eduKateSG · Why Science?
Compare many genomes in alignment blocks—while keeping reference order, strand, gaps, missing species and evolutionary limits visible
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. UCSC’s Multiple Alignment Format (MAF) stores a series of DNA-level whole-genome multiple alignments as blocks. The file begins with a ##maf header; each alignment paragraph begins with an a line and contains one s line per aligned sequence, with the first s line representing the reference genome. Sequence rows record source, zero-based start, ungapped size, strand, full source size and gapped alignment text. Optional i, e and q lines can describe flanking context, empty components and base quality. A MAF block represents an alignment model for the included sequences and region; it does not prove functional conservation, species absence, direct ancestry or an unchanged evolutionary mechanism. 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: FASTA sequence records and reference data; BED and genomic feature coordinates; 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 MAF format specification, checked 12 October 2026; UCSC Genome Browser multiple-alignment track documentation, checked 12 October 2026; UCSC Genome Browser Data File Formats FAQ, 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
MAF is block-oriented because whole genomes contain rearrangements, missing sequence and strand changes that do not fit one uninterrupted alignment. Each paragraph is a local alignment with a chosen reference row followed by other contributing sequences. Blank lines are therefore structural delimiters, not cosmetic spacing.
The s line makes every alignment row self-describing. src identifies the source sequence, start locates the ungapped region, size counts non-dash characters, strand gives orientation, srcSize gives the full sequence length and text shows aligned bases with dashes. Text strings within one block must occupy the same number of alignment columns even though their ungapped sizes differ.
Negative-strand coordinates need special attention. UCSC defines start relative to the reverse-complemented source when strand is minus. To compare with forward coordinates, transform using srcSize and the ungapped size. Mixing stored start with a forward-strand interpretation shifts the region and can reverse comparative conclusions.
The first sequence is the reference for that alignment block. Reference-first ordering supports browser display and coordinate anchoring but does not make that species an evolutionary ancestor or the scientifically neutral centre of the analysis. A different reference can expose a different set of aligned blocks and missing components.
Gaps do not all mean the same thing. A dash in alignment text can reflect an insertion or deletion relative to another row under the alignment model. Missing species from a block may reflect true absence, incomplete assembly, filtering or failure to align. Optional e and i lines can add context, but interpretation still needs the pipeline and assemblies.
Alignment scores are meaningful only with their scoring scheme. The header can name the generating program and scheme, while each a line can carry an optional score or pass number. Cross-file score comparisons are unsafe unless methods and scales are compatible.
Validation checks structure and biological context. Confirm header version, block delimiters, reference-first rows, equal text widths, size versus non-dash count, source bounds, strand transforms and permitted optional lines. Then inspect representative regions against source assemblies and known rearrangements.
For students, MAF creates a cheerful multi-column view of evolution. Shared letters can suggest conserved sequence, while differences invite questions. The durable lesson is that alignment is a model built from chosen genomes, assemblies, algorithms and scores—powerful evidence when those choices remain visible.
Coverage summaries need reference-aware denominators. The proportion of a reference aligned to another species depends on excluded regions, assembly gaps, minimum score and block filtering. Report the number of eligible reference bases, aligned bases and blocks, along with the exact species pair and assembly versions. Without those denominators, a conservation percentage is not reproducible.
Column-level identity should not be confused with independence. Neighbouring alignment columns share evolutionary history, and species are related by a phylogenetic tree rather than being independent replicates. Formal conservation or selection analysis needs an evolutionary model that accounts for those relationships; counting matching letters is an informative description, not a complete statistical test.
Reference bias can be explored by comparing another alignment or re-anchoring where possible. Regions missing from the chosen reference cannot appear as ordinary reference-anchored blocks, and assembly quality differs among species. Discuss which genomes were high quality, which regions were filtered and how reference choice shaped the visible comparison.
A durable MAF archive should include block statistics that can be recomputed: number of blocks, row-count distribution, total reference aligned bases, strand counts, optional-line counts and per-source participation. Add known plus and minus blocks as fixtures. These summaries catch truncation or parsing changes without pretending to validate evolutionary interpretation by themselves.
Block boundaries are analytical choices. Two adjacent blocks may be separated because of filtering, rearrangement, missing sequence or the behaviour of the alignment pipeline, and a feature can overlap several blocks with different species participation. Define whether summaries count blocks, columns, reference bases or features, then handle overlaps without double-counting. When extracting a gene or regulatory region, retain the original block identifiers and clipping coordinates so another reader can reconstruct exactly which alignment columns contributed to the conclusion.
Missingness should be classified before it is interpreted. A species row can be absent because the sequence did not align, the assembly lacks that region, the pipeline filtered the candidate, or the source was not included in that alignment set. Optional empty-component lines may add context, but they still do not prove biological deletion. Compare assembly completeness and supporting alignments, and use an explicit state such as unavailable or unresolved instead of turning every blank into absence. This discipline keeps evolutionary stories proportional to what the file actually records.
| Stage | Question | Evidence | Responsible output |
|---|---|---|---|
| Define | Which genomes and reference anchor are compared? | Assembly manifests, header and first s row | A bounded comparative context |
| Align | Which sequences occupy one local block? | a line plus equal-width s rows | A multiple-alignment column model |
| Contextualise | What do gaps, strands and absent rows mean? | Coordinates, i/e/q lines and pipeline metadata | Qualified structural evidence |
| Interpret | What evolutionary claim is supported? | Replicates across regions, models and functional evidence | A cautious comparative conclusion |
Inside this guide
1–12 · Foundations and design
- 1. Start with the comparative question
- 2. Recognise a series of blocks
- 3. Use the required header
- 4. Keep comments and metadata distinct
- 5. Begin each block with a
- 6. Read score with its scheme
- 7. Read pass metadata cautiously
- 8. Require reference first
- 9. Read src
- 10. Read start as zero-based
- 11. Read size as ungapped length
- 12. Read strand
13–24 · Models and estimation
- 13. Read srcSize
- 14. Read alignment text
- 15. Transform minus coordinates
- 16. Interpret a column
- 17. Interpret dashes cautiously
- 18. Understand optional i lines
- 19. Understand optional e lines
- 20. Understand optional q lines
- 21. Allow ignorable extensions
- 22. Keep blocks independent
- 23. Do not equate absence with deletion
- 24. Protect the reference choice
25–36 · Diagnostics and interpretation
- 25. Handle rearrangements
- 26. Summarise conservation carefully
- 27. Validate against source sequence
- 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 alignment note
- 35. Build a validation ladder
- 36. Finish with a reproducible bundle
Section 1 of 36
1. Start with the comparative question
A MAF should answer a defined question about which assemblies and regions are compared. More species do not automatically make a fairer analysis.
Working checkpoint — MAF alignment: Write reference, species set and intended inference. Keep assembly manifests, reference order, alignment program, scoring scheme, filters, MAF hash and source checks together.
For a classroom model, align invented strings and distinguish dash from unknown.
Section 2 of 36
2. Recognise a series of blocks
MAF stores many local multiple alignments rather than one uninterrupted whole-genome matrix.
Working checkpoint — alignment block: Preserve blank-line paragraph boundaries. Read every block beside strand, source bounds, species participation, gaps and quality context.
For future study, connect evolution to matrices and algorithms while checking pathways separately.
Section 3 of 36
3. Use the required header
The first line begins with ##maf and declares version one, with optional scoring and program metadata.
Working checkpoint — reference genome: Validate header before reading blocks. Preserve absent rows and optional lines; a tidy reference view is not the complete comparative evidence.
For scientific writing, report blocks, assemblies, scoring and bounded claims.
Section 4 of 36
4. Keep comments and metadata distinct
Lines beginning with # are comments, while ## lines can carry metadata for parsers.
Working checkpoint — sequence row: Preserve provenance comments without mistaking them for sequence rows. Separate alignment columns, evolutionary models, functional annotation and experiment.
For students, ask which letters share a column and which conclusion that does not yet prove.
Section 5 of 36
5. Begin each block with a
Every alignment paragraph begins with an a line followed by name=value attributes.
Working checkpoint — reverse strand: Reject sequence rows that occur outside a block. Treat reference choice, gaps, orientation and sampling as scientific inputs.
For parents, ask whether the comparison names all genomes and limitations.
Section 6 of 36
6. Read score with its scheme
An a line can carry a floating-point score, and the header can identify the scoring system.
Working checkpoint — comparative genomics: Never compare scores across unknown schemes. Keep assembly manifests, reference order, alignment program, scoring scheme, filters, MAF hash and source checks together.
For a classroom model, align invented strings and distinguish dash from unknown.
Section 7 of 36
7. Read pass metadata cautiously
A pass number can identify stages of a multi-pass alignment procedure. It is method context, not an evolutionary rank.
Working checkpoint — MAF alignment: Record the producing program and parameters. Read every block beside strand, source bounds, species participation, gaps and quality context.
For future study, connect evolution to matrices and algorithms while checking pathways separately.
Section 8 of 36
8. Require reference first
The first s line in each block must be the reference genome on which the other sequences map.
Working checkpoint — alignment block: Check block order consistently. Preserve absent rows and optional lines; a tidy reference view is not the complete comparative evidence.
For scientific writing, report blocks, assemblies, scoring and bounded claims.
Section 9 of 36
9. Read src
src names the source sequence, often using database.chromosome for browser-resident assemblies.
Working checkpoint — reference genome: Resolve names against a versioned manifest. Separate alignment columns, evolutionary models, functional annotation and experiment.
For students, ask which letters share a column and which conclusion that does not yet prove.
Section 10 of 36
10. Read start as zero-based
start gives the zero-based beginning of the aligning region in its strand-specific source coordinates.
Working checkpoint — sequence row: Convert explicitly for one-based reports. Treat reference choice, gaps, orientation and sampling as scientific inputs.
For parents, ask whether the comparison names all genomes and limitations.
Section 11 of 36
11. Read size as ungapped length
size equals the number of non-dash characters in the alignment text, not the text width.
Working checkpoint — reverse strand: Recount non-dash symbols for every row. Keep assembly manifests, reference order, alignment program, scoring scheme, filters, MAF hash and source checks together.
For a classroom model, align invented strings and distinguish dash from unknown.
Section 12 of 36
12. Read strand
Strand is plus or minus and determines whether text comes from the source or its reverse complement.
Working checkpoint — comparative genomics: Carry orientation into coordinate interpretation. Read every block beside strand, source bounds, species participation, gaps and quality context.
For future study, connect evolution to matrices and algorithms while checking pathways separately.
Section 13 of 36
13. Read srcSize
srcSize is the length of the entire source sequence and supports bounds and reverse-strand transforms.
Working checkpoint — MAF alignment: Validate it against the exact assembly. Preserve absent rows and optional lines; a tidy reference view is not the complete comparative evidence.
For scientific writing, report blocks, assemblies, scoring and bounded claims.
Section 14 of 36
14. Read alignment text
text contains aligned bases or amino acids plus insertion dashes. All rows in a block align by column.
Working checkpoint — alignment block: Require equal text widths inside each block. Separate alignment columns, evolutionary models, functional annotation and experiment.
For students, ask which letters share a column and which conclusion that does not yet prove.
Section 15 of 36
15. Transform minus coordinates
For a minus row, the stored start is relative to the reverse-complemented source. Forward bounds derive from srcSize, start and size.
Working checkpoint — reference genome: Keep raw and transformed values together. Treat reference choice, gaps, orientation and sampling as scientific inputs.
For parents, ask whether the comparison names all genomes and limitations.
Section 16 of 36
16. Interpret a column
One alignment column proposes homologous positions or gaps under the alignment model. It is not itself a functional annotation.
Working checkpoint — sequence row: Separate alignment evidence from biological consequence. Keep assembly manifests, reference order, alignment program, scoring scheme, filters, MAF hash and source checks together.
For a classroom model, align invented strings and distinguish dash from unknown.
Section 17 of 36
17. Interpret dashes cautiously
A dash marks an alignment gap, whose evolutionary direction depends on the model, outgroups and context.
Working checkpoint — reverse strand: Avoid calling insertion or deletion from two rows alone. Read every block beside strand, source bounds, species participation, gaps and quality context.
For future study, connect evolution to matrices and algorithms while checking pathways separately.
Section 18 of 36
18. Understand optional i lines
An i line describes what lies before and after the immediately preceding sequence component.
Working checkpoint — comparative genomics: Preserve it when continuity matters. Preserve absent rows and optional lines; a tidy reference view is not the complete comparative evidence.
For scientific writing, report blocks, assemblies, scoring and bounded claims.
Section 19 of 36
19. Understand optional e lines
An e line can describe an empty component and the size of the gap between alignments spanning the block.
Working checkpoint — MAF alignment: Distinguish documented emptiness from an omitted species. Separate alignment columns, evolutionary models, functional annotation and experiment.
For students, ask which letters share a column and which conclusion that does not yet prove.
Section 20 of 36
20. Understand optional q lines
A q line can provide per-base quality information for the species row.
Working checkpoint — alignment block: Read quality symbols through the documented encoding. Treat reference choice, gaps, orientation and sampling as scientific inputs.
For parents, ask whether the comparison names all genomes and limitations.
Section 21 of 36
21. Allow ignorable extensions
The specification permits parsers to ignore other line types, but silent loss can matter to an analysis.
Working checkpoint — reference genome: Log every unknown line type. Keep assembly manifests, reference order, alignment program, scoring scheme, filters, MAF hash and source checks together.
For a classroom model, align invented strings and distinguish dash from unknown.
Section 22 of 36
22. Keep blocks independent
A species can appear in one block and be absent from the next because alignment and assembly support vary by region.
Working checkpoint — sequence row: Report per-block participation. Read every block beside strand, source bounds, species participation, gaps and quality context.
For future study, connect evolution to matrices and algorithms while checking pathways separately.
Section 23 of 36
23. Do not equate absence with deletion
A missing row may reflect assembly gaps, filters, weak similarity or lineage change.
Working checkpoint — reverse strand: Check source sequence and pipeline status before interpretation. Preserve absent rows and optional lines; a tidy reference view is not the complete comparative evidence.
For scientific writing, report blocks, assemblies, scoring and bounded claims.
Section 24 of 36
24. Protect the reference choice
Reference-anchored blocks privilege coordinates and sequence available in the chosen assembly.
Working checkpoint — comparative genomics: Discuss how another reference could change coverage. Separate alignment columns, evolutionary models, functional annotation and experiment.
For students, ask which letters share a column and which conclusion that does not yet prove.
Section 25 of 36
25. Handle rearrangements
Whole-genome comparisons can include inversions and rearranged pieces across separate blocks and strands.
Working checkpoint — MAF alignment: Inspect order and orientation beyond one block. Treat reference choice, gaps, orientation and sampling as scientific inputs.
For parents, ask whether the comparison names all genomes and limitations.
Section 26 of 36
26. Summarise conservation carefully
High identity can support sequence conservation under the sampled genomes and alignment method, but function needs additional evidence.
Working checkpoint — alignment block: Pair alignment with annotation or experiment when claiming function. Keep assembly manifests, reference order, alignment program, scoring scheme, filters, MAF hash and source checks together.
For a classroom model, align invented strings and distinguish dash from unknown.
Section 27 of 36
27. Validate against source sequence
Extract selected source intervals, orient them correctly and confirm non-gap characters match the MAF row.
Working checkpoint — reference genome: Test plus and minus fixtures. Read every block beside strand, source bounds, species participation, gaps and quality context.
For future study, connect evolution to matrices and algorithms while checking pathways separately.
Section 28 of 36
28. Connect to Primary Science
Aligned fictional symbol strings can show shared positions, differences and missing symbols.
Working checkpoint — sequence row: Explain that the activity is a model of comparison. Preserve absent rows and optional lines; a tidy reference view is not the complete comparative evidence.
For scientific writing, report blocks, assemblies, scoring and bounded claims.
Section 29 of 36
29. Connect to Secondary Science
Comparative sequence evidence connects variation, inheritance and evolution while reminding students that similarity is not identity of function.
Working checkpoint — reverse strand: Write the narrowest supported claim. Separate alignment columns, evolutionary models, functional annotation and experiment.
For students, ask which letters share a column and which conclusion that does not yet prove.
Section 30 of 36
30. Connect to mathematics
Matrices, sequences, gaps, proportions and coordinate transforms make MAF a practical discrete-data example.
Working checkpoint — comparative genomics: Calculate ungapped size and identity for a small block. Treat reference choice, gaps, orientation and sampling as scientific inputs.
For parents, ask whether the comparison names all genomes and limitations.
Section 31 of 36
31. Connect to computing
Block parsers, streaming, equal-width validation and reference manifests show how software supports comparative biology.
Working checkpoint — MAF alignment: Test truncated and malformed paragraphs. Keep assembly manifests, reference order, alignment program, scoring scheme, filters, MAF hash and source checks together.
For a classroom model, align invented strings and distinguish dash from unknown.
Section 32 of 36
32. Connect to pathways
Comparative genomics connects evolution, ecology, computing, statistics and data curation and can illuminate STEM pathways.
Working checkpoint — alignment block: Check current official qualifications separately. Read every block beside strand, source bounds, species participation, gaps and quality context.
For future study, connect evolution to matrices and algorithms while checking pathways separately.
Section 33 of 36
33. Did you know?
In a MAF s line, size counts only non-dash characters, while the alignment text can be longer because gaps occupy columns.
Working checkpoint — reference genome: Count both values in a classroom example. Preserve absent rows and optional lines; a tidy reference view is not the complete comparative evidence.
For scientific writing, report blocks, assemblies, scoring and bounded claims.
Section 34 of 36
34. Write a reader-ready alignment note
Name species and assemblies, reference order, generator, scoring, filters, block counts, quality and limits.
Working checkpoint — sequence row: Keep source manifests and file hash beside the MAF. Separate alignment columns, evolutionary models, functional annotation and experiment.
For students, ask which letters share a column and which conclusion that does not yet prove.
Section 35 of 36
35. Build a validation ladder
Check header, paragraphs, row order, widths, sizes, bounds, strands, optional lines, sequences and comparative claims.
Working checkpoint — reverse strand: Stop at the first unexplained block. Treat reference choice, gaps, orientation and sampling as scientific inputs.
For parents, ask whether the comparison names all genomes and limitations.
Section 36 of 36
36. Finish with a reproducible bundle
Deliver MAF hash, assembly manifests, commands, software, scoring, fixtures, summaries and bounded interpretation.
Working checkpoint — comparative genomics: Make every comparative statement traceable to blocks and sources. Keep assembly manifests, reference order, alignment program, scoring scheme, filters, MAF hash and source checks together.
For a classroom model, align invented strings and distinguish dash from unknown.
Contents · Previous section · Continue to the Science Learning Hub