Glossary
The vocabulary of bioinformatics, in plain words.
- Sequence
- The order of bases in DNA or RNA, or of amino acids in a protein, written as a string of letters.
- Alignment
- An arrangement of two or more sequences so that corresponding positions line up.
- Gap
- A position in an alignment where one sequence has nothing opposite the other: an insertion or deletion.
- Substitution matrix
- A table scoring how likely each letter is to be replaced by each other letter. PAM and BLOSUM are the classic ones.
- Homology
- Similarity due to common ancestry. Orthologues were separated by speciation; paralogues by gene duplication.
- E-value
- The number of database matches at least this good that would be expected by chance.
- Read
- One stretch of sequence produced by a sequencing machine.
- Coverage
- The average number of reads covering each position of a genome. Also called depth.
- Assembly
- Reconstructing a long sequence from overlapping reads.
- Contig
- A continuous stretch of assembled sequence with no gaps.
- k-mer
- A substring of length k. Counting and linking k-mers underlies many assemblers.
- De Bruijn graph
- A graph whose edges are k-mers; a genome corresponds to a path through it.
- Reference genome
- An agreed representative sequence for a species, against which others are compared.
- Mapping
- Finding where each read belongs on a reference genome.
- Variant calling
- Identifying positions where a sample differs from the reference.
- Annotation
- Marking the features of a sequence: genes, regulatory regions, repeats.
- FASTA and FASTQ
- Plain-text formats for sequences; FASTQ adds a quality score for every base.
- A probabilistic model of a sequence family, used to find distant relatives and to predict genes.
- Phylogenetic tree
- A branching diagram of the inferred evolutionary relationships among sequences or species.
- Ontology
- A controlled vocabulary with defined relationships between its terms.
- Workflow
- A recorded chain of analysis steps, with software versions, that can be rerun.
- False discovery rate
- The expected share of false positives among the results declared significant.