BIOLOGY · GENETICS · GENOMICS

Read the sequence.
Keep the inference layers separate.

Move from sequence classes to reads, coverage, assembly, annotation, and functional comparison without treating genome size, similarity, or bulk abundance as proof.

Start the lessons ↓Go to practice →
3guided lessons
12practice questions
5choices per item
$0free, always

The Genomics reasoning loop

Use one class-stage-evidence ledger.

  1. 01Class

    Identify coding, intronic, regulatory, repeated, structural, or organelle sequence.

  2. 02Stage

    Distinguish raw reads, quality control, alignment, assembly, contigs, and annotation.

  3. 03Localize

    Read coverage at the relevant position and test whether repeats make placement ambiguous.

  4. 04Layer

    Name whether the assay measures DNA, RNA, protein, or a mixed community.

  5. 05Bound

    Keep similarity, abundance, composition, and perturbation conclusions within their evidence.

Genomics instruction is cross-checked against OpenStax Biology 2e · Whole-Genome Sequencing ↗.

Three linked lessons

From genome organization to bounded function.

A read is not an assembly, an assembly is not an annotation, and an annotation is not automatic functional proof. Track which inference was actually made.

01

LESSON 1 · 21 MIN

Study + retrieve

Genome organization and sequence classes

Distinguish coding, regulatory, intronic, repetitive, structural, and organelle sequence while treating function as an evidence question rather than a coding-status label.

ESSENTIAL QUESTIONWhat sequence class is described, where is it located, and what evidence supports a functional claim?
Genome sequence-class and function-evidence ledgerA whole-genome bar is partitioned into protein-coding exons, introns, regulatory DNA, repeated sequence, structural regions, and other noncoding sequence. A comparison shows species A with a genome twice the size of species B but a similar protein-coding gene count, and attributes the size difference to noncoding or repeated content rather than automatic complexity. A transcript panel shows an intron inside pre-mRNA and removed from mature RNA. A function-evidence ladder lists conservation, expression, binding, and controlled perturbation as different evidence types. A separate organelle card distinguishes mitochondrial or chloroplast DNA from nuclear chromosomes. A footer states GENOME SIZE IS NOT PROTEIN-CODING GENE COUNT. Every class and claim is written in text.A GENOME IS MORE THAN PROTEIN-CODING EXONSCODINGEXONSINTRONSREGULATORYREPEATSSTRUCTURALOTHERSPECIES A · 2× GENOME SIZEsimilar protein-coding gene countMORE NONCODING / REPEATED DNAPRE-mRNA → MATURE RNAexon 1 · [ intron ] · exon 2INTRON TRANSCRIBED, THEN REMOVEDFUNCTION IS AN EVIDENCE CLAIMCONSERVATIONEXPRESSIONBINDINGPERTURBATIONdifferent supportUNKNOWN FUNCTION ≠ PROVEN USELESS · CODING ≠ AUTOMATICALLY ESSENTIALGENOME SIZE ≠ PROTEIN-CODING GENE COUNT · SEQUENCE CLASSES HAVE DIFFERENT ROLESSTUDY DIAGRAM · TEXT DESCRIPTION AVAILABLE
01

Do not count genes from genome size

A genome includes protein-coding genes plus introns, regulatory regions, repeated sequences, structural regions, and other noncoding DNA. Species with larger genomes do not necessarily have more protein-coding genes because sequence-class proportions differ.

  • Genome size includes coding and noncoding DNA
  • Gene count is not genome length
  • Repeat content can expand genome size
02

Keep sequence roles distinct

Exons retained in a mature coding RNA can contribute protein-coding sequence, whereas introns are transcribed into pre-mRNA and then removed in typical splicing. Enhancers and other regulatory DNA can affect expression without encoding protein. Some noncoding regions produce functional RNAs or contribute chromosome structure.

  • Introns can be transcribed then removed
  • Regulatory DNA can act without translation
  • Noncoding does not mean automatically useless
03

Require evidence for function

Conservation, expression, biochemical binding, or a perturbation can support different functional hypotheses. Coding status alone does not prove importance, and lack of a known function does not prove that a sequence is useless. Nuclear and organelle genomes are physically distinct information systems with different inheritance contexts.

  • Known function requires evidence
  • Unknown is not useless
  • Organelle genome ≠ nuclear chromosome

Worked example

Species A has twice the genome size of species B but similar protein-coding gene counts. What could explain the difference?

  1. 1

    Genome size counts all DNA, not only coding exons.

  2. 2

    Species A may contain more intronic, repeated, or other noncoding sequence.

  3. 3

    The observation does not require twice as many proteins or greater organismal complexity.

ConclusionDifferent noncoding and repeated-sequence content can change genome size without a proportional change in protein-coding gene count.

Close the notes first

Retrieve the evidence boundary.

01Why is genome size not a direct gene count?
Genomes include extensive noncoding, intronic, regulatory, and repeated sequence.

Protein-coding genes are only one part of total DNA.

02Are introns absent from the primary transcript?
No; typical introns are transcribed into pre-mRNA and then removed during splicing.

Their removal is a post-transcriptional processing step.

03Does noncoding mean nonfunctional?
No; some noncoding sequences regulate, encode RNAs, or contribute structure, while other functions may be unknown.

Function is an evidence-based claim.

02

LESSON 2 · 23 MIN

Study + retrieve

Reads, assembly, alignment, and annotation

Trace raw reads through quality control, reference alignment or de novo assembly, contigs, coverage, and annotation while locating uncertainty at each stage.

ESSENTIAL QUESTIONIs this observation a raw sequence read, an assembled interval, an alignment to a reference, or a proposed biological annotation?
Genome sequencing inference pipelineA left-to-right pipeline labels sampled DNA fragments, raw reads, base-quality control, either reference alignment or overlap-based assembly, contigs, and biological annotation. A coverage strip shows regions with forty reads, twenty reads, and zero reads despite a high average, emphasizing local gaps. A repeat panel shows a fifty-base repeat-only read matching five locations and a longer read reaching unique flank sequence that can anchor placement. An annotation card combines sequence pattern, RNA evidence, homology, and perturbation evidence and labels the proposed feature revisable. A footer states READS ARE NOT A FINISHED ANNOTATED GENOME. Text labels carry all distinctions.SEPARATE EACH INFERENCE STAGEFRAGMENTSREADSQUALITYALIGN / ASSEMBLEANNOTATEAVERAGE COVERAGE CAN HIDE LOCAL GAPSregion 140×region 220×region 30× GAP30× AVERAGE DOES NOT MEAN EVERY BASE HAS 30 READSREPEATS CAN BE READ BUT NOT PLACED UNIQUELY50-BASE REPEAT-ONLY READmatches location 1 · 2 · 3 · 4 · 5LONGER READ + UNIQUE FLANKANCHORS ONE LOCATIONREADS ≠ FINISHED ANNOTATED GENOME · ANNOTATION REMAINS A REVISABLE HYPOTHESISSTUDY DIAGRAM · TEXT DESCRIPTION AVAILABLE
01

Treat reads as fragments

Sequencing produces reads from sampled DNA fragments. Quality control identifies unreliable bases or reads. Overlap among reads can support assembly into longer contigs, while reference alignment places reads against an existing sequence. Alignment and de novo assembly answer related but distinct questions.

  • Read = sampled fragment
  • Overlap can support a contig
  • Reference alignment uses an existing coordinate system
02

Read coverage locally

Coverage describes how many reads support positions or regions, often summarized as an average. Coverage can be uneven, so a high average does not guarantee every base is observed. Repeats longer than or too similar for the available reads can create multiple plausible placements and assembly gaps.

  • Average depth can hide gaps
  • Repeats create ambiguity
  • Longer context can resolve some placements
03

Keep annotation provisional

Annotation proposes genes, regulatory features, repeats, and other elements using sequence patterns, RNA or protein evidence, homology, and experiments. It is a layer of biological inference added after sequence processing, not a perfect label returned automatically by the sequencing instrument.

  • Sequence first, labels later
  • Multiple evidence types improve annotation
  • Annotation can be revised

Worked example

A 100-base repeat occurs in five locations, but all reads are 50 bases long and contain only repeat sequence. Why is placement ambiguous?

  1. 1

    Each repeat-only read matches several genomic locations equally well.

  2. 2

    The reads lack unique flanking sequence that would anchor a location.

  3. 3

    A longer read or paired context reaching unique sequence could resolve some placements.

ConclusionThe sequence can be read correctly yet still be impossible to place uniquely; base calling and mapping are different uncertainties.

Close the notes first

Retrieve the evidence boundary.

01What is a contig?
A longer continuous sequence inferred by assembling overlapping reads.

Reads are fragments; overlap supports their relative order.

02Does 30× average coverage mean every base has exactly 30 reads?
No; coverage can vary and some positions may have much less or none.

An average hides local variation.

03Is annotation the same as sequencing?
No; annotation interprets the processed sequence using computational and experimental evidence.

A sequence instrument does not automatically establish feature function.

03

LESSON 3 · 22 MIN

Study + retrieve

Comparative and functional genomic evidence

Use sequence, transcript, protein, or community data to generate bounded hypotheses while separating similarity, abundance, cell composition, and causal perturbation.

ESSENTIAL QUESTIONWhat molecular layer was measured, what comparison was made, and does the result show association or a context-bounded causal effect?
Comparative and functional genomics claim ladderA sequence-comparison panel shows high similarity leading to hypotheses of shared ancestry or related function, followed by an experiment requirement rather than an identical-function conclusion. A measurement-layer panel distinguishes transcriptomics as RNA, proteomics as protein, and metagenomics as community DNA. A bulk-tissue mixture shows equal per-cell expression but twice as many G-high cells producing a doubled bulk G signal. An evidence ladder progresses from association and abundance through cell-resolved measurement to controlled perturbation, with every conclusion bounded to the tested organism, tissue, cell type, and condition. A footer states SIMILARITY IS NOT IDENTICAL FUNCTION. Every inference is text-labeled without color dependence.NAME THE MEASUREMENT LAYERTRANSCRIPTOMICSRNA abundance / identityPROTEOMICSprotein abundance / stateMETAGENOMICScommunity DNA sampleBULK ABUNDANCE MIXES CELL TYPESBEFORE2 G-high cells + 8 G-low cellsBULK G = 1×AFTER4 G-high cells + 6 G-low cellsBULK G = 2×per-cell expression can remain unchanged while composition changesSIMILARITY → RELATED-FUNCTION HYPOTHESIS → CONTEXTUAL EXPERIMENTCONTROLLED PERTURBATION SUPPORTS A BOUNDED CAUSAL CLAIMSIMILARITY ≠ IDENTICAL FUNCTION · BULK CHANGE ≠ WITHIN-CELL REGULATIONSTUDY DIAGRAM · TEXT DESCRIPTION AVAILABLE
01

Use similarity as a hypothesis

Sequence similarity can support shared ancestry and a possible related function, especially when combined with conserved structure or experimental evidence. It does not guarantee identical function in every organism, tissue, or environment because regulation and cellular context can differ.

  • Similarity supports a candidate
  • Context can change function
  • Experimental evidence strengthens assignment
02

Name the measured layer

Transcriptomics measures RNA, proteomics measures proteins, and metagenomics samples genetic material from communities. High RNA abundance does not by itself establish high active protein, and a detected microbial sequence does not alone prove which organism is metabolically active.

  • RNA ≠ active protein
  • Community DNA ≠ activity by itself
  • Layer determines the claim
03

Watch mixtures and causality

Bulk samples average across cell types. A bulk abundance change can reflect regulation within cells, changed cell proportions, or both. Controlled perturbation can test necessity or sufficiency in a defined context, but one edited cell line does not prove a universal organism-level mechanism.

  • Bulk change can be composition
  • Association locates candidates
  • Perturbation claim stays context-bounded

Worked example

Bulk tissue RNA for gene G doubles after treatment, but the tissue also contains twice as many G-high cells. What can be concluded?

  1. 1

    The bulk assay averages RNA across the cell mixture.

  2. 2

    A larger proportion of G-high cells can increase the bulk signal without per-cell induction.

  3. 3

    Cell-type-resolved or controlled evidence is needed to separate composition from regulation.

ConclusionThe tissue-level RNA increase is real, but it does not by itself prove that treatment doubled G transcription within each cell.

Close the notes first

Retrieve the evidence boundary.

01Does the most similar sequence prove identical function?
No; it supports a functional hypothesis that still depends on context and evidence.

Related sequences can diverge in regulation or activity.

02What does transcriptomics directly measure?
RNA abundance or identity under the assay conditions.

Protein abundance and activity are downstream layers.

03Why can bulk data mislead?
A change can reflect altered cell-type composition rather than a within-cell regulatory change.

Bulk measurements average mixed populations.

Randomized retrieval set

Now locate the sequence class, processing stage, or evidence boundary.

Genome organization, repeats, coverage, assembly, annotation, transcriptomics, proteomics, metagenomics, bulk mixtures, and perturbation are interleaved.

12 PRACTICE QUESTIONS

Retrieve before you review.

Question order and all five answer options are shuffled when you begin. The correct answer stays attached to the same underlying choice.

Scope and score notice

Genomics foundations, not personal genome interpretation.

The ADA lists genomics within Genetics but does not publish a subtopic item quota. DAT TRAIN does not invent one.

Assembly algorithms, platform-specific workflows, personal genomic results, clinical variant classification, and population-genetic modeling remain outside this route unless a prompt supplies the model.

Use your results to choose what to review next—not as an official DAT score prediction.