eduKateSG · Why Science?
Give every quantitative value an explicit genomic interval—then protect coordinate rules, exact values and missing regions
Choose the closest route, then return to the full index whenever you need the wider evidence chain.
Science learning becomes powerful when students can tell a measured pattern from a model, a statistical decision from a mechanism, and a useful lead from a finished conclusion. bedGraph is a line-oriented UCSC track format for continuous-valued genomic data. After an optional browser-specific track line, each data row contains chrom, chromStart, chromEnd and dataValue. UCSC specifies zero-based half-open coordinates: chromStart is included and chromEnd is excluded. Values may be positive or negative integers or reals, rows must be in numerical order, and only specified positions are graphed. Unlike uploaded Wiggle data, bedGraph export preserves original values; for very large data UCSC recommends conversion to indexed binary bigWig. A valid interval-value table still needs assembly, units, sample provenance and experimental limits. This guide supports Primary Science, PSLE Science, Secondary Science, O-Level Science and STEM education as a longform evidence-reading route; it is not medical advice, a laboratory protocol, an admission promise or a career guarantee.
Related eduKate reading: BED and genomic feature coordinates; tabix and indexed genomic text files; GFF3 genome feature annotations; Science Learning Hub; Education Hub; STEM Education; Careers by Subject Capability.
Primary and official sources checked on 12 October 2026: UCSC Genome Browser bedGraph Track Format, checked 12 October 2026; UCSC Genome Browser Wiggle Track ASCII Text Format, checked 12 October 2026; UCSC Genome Browser bigWig Track Format, checked 12 October 2026; 2026 Singapore–Cambridge O-Level Biology syllabus; MOE G2/G3 Lower Secondary Science teaching and learning syllabus; 2026 MOE G2 Computing syllabus. Specifications, tools and recommendations can change, so preserve the dated documentation, exact source data, coordinate rules and software versions used.
A deeper reading route
bedGraph makes interval coverage explicit. Every row says which chromosome, which half-open interval and which value belong together. That clarity supports variable-width regions and sparse data, but it also exposes every coordinate mistake directly. One wrong end can create an overlap, a gap or a value assigned to an unintended base.
The four-column contract is small enough to teach and strict enough to matter. chrom identifies the reference sequence; chromStart and chromEnd define an interval; dataValue gives the quantitative signal. Extra browser control lines are useful inside a UCSC custom track but can make a raw file fail other validators, so presentation and data should be kept distinguishable.
Zero-based half-open intervals make length equal to chromEnd minus chromStart and allow adjacent intervals to meet cleanly. The first base occupies [0,1). This differs from one-based inclusive systems such as GFF3 and from the positional declarations of Wiggle. Coordinate conversion must be named, tested and reversible.
Only specified positions are graphed. An uncovered region can mean not measured, filtered, below a reporting threshold or genuinely absent under a declared rule. It does not automatically mean zero. A pipeline should report how gaps were created and whether downstream summaries ignore, impute or treat them as zeros.
bedGraph preserves exact exported values in the UCSC workflow described by its documentation, which distinguishes it from uploaded Wiggle custom tracks. Exact storage is still not exact science: rounding may already have occurred upstream, and a value may be a windowed summary rather than a base-level observation. Units and generation method remain essential.
Rows need numerical order for display. Production conversion to bigWig additionally depends on chromosome sizes and compatible sorting. Lexicographic chromosome sorting can place chr10 before chr2, while another tool expects the order of chrom.sizes. Use the converter contract and validate the complete contig set.
Overlaps require a stated meaning. Two rows can assign different values to overlapping bases, and generic parsers may keep both, choose one or refuse the file. Even when a tool accepts overlaps, a scientific workflow should define whether they are impossible, additive, alternative measurements or an error.
Converting a large bedGraph to bigWig enables indexed regional transfer, but the binary output should be compared with the text source. Query the first and last intervals of multiple chromosomes, check value extrema and totals, and verify that gaps and negative values survive. Preserve source and output hashes with the exact chrom.sizes file.
A visual track is a view, not an audit. Autoscaling and window functions can make two tracks appear similar or different depending on the region. Claim-making should rely on recorded values, comparable scales and statistical analysis tied to independent samples, not merely on a striking peak.
For students, bedGraph is an elegant lesson in functions defined over intervals. It connects number lines, graphs and measurement to biology while making boundaries visible. The cheerful takeaway is practical: if you can state where a value applies, what it measures and what is missing, you are already thinking like a careful data scientist.
Interval algebra becomes especially important when a workflow bins, merges or clips a bedGraph. Merging adjacent rows with equal values can reduce file size without changing the piecewise signal, but only if the rows touch exactly and share the same reference and meaning. Splitting an interval should preserve coverage and value, while averaging several intervals changes the data model. Record whether a transformation was lossless, value-preserving under a stated query, or a statistical summary with new resolution.
Comparison across samples requires denominator discipline. Coverage-like values may be raw counts, reads per million, fold enrichment, log ratios or model outputs. Even identical coordinates do not make unlike scales comparable. The manifest should state library size treatment, excluded regions, pseudocounts and any clipping. When tracks are displayed together, use either a justified shared scale or clearly labelled separate scales, and confirm visual impressions with an analysis that respects biological replicates.
A durable bedGraph archive benefits from simple invariants that anyone can recompute: row count, covered bases, number of gaps, number of overlaps, minimum and maximum coordinate per contig, value range and a checksum. Add several known interval queries that exercise contig starts, ends and gaps. These checks are inexpensive, catch truncation and sorting mistakes quickly, and make later conversion to bigWig auditable even when software versions or hosting locations change.
Clipping deserves special care. Restricting a track to a study region can shorten intervals at the boundary, discard rows outside the target and alter covered-base denominators. The subset should name the selection intervals, coordinate convention and treatment of partial overlaps. Retain the parent hash and before-after counts so a later reader can distinguish a legitimate regional product from accidental truncation.
Educational use becomes stronger when learners audit two representations of the same tiny signal. Give them a four-row bedGraph, ask them to draw the step function, then ask which bases are uncovered and which carry an explicit zero. Finally, change one end coordinate and let them observe the gap or overlap. The exercise makes data literacy concrete: one symbol can change the scientific object even when the plotted picture looks almost the same.
Publication should include a plain-language boundary statement. Say whether the track is exploratory, quality-control, descriptive or intended for formal testing; name the population and assay it represents; and state which regions were excluded. This prevents readers from extending a tidy interval-value file beyond the study that produced it. The happiest scientific outcome is not a dramatic graph but a result whose limits are easy to see, question and reproduce.
| Stage | Question | Evidence | Responsible output |
|---|---|---|---|
| Locate | Where does the value apply? | chrom, chromStart and chromEnd | An explicit half-open interval |
| Measure | What quantitative value is assigned? | Signed integer or real dataValue plus units | A bounded signal claim |
| Organise | How are rows ordered and gaps handled? | Sorting, overlap and missingness rules | A coherent signal table |
| Scale | How is a large track served and checked? | chrom.sizes, bigWig conversion and regional comparison | A reproducible scalable track |
Inside this guide
1–12 · Foundations and design
- 1. Start with the signal question
- 2. Recognise four data columns
- 3. Separate track metadata
- 4. Bind chrom to an assembly
- 5. Read chromStart as included
- 6. Read chromEnd as excluded
- 7. Use half-open adjacency
- 8. Store variable widths explicitly
- 9. Use numerical order
- 10. Define chromosome order
- 11. Keep integer and real values
- 12. Name units and baseline
13–24 · Models and estimation
- 13. Treat gaps as no data
- 14. Handle explicit zero separately
- 15. Detect overlaps
- 16. Detect duplicate intervals
- 17. Check bounds against contig length
- 18. Preserve exact source values
- 19. Keep upstream precision visible
- 20. Separate intervals from features
- 21. Use display options cautiously
- 22. Understand window aggregation
- 23. Convert large data to bigWig
- 24. Supply the correct chrom.sizes
25–36 · Diagnostics and interpretation
- 25. Remove browser lines before conversion
- 26. Compare source and bigWig
- 27. Protect sensitive signals
- 28. Connect to Primary Science
- 29. Connect to Secondary Science
- 30. Connect to mathematics
- 31. Connect to computing
- 32. Connect to pathways
- 33. Did you know?
- 34. Write a reader-ready signal note
- 35. Build a validation ladder
- 36. Finish with a reproducible bundle
Section 1 of 36
1. Start with the signal question
bedGraph can store coverage, scores or other continuous-valued data over genomic intervals. The file does not explain the assay by itself.
Working checkpoint — bedGraph: Name the measurement, unit, sample and aggregation method. Keep source hash, assembly, units, interval convention, sorting, gaps and transformation receipt together.
For a classroom model, assign rainfall values to invented time intervals.
Section 2 of 36
2. Recognise four data columns
The UCSC specification defines chrom, chromStart, chromEnd and dataValue for each data row.
Working checkpoint — chromStart: Reject missing, shifted or silently extra fields under the chosen parser. Read each value beside its half-open interval, sample, scale and generation method.
For future study, connect genomics to piecewise functions and computing while checking pathways separately.
Section 3 of 36
3. Separate track metadata
A UCSC track definition controls display, but other tools may require a raw four-column data file.
Working checkpoint — chromEnd: Keep presentation lines separate from portable data. Preserve omitted intervals, explicit zeros and rejected overlaps; they answer different questions.
For scientific writing, report intervals, units, gaps, overlaps and conversion checks.
Section 4 of 36
4. Bind chrom to an assembly
A chromosome label is meaningful only within a named reference sequence set.
Working checkpoint — dataValue: Store assembly, contig lengths and source beside the file. Separate exact file values, visual summaries, statistical results and biological interpretation.
For students, ask where each value begins and where it stops.
Section 5 of 36
5. Read chromStart as included
bedGraph starts are zero-based and belong to the interval. The first possible start is zero.
Working checkpoint — half-open interval: Test a record beginning at the first base. Treat bounds, order, precision and missingness as scientific inputs.
For parents, ask whether gaps mean missing measurements or measured zero.
Section 6 of 36
6. Read chromEnd as excluded
chromEnd marks the first base after the interval. It is not itself covered.
Working checkpoint — continuous signal: Calculate interval length by subtraction. Keep source hash, assembly, units, interval convention, sorting, gaps and transformation receipt together.
For a classroom model, assign rainfall values to invented time intervals.
Section 7 of 36
7. Use half-open adjacency
Intervals [0,10) and [10,20) meet without sharing a base.
Working checkpoint — bedGraph: Check touching intervals for accidental gaps or overlaps. Read each value beside its half-open interval, sample, scale and generation method.
For future study, connect genomics to piecewise functions and computing while checking pathways separately.
Section 8 of 36
8. Store variable widths explicitly
Unlike fixedStep Wiggle, every bedGraph row can state its own start and end.
Working checkpoint — chromStart: Validate positive length for every interval. Preserve omitted intervals, explicit zeros and rejected overlaps; they answer different questions.
For scientific writing, report intervals, units, gaps, overlaps and conversion checks.
Section 9 of 36
9. Use numerical order
UCSC requires input positions in numerical order. Sorting is part of the format contract.
Working checkpoint — chromEnd: Check order within every chromosome. Separate exact file values, visual summaries, statistical results and biological interpretation.
For students, ask where each value begins and where it stops.
Section 10 of 36
10. Define chromosome order
Natural, lexical and chrom.sizes order can differ. Converter behavior should determine the required ordering.
Working checkpoint — dataValue: Sort with a documented contig-order file. Treat bounds, order, precision and missingness as scientific inputs.
For parents, ask whether gaps mean missing measurements or measured zero.
Section 11 of 36
11. Keep integer and real values
UCSC permits positive or negative integer or real values.
Working checkpoint — half-open interval: Parse numerically and preserve intended precision. Keep source hash, assembly, units, interval convention, sorting, gaps and transformation receipt together.
For a classroom model, assign rainfall values to invented time intervals.
Section 12 of 36
12. Name units and baseline
A value of one can mean coverage, probability, enrichment or a normalised score.
Working checkpoint — continuous signal: Declare units, zero meaning and transformation. Read each value beside its half-open interval, sample, scale and generation method.
For future study, connect genomics to piecewise functions and computing while checking pathways separately.
Section 13 of 36
13. Treat gaps as no data
Only positions present in the input are graphed. A gap is not automatically a measured zero.
Working checkpoint — bedGraph: Record the missingness policy before aggregation. Preserve omitted intervals, explicit zeros and rejected overlaps; they answer different questions.
For scientific writing, report intervals, units, gaps, overlaps and conversion checks.
Section 14 of 36
14. Handle explicit zero separately
A row with value zero records a value over an interval; an omitted interval records no row.
Working checkpoint — chromStart: Keep explicit zero distinct from absence. Separate exact file values, visual summaries, statistical results and biological interpretation.
For students, ask where each value begins and where it stops.
Section 15 of 36
15. Detect overlaps
Overlapping rows can assign competing values to the same base and are not self-explanatory.
Working checkpoint — chromEnd: Reject or define overlap semantics explicitly. Treat bounds, order, precision and missingness as scientific inputs.
For parents, ask whether gaps mean missing measurements or measured zero.
Section 16 of 36
16. Detect duplicate intervals
Repeated identical coordinates may reflect replicates, accidental concatenation or alternative measures.
Working checkpoint — dataValue: Resolve duplicates before display or conversion. Keep source hash, assembly, units, interval convention, sorting, gaps and transformation receipt together.
For a classroom model, assign rainfall values to invented time intervals.
Section 17 of 36
17. Check bounds against contig length
Starts must not be negative and ends must fit the intended reference sequence.
Working checkpoint — half-open interval: Validate every row against chrom.sizes. Read each value beside its half-open interval, sample, scale and generation method.
For future study, connect genomics to piecewise functions and computing while checking pathways separately.
Section 18 of 36
18. Preserve exact source values
UCSC says bedGraph export preserves original data values, unlike its compressed Wiggle custom-track storage.
Working checkpoint — continuous signal: Compare source and exported text byte or number wise. Preserve omitted intervals, explicit zeros and rejected overlaps; they answer different questions.
For scientific writing, report intervals, units, gaps, overlaps and conversion checks.
Section 19 of 36
19. Keep upstream precision visible
Exact storage cannot restore digits discarded during signal calculation.
Working checkpoint — bedGraph: Report upstream rounding and numeric type. Separate exact file values, visual summaries, statistical results and biological interpretation.
For students, ask where each value begins and where it stops.
Section 20 of 36
20. Separate intervals from features
bedGraph assigns numerical signal to spans; BED and GFF3 can encode richer named or hierarchical features.
Working checkpoint — chromStart: Choose a model that fits the scientific object. Treat bounds, order, precision and missingness as scientific inputs.
For parents, ask whether gaps mean missing measurements or measured zero.
Section 21 of 36
21. Use display options cautiously
Autoscale, colours and graph type can change emphasis without changing the file.
Working checkpoint — chromEnd: Use common scales for visual comparisons. Keep source hash, assembly, units, interval convention, sorting, gaps and transformation receipt together.
For a classroom model, assign rainfall values to invented time intervals.
Section 22 of 36
22. Understand window aggregation
Zoomed-out browser views may summarise many intervals. Peaks can change with aggregation rule.
Working checkpoint — dataValue: Inspect raw values and record windowing settings. Read each value beside its half-open interval, sample, scale and generation method.
For future study, connect genomics to piecewise functions and computing while checking pathways separately.
Section 23 of 36
23. Convert large data to bigWig
UCSC recommends bigWig for very large datasets and rejects extremely large custom bedGraph uploads.
Working checkpoint — half-open interval: Plan binary conversion before distribution. Preserve omitted intervals, explicit zeros and rejected overlaps; they answer different questions.
For scientific writing, report intervals, units, gaps, overlaps and conversion checks.
Section 24 of 36
24. Supply the correct chrom.sizes
bedGraphToBigWig needs contig lengths that match the data assembly and names.
Working checkpoint — continuous signal: Hash and cite the exact sizes file. Separate exact file values, visual summaries, statistical results and biological interpretation.
For students, ask where each value begins and where it stops.
Section 25 of 36
25. Remove browser lines before conversion
Conversion utilities expect data rather than custom-track browser controls.
Working checkpoint — bedGraph: Validate a clean four-column input. Treat bounds, order, precision and missingness as scientific inputs.
For parents, ask whether gaps mean missing measurements or measured zero.
Section 26 of 36
26. Compare source and bigWig
Regional queries should reproduce intervals and values within declared numeric tolerance.
Working checkpoint — chromStart: Test multiple chromosomes, gaps and extrema. Keep source hash, assembly, units, interval convention, sorting, gaps and transformation receipt together.
For a classroom model, assign rainfall values to invented time intervals.
Section 27 of 36
27. Protect sensitive signals
Genomic tracks derived from human data can remain sensitive even without read-level records.
Working checkpoint — chromEnd: Apply consent, access and governance separately from format. Read each value beside its half-open interval, sample, scale and generation method.
For future study, connect genomics to piecewise functions and computing while checking pathways separately.
Section 28 of 36
28. Connect to Primary Science
A rainfall chart can assign one value to each time interval and distinguish zero rain from missing observations.
Working checkpoint — dataValue: Use a number line with half-open intervals. Preserve omitted intervals, explicit zeros and rejected overlaps; they answer different questions.
For scientific writing, report intervals, units, gaps, overlaps and conversion checks.
Section 29 of 36
29. Connect to Secondary Science
Signal tracks connect experiments, graphs, units, normalisation and uncertainty.
Working checkpoint — half-open interval: Ask what process produced each interval value. Separate exact file values, visual summaries, statistical results and biological interpretation.
For students, ask where each value begins and where it stops.
Section 30 of 36
30. Connect to mathematics
Half-open intervals, piecewise functions, signed values and aggregation connect directly to algebra and statistics.
Working checkpoint — continuous signal: Plot a short bedGraph by hand. Treat bounds, order, precision and missingness as scientific inputs.
For parents, ask whether gaps mean missing measurements or measured zero.
Section 31 of 36
31. Connect to computing
Delimited parsing, sorting, validation and indexed conversion demonstrate data-engineering fundamentals.
Working checkpoint — bedGraph: Test malformed coordinates and duplicate regions. Keep source hash, assembly, units, interval convention, sorting, gaps and transformation receipt together.
For a classroom model, assign rainfall values to invented time intervals.
Section 32 of 36
32. Connect to pathways
Genomic signal analysis links biology, statistics, visualisation and software and can illuminate STEM pathways.
Working checkpoint — chromStart: Check current official qualifications separately. Read each value beside its half-open interval, sample, scale and generation method.
For future study, connect genomics to piecewise functions and computing while checking pathways separately.
Section 33 of 36
33. Did you know?
In bedGraph, the interval from chromStart 0 to chromEnd 1 covers exactly the first base because the end is excluded.
Working checkpoint — chromEnd: Use subtraction to verify the length. Preserve omitted intervals, explicit zeros and rejected overlaps; they answer different questions.
For scientific writing, report intervals, units, gaps, overlaps and conversion checks.
Section 34 of 36
34. Write a reader-ready signal note
Name assembly, units, generation method, interval convention, sorting, gaps, overlaps, precision, display and hash.
Working checkpoint — dataValue: Keep the note beside every derived view. Separate exact file values, visual summaries, statistical results and biological interpretation.
For students, ask where each value begins and where it stops.
Section 35 of 36
35. Build a validation ladder
Check four columns, numeric types, bounds, order, overlaps, missingness, source identity and conversion equivalence.
Working checkpoint — half-open interval: Stop at the earliest unexplained interval. Treat bounds, order, precision and missingness as scientific inputs.
For parents, ask whether gaps mean missing measurements or measured zero.
Section 36 of 36
36. Finish with a reproducible bundle
Deliver bedGraph, manifest, chrom.sizes, commands, hashes, query fixtures, validation logs and bounded claims.
Working checkpoint — continuous signal: Make every signal value traceable to one interval and method. Keep source hash, assembly, units, interval convention, sorting, gaps and transformation receipt together.
For a classroom model, assign rainfall values to invented time intervals.
Contents · Previous section · Continue to the Science Learning Hub