 (Zoologists and Wildlife Biologists). Distinctively strong in Florida: UF/IFAS crop genomics, citrus greening research, aquaculture genetics, and coral and marine genomics work at Mote Marine Laboratory and Florida Atlantic University's Harbor Branch.
• Graduate study — the most common destination. Florida programmes include UF's bioinformatics and genetics graduate programmes, FSU's biomedical sciences, USF's bioinformatics and computational biology, and FIU's biomedical and biological sciences.
• Florida employers of note: Moffitt Cancer Center, UF Health and the UF Genetics Institute, Mayo Clinic Florida in Jacksonville, Nemours Children's Health, Sanford Burnham Prebys (Orlando), Max Planck Florida Institute for Neuroscience (Jupiter), the Herbert Wertheim UF Scripps Institute, Mote Marine Laboratory (Sarasota), and the Florida Department of Health Bureau of Public Health Laboratories.

Special Information

Position in the curriculum

BSC4434C is a junior- or senior-year elective in biology, microbiology, biotechnology, biochemistry and related majors. It follows the molecular foundation — at the University of Florida the prerequisite is MCB3020 or MCB3023 (microbiology) or BCH4024 or CHM3218 (biochemistry), each with a minimum grade of C. It is normally an elective rather than a required core course, though it is a required or strongly recommended course in biotechnology and molecular biology concentrations. It pairs naturally with genetics (PCB3063), molecular biology (PCB4522) and a statistics course.

Prerequisites narrative

Prerequisites vary by institution and by owning department, which is the practical consequence of the emphasis variation described above. A course taught out of microbiology will gate on microbiology or biochemistry; one taught out of public health or data science may gate on statistics or an introductory computing course instead, or on programme admission. Programming is generally not a prerequisite — most Florida sections are built around web-based tools precisely so that students without a computing background can take them — but sections that do include Python or R will say so, and students should check before enrolling if that distinction matters to them. Where the course is offered concurrently with a graduate section (UWF runs BSC4434 alongside BSC5459), undergraduates should expect graduate-level pace with a reduced assignment load.

Course format and workload

Three credits with an integrated computer laboratory, approximately 60 contact hours. At the University of Florida the course meets twice weekly in sessions of just under two hours, which is the practical signature of a C-suffix course: contact time is long enough for hands-on work in class. Laboratory work is done on computers, so a laptop is usually expected. Assessment typically combines problem sets built around real datasets, laboratory reports, and often a final independent analysis project. Students should expect the workload of a laboratory course rather than a lecture course, and the analyses accumulate — falling behind is difficult to recover from, because each week's tools build on the previous week's output.

Transfer and articulation

This is a 4000-level SCNS course, so the statewide guarantee behaves as it does for all upper-division work: the number is recognised, but applicability inside a major is the receiving department's decision, and the A.A. transfer guarantee does not cover it. Two specific cautions apply here. First, the C suffix and the credit count: a bare BSC4434 taken elsewhere may or may not be accepted where BSC4434C is required, depending on whether the receiving programme requires the laboratory contact. Second, the emphasis divergence: a course taught as sequence-based molecular bioinformatics and one taught as biological data science cover overlapping but not identical ground, and a programme with a specific downstream course in mind may care about which one a student took. Students transferring should carry a syllabus, not just a transcript line.

Course-code variations across Florida

Related content appears under several prefixes. BSC4434 / BSC4434C is the general biology-department course. PCB-prefix courses carry computational genomics and genome analysis at some institutions. BCH covers structural and computational biochemistry. CAP and CIS prefixes carry bioinformatics taught from the computer science side, which is a genuinely different course — algorithm-centred rather than biology-centred — and is usually not interchangeable with BSC4434C for a biology major's requirements. Graduate versions appear as BSC5459 and similar. Titles across the family include Introduction to Bioinformatics, Intro to Bioinformatics, Bioinformatics for Biologists and Bioinformatics and Data Science.

A note on tool and database currency

Bioinformatics coursework ages faster than most. Tool interfaces change, databases are reversioned, and a tutorial written three years ago may reference a page that no longer exists. This is a feature of the field rather than a defect of the course, and it is why documentation of tool and database versions is taught as a core skill: an analysis that does not record which build of which database it queried cannot be reproduced or defended later.

AI Integration

Bioinformatics is one of the few undergraduate biology courses where machine learning is not a commentary on the field but a substantial part of its current content. Treating AI here as an academic-integrity footnote would misrepresent the discipline.

AI as subject matter. AlphaFold changed structural biology inside a few years: predicted structures for essentially every known protein are now freely available from the AlphaFold Protein Structure Database, and a course that teaches protein structure without teaching how to obtain and evaluate a predicted structure is teaching an obsolete workflow. The critical skill is interpretive, not operational — reading the pLDDT per-residue confidence score and the PAE matrix, recognising that low-confidence regions frequently correspond to genuinely disordered regions rather than to prediction failure, and understanding that a predicted structure is a hypothesis rather than an experimental observation. Machine learning is likewise embedded throughout the standard toolkit: gene prediction, splice-site and promoter identification, signal peptide and transmembrane topology prediction, variant effect prediction, and protein function annotation are all now dominated by learned models. Students using these tools are already using AI whether or not the syllabus says so.

AI as a working tool. Large language models are genuinely effective at the tasks that block biology students in this course: explaining what an error message means, writing or fixing a short Python or R script, translating a described analysis into command-line syntax, and explaining an unfamiliar file format. For a student whose barrier is programming rather than biology, this is a real reduction in friction, and it is worth saying plainly rather than pretending otherwise.

Where AI fails. Language models fabricate gene names, accession numbers, and citations that look entirely correct. They will state a gene's function with confidence and no source. They generate code that runs and silently produces a wrong answer — the most dangerous failure mode in this field, because a bioinformatics pipeline rarely crashes when it is wrong, it simply outputs a plausible table. They do not know the current version of a database and will describe interfaces and parameters that have changed. And they cannot tell you whether a result is biologically meaningful: a significant enrichment score for a pathway that has nothing to do with your organism is exactly the kind of result a model will happily narrate.

The analyst's responsibility. Every accession must be resolved in the actual database. Every generated script must be understood well enough that you can say what each step does and why. Every result needs a sanity check against biological expectation — positive controls, known genes, a case where you already know the answer. Record tool and database versions, because a result you cannot reproduce is not a result. These are the same standards the field already applies to any pipeline; AI has not changed the standard, it has increased the volume of unverified output that has to meet it.

Ethics and data governance. The University of West Florida's version of this course names "ethical protocols for data collection" explicitly, and the point generalises. Genomic data are not anonymisable in the ordinary sense — a genome identifies its owner and partially identifies their relatives. Applying AI to clinical or human genomic data raises consent questions that go beyond the technical: what a participant agreed to when the sample was collected, whether controlled-access data may be submitted to a third-party service, and how model-derived predictions should be communicated when they carry clinical implications. Uploading identifiable or controlled-access data to a commercial AI service is a governance violation regardless of how useful the answer would be.

Academic integrity. Instructor policies vary, and in this course they often differ by assignment type — AI permitted for debugging code, prohibited for interpretation, is a common split, because interpretation is the skill being assessed. Read the syllabus and ask when it is not explicit.