eduKateSG · Why Science?
Turn peptide evidence into a readable genome-browser track—while preserving intervals, blocks, scores, modifications, references and uncertainty
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. proBed 1.0 is a HUPO-PSI proteogenomics format for storing peptide-spectrum matches or peptide groups in genomic context. The official page says it extends the 12-column BED structure with 13 additional proteomics fields. Because it remains tab-delimited and interval-oriented, records can support genome-browser tracks and can be converted to bigBed with an appropriate autoSQL definition. A visible track is an efficient map, not proof that a peptide is unique, a coding region is novel or the supporting spectrum is correct. 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; bigBed and indexed genome annotation tracks; mzIdentML and proteomics identification results; mzTab and tabular proteomics results; Science Learning Hub; Education Hub; STEM Education; Careers by Subject Capability.
Primary and official sources checked on 12 October 2026: HUPO-PSI official proBed 1.0.0 page, specification and examples, checked 12 October 2026; Peer-reviewed HUPO-PSI proBAM and proBed standard article, Genome Biology 2018, checked 12 October 2026; UCSC official BED format reference, 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, drafts, vocabularies, tools and recommendations can change, so preserve dated documentation, exact source data and software versions.
A deeper reading route
BED is successful because a few tab-separated fields can place features on a genome and optional fields can describe blocks, strand and display. proBed keeps that interval-first idea and adds proteomics-specific columns. The official PSI page says the extension adds 13 fields beyond the 12 original BED fields, allowing peptide-spectrum matches or grouped peptides to be viewed with genomic annotations.
The representation is deliberately lighter than proBAM. proBAM uses alignment-style records and a binary BAM option; proBed uses a browser-track table. That makes proBed convenient for sorting, intersections and visualisation, but a compact interval may not preserve every detail of a rich identification result. The source mzIdentML or mzTab should remain part of the evidence bundle.
Coordinates inherit BED conventions. Chromosome start is zero-based and end is excluded; blocks can describe exon-separated placements. Display columns such as score, strand, thick start, thick end, colour and block arrays have defined meanings. Off-by-one errors are especially dangerous because a track can still look plausible while the peptide is placed on the wrong residue or exon boundary.
Proteomics extension fields connect the interval to protein accessions, peptide sequence, uniqueness, genome reference, score, false-discovery context, modifications, charge, observed and calculated mass-to-charge values, spectrum reference and annotation or novelty status under the specification. Each value remains dependent on upstream databases and scoring rules.
Uniqueness must be scoped. A peptide may be unique within one protein database release yet shared in another, and unique protein mapping does not always mean unique genomic locus. Reports should name the database, mapping rule and level of uniqueness rather than treating a Boolean field as universal truth.
Block structure is central for splice-junction peptides. A proBed record can use BED block sizes and starts to represent peptide segments across exons. The total translated coding sequence across blocks should reproduce the peptide under the declared strand and transcript model. Browser appearance alone cannot verify the translation.
bigBed conversion can make large tracks efficient to browse remotely. The official proBed page links an autoSQL definition, chromosome-size file and conversion examples. The binary file and index-like structures are derived from exact source proBed, reference chromosome sizes and schema. Preserve all of those inputs and validate counts after conversion.
Genome-browser styling is communication, not evidence strength. Colour, bar height or a bold label can make one interval look important. The legend should explain score transforms, novelty categories and filters. Users should be able to open the source PSM, spectrum and mapping receipt rather than infer confidence from aesthetics.
A practical audit starts with one track item. Decode its BED interval and blocks, retrieve the peptide and spectrum reference, check score, FDR, charge and modifications, resolve the protein and transcript database, translate the genomic blocks, then compare observed and calculated m/z. This joins the visible feature to its complete evidence chain.
The optimistic value is navigable integration. Researchers and students can see where protein-level evidence sits beside genes, variants and RNA features. The track makes relationships easier to notice; careful provenance and conservative interpretation stop that visual clarity from becoming false certainty.
Interval operations can answer powerful questions—whether peptide evidence overlaps an exon, variant, untranslated region or alternative frame—but only after coordinate systems are reconciled. Record assembly, chromosome naming and zero-based half-open convention for every input track before using intersect or merge operations.
Merging proBed items can destroy evidence granularity. Two overlapping peptides may have different spectra, charges, modifications or confidence. Keep source rows and use derived summary tracks with explicit aggregation rules rather than replacing detailed records with one visually tidy interval.
Track hubs and remote bigBed files need stable hosting, byte-range support and versioned URLs. Publish the underlying proBed and autoSQL schema, test representative remote queries and use hashes so a browser session can be tied to exact bytes rather than a mutable latest file.
Visual encoding should be accessible. Use a restrained palette, readable labels and a legend that does not rely on colour alone. Distinguish mapping category from identification confidence, and let users click through to the source record instead of compressing every property into colour.
Novel coding claims need comparison with current annotation releases, alternative translations, contaminant databases and independent evidence. A peptide interval outside a known coding region is an investigation route, not a completed discovery. Versioned labels allow today’s novel feature to become tomorrow’s annotated feature without rewriting history.
A publication bundle should contain the source identification result, proBed text, chromosome sizes, autoSQL, generated bigBed, conversion commands, validation output and a small screenshot only as illustration. The data and receipts—not the picture—support reproducibility.
Browser sessions should declare the reference assembly and supporting annotation tracks. A peptide item can appear to overlap a gene only because two tracks use different assemblies or naming conventions. Save a session description or track configuration, but keep canonical URLs and hashes outside temporary browser state so the analysis remains reconstructable.
Score scaling deserves explicit documentation. BED display scores usually occupy a limited integer range, while PSM scores and false-discovery values follow method-specific scales. A transformation may improve visual contrast but must not replace the original scientific value. Store both, publish the mapping and ensure that a high bar is not mistaken for universal confidence.
Shared peptides can create repeated intervals across genes or paralogues. Deduplicating by peptide sequence alone can erase biologically important ambiguity, while counting every interval as independent evidence can inflate support. Keep a mapping group identifier and report unique peptides separately from the complete set of valid placements.
Coordinate tests should include the first base of a chromosome, single-base intervals, exon junctions and reverse-strand features. These fixtures expose off-by-one and orientation errors before a full track is published. Keep the expected browser display and translated peptide beside each fixture so both software and scientific interpretation are tested.
Finally, validate that every displayed item opens the intended evidence record. Broken identifiers can leave a beautiful track scientifically unauditable.
| Stage | Question | Evidence | Responsible output |
|---|---|---|---|
| Locate | Where does peptide evidence map? | BED chromosome, start, end, strand and blocks | A genome-browser interval |
| Describe | Which proteomics evidence supports it? | Peptide, protein, score, FDR, charge and modifications | A linked PSM or peptide record |
| Display | How should the feature be explored? | Track labels, colours, bigBed conversion and browser settings | A navigable visual layer |
| Verify | What conclusion is justified? | Source spectrum, reference releases, translation and alternatives | A bounded proteogenomic claim |
Inside this guide
1–12 · Foundations and design
13–24 · Models and estimation
- 13. Read block sizes and starts
- 14. Read protein accessions
- 15. Read peptide sequences
- 16. Read uniqueness
- 17. Read genome reference fields
- 18. Read PSM scores and FDR
- 19. Read modifications
- 20. Read charge and m/z
- 21. Read spectrum references
- 22. Read annotation status
- 23. Validate tabular structure
- 24. Validate translation
25–36 · Diagnostics and interpretation
- 25. Convert to bigBed carefully
- 26. Protect sensitive findings
- 27. Separate visual salience from certainty
- 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 methods note
- 35. Build a validation ladder
- 36. Finish with a reproducible bundle
Section 1 of 36
1. Start with the track question
Decide whether the track supports visual review, overlap analysis, teaching or publication.
Working checkpoint — proBed: Write the intended user and claim. Keep source result, 25 fields, reference releases, interval blocks, spectrum links and validation together.
For a classroom model, draw fictional exon blocks and peptide spans.
Section 2 of 36
2. Use proBed 1.0.0
The official PSI page identifies proBed 1.0.0 as the final version and supplies examples.
Working checkpoint — proteogenomics track: Pin specification, schema and converter. Read every displayed feature beside peptide uniqueness, score, FDR, charge and m/z evidence.
For future study, connect molecules to maps while checking pathways separately.
Section 3 of 36
3. Preserve the source result
proBed is a mapping view of peptide or PSM evidence, not a replacement for mzIdentML or mzTab.
Working checkpoint — BED interval: Keep source result and spectrum links. Preserve the original proBed before bigBed conversion; a browser track is a derived view.
For writing, cite source files, assembly, proBed version, converter and limits.
Section 4 of 36
4. Count the columns
proBed extends the 12 BED columns with 13 proteomics fields. Tabs and order define the record.
Working checkpoint — peptide-spectrum match: Validate exactly 25 expected fields for full records. Separate interval validity from identification confidence, novelty and function.
For students, ask what evidence sits behind a track item.
Section 5 of 36
5. Read chromosome names
Chromosome labels depend on a particular assembly and naming convention.
Working checkpoint — genome browser: Match the chromosome-size file and reference. Treat tabs, zero-based starts, excluded ends and blocks as scientific inputs: one offset can change the locus.
For parents, ask whether a clear visual still shows uncertainty.
Section 6 of 36
6. Read zero-based starts
BED chromStart is zero-based while many biological descriptions use one-based positions.
Working checkpoint — bigBed: Convert coordinates explicitly. Keep source result, 25 fields, reference releases, interval blocks, spectrum links and validation together.
For a classroom model, draw fictional exon blocks and peptide spans.
Section 7 of 36
7. Read excluded ends
BED chromEnd is not included in the interval.
Working checkpoint — proBed: Test interval length and boundaries. Read every displayed feature beside peptide uniqueness, score, FDR, charge and m/z evidence.
For future study, connect molecules to maps while checking pathways separately.
Section 8 of 36
8. Read feature names
The name helps display but is not a stable evidence key unless governed.
Working checkpoint — proteogenomics track: Link to peptide or PSM identifier. Preserve the original proBed before bigBed conversion; a browser track is a derived view.
For writing, cite source files, assembly, proBed version, converter and limits.
Section 9 of 36
9. Read display scores
BED score has a constrained display role and may transform scientific confidence.
Working checkpoint — BED interval: Publish the transformation and original score. Separate interval validity from identification confidence, novelty and function.
For students, ask what evidence sits behind a track item.
Section 10 of 36
10. Read strand
Strand determines how genomic sequence is interpreted for coding evidence.
Working checkpoint — peptide-spectrum match: Verify transcript orientation. Treat tabs, zero-based starts, excluded ends and blocks as scientific inputs: one offset can change the locus.
For parents, ask whether a clear visual still shows uncertainty.
Section 11 of 36
11. Read thick and colour fields
Browser fields can highlight coding or category information but do not add evidence.
Working checkpoint — genome browser: Provide an honest legend. Keep source result, 25 fields, reference releases, interval blocks, spectrum links and validation together.
For a classroom model, draw fictional exon blocks and peptide spans.
Section 12 of 36
12. Read block counts
Blocks represent separated genomic pieces such as exonic segments.
Working checkpoint — bigBed: Check count against both arrays. Read every displayed feature beside peptide uniqueness, score, FDR, charge and m/z evidence.
For future study, connect molecules to maps while checking pathways separately.
Section 13 of 36
13. Read block sizes and starts
Block arrays are relative to the feature start and must reconstruct the full mapped interval.
Working checkpoint — proBed: Recalculate every boundary. Preserve the original proBed before bigBed conversion; a browser track is a derived view.
For writing, cite source files, assembly, proBed version, converter and limits.
Section 14 of 36
14. Read protein accessions
Protein identifiers connect peptide evidence to a database entry and release.
Working checkpoint — proteogenomics track: Preserve namespace and version. Separate interval validity from identification confidence, novelty and function.
For students, ask what evidence sits behind a track item.
Section 15 of 36
15. Read peptide sequences
Exact sequence, ambiguity and modification context determine mapping.
Working checkpoint — BED interval: Avoid display-only abbreviations. Treat tabs, zero-based starts, excluded ends and blocks as scientific inputs: one offset can change the locus.
For parents, ask whether a clear visual still shows uncertainty.
Section 16 of 36
16. Read uniqueness
Uniqueness depends on the reference database and mapping scope.
Working checkpoint — peptide-spectrum match: State protein and genome criteria. Keep source result, 25 fields, reference releases, interval blocks, spectrum links and validation together.
For a classroom model, draw fictional exon blocks and peptide spans.
Section 17 of 36
17. Read genome reference fields
The assembly and annotation release give coordinates their meaning.
Working checkpoint — genome browser: Archive source references. Read every displayed feature beside peptide uniqueness, score, FDR, charge and m/z evidence.
For future study, connect molecules to maps while checking pathways separately.
Section 18 of 36
18. Read PSM scores and FDR
Score and false-discovery fields need method, direction and threshold definitions.
Working checkpoint — bigBed: Keep unfiltered results available. Preserve the original proBed before bigBed conversion; a browser track is a derived view.
For writing, cite source files, assembly, proBed version, converter and limits.
Section 19 of 36
19. Read modifications
Modified residues need position and controlled terminology and may be artefactual or biological.
Working checkpoint — proBed: Separate modification from genomic change. Separate interval validity from identification confidence, novelty and function.
For students, ask what evidence sits behind a track item.
Section 20 of 36
20. Read charge and m/z
Observed and calculated mass-to-charge values provide a numerical consistency check.
Working checkpoint — proteogenomics track: Recalculate with stated charge. Treat tabs, zero-based starts, excluded ends and blocks as scientific inputs: one offset can change the locus.
For parents, ask whether a clear visual still shows uncertainty.
Section 21 of 36
21. Read spectrum references
A track item should point back to the spectrum that supports the PSM.
Working checkpoint — BED interval: Test representative retrieval. Keep source result, 25 fields, reference releases, interval blocks, spectrum links and validation together.
For a classroom model, draw fictional exon blocks and peptide spans.
Section 22 of 36
22. Read annotation status
Known or novel labels depend on the chosen annotation release and decision rules.
Working checkpoint — peptide-spectrum match: Treat novelty as provisional and versioned. Read every displayed feature beside peptide uniqueness, score, FDR, charge and m/z evidence.
For future study, connect molecules to maps while checking pathways separately.
Section 23 of 36
23. Validate tabular structure
Tabs, field types, ranges and array cardinalities should pass a specification-aware validator.
Working checkpoint — genome browser: Keep the complete report. Preserve the original proBed before bigBed conversion; a browser track is a derived view.
For writing, cite source files, assembly, proBed version, converter and limits.
Section 24 of 36
24. Validate translation
Genome blocks, strand and frame should reproduce the peptide sequence.
Working checkpoint — bigBed: Translate independent fixtures. Separate interval validity from identification confidence, novelty and function.
For students, ask what evidence sits behind a track item.
Section 25 of 36
25. Convert to bigBed carefully
Conversion needs the exact proBed table, autoSQL schema and chromosome sizes.
Working checkpoint — proBed: Compare row counts and sample records. Treat tabs, zero-based starts, excluded ends and blocks as scientific inputs: one offset can change the locus.
For parents, ask whether a clear visual still shows uncertainty.
Section 26 of 36
26. Protect sensitive findings
Variant or disease-linked peptide tracks can expose governed sample information.
Working checkpoint — proteogenomics track: Apply consent and access review. Keep source result, 25 fields, reference releases, interval blocks, spectrum links and validation together.
For a classroom model, draw fictional exon blocks and peptide spans.
Section 27 of 36
27. Separate visual salience from certainty
A bright track feature can still represent a shared peptide or weak PSM.
Working checkpoint — BED interval: Let evidence, not styling, control the claim. Read every displayed feature beside peptide uniqueness, score, FDR, charge and m/z evidence.
For future study, connect molecules to maps while checking pathways separately.
Section 28 of 36
28. Connect to Primary Science
A fictional genome-track card helps younger learners see that an observation, a label and an explanation are different layers.
Working checkpoint — peptide-spectrum match: Define every symbol and use no personal data. Preserve the original proBed before bigBed conversion; a browser track is a derived view.
For writing, cite source files, assembly, proBed version, converter and limits.
Section 29 of 36
29. Connect to Secondary Science
proBed supports lessons about molecules, evidence, models, controls and uncertainty.
Working checkpoint — genome browser: Trace one conclusion back to the smallest supporting record. Separate interval validity from identification confidence, novelty and function.
For students, ask what evidence sits behind a track item.
Section 30 of 36
30. Connect to mathematics
Coordinates, masses, charges, intervals, scores and probabilities reward exact definitions and compatible units.
Working checkpoint — bigBed: Recalculate one small example by hand. Treat tabs, zero-based starts, excluded ends and blocks as scientific inputs: one offset can change the locus.
For parents, ask whether a clear visual still shows uncertainty.
Section 31 of 36
31. Connect to computing
Typed tabular records, interval indexes and browser rendering show how scientific evidence becomes navigable.
Working checkpoint — proBed: Break one field or reference, run a validator and repair it transparently. Keep source result, 25 fields, reference releases, interval blocks, spectrum links and validation together.
For a classroom model, draw fictional exon blocks and peptide spans.
Section 32 of 36
32. Connect to pathways
Proteogenomic tracks connect molecular biology, genome browsers, statistics and software and can illuminate STEM education and career pathways.
Working checkpoint — proteogenomics track: Verify current official education and qualification routes separately. Read every displayed feature beside peptide uniqueness, score, FDR, charge and m/z evidence.
For future study, connect molecules to maps while checking pathways separately.
Section 33 of 36
33. Did you know?
proBed adds 13 proteomics columns to the familiar 12-column BED structure.
Working checkpoint — BED interval: Explain the fact in plain language, then name the evidence boundary. Preserve the original proBed before bigBed conversion; a browser track is a derived view.
For writing, cite source files, assembly, proBed version, converter and limits.
Section 34 of 36
34. Write a reader-ready methods note
State format status, version, inputs, reference resources, software, validation and the narrow claim supported.
Working checkpoint — peptide-spectrum match: Keep the note beside the machine-readable record. Separate interval validity from identification confidence, novelty and function.
For students, ask what evidence sits behind a track item.
Section 35 of 36
35. Build a validation ladder
Check bytes, structure, identifiers, references, controlled meaning, scientific plausibility and a known fixture in sequence.
Working checkpoint — genome browser: Stop at the earliest unresolved layer. Treat tabs, zero-based starts, excluded ends and blocks as scientific inputs: one offset can change the locus.
For parents, ask whether a clear visual still shows uncertainty.
Section 36 of 36
36. Finish with a reproducible bundle
Deliver source files or stable identifiers, format documents, reference releases, commands, validator output, hashes and limitations.
Working checkpoint — bigBed: Make every published claim traceable. Keep source result, 25 fields, reference releases, interval blocks, spectrum links and validation together.
For a classroom model, draw fictional exon blocks and peptide spans.
Contents · Previous section · Continue to the Science Learning Hub