Bioinformatics is full of jargon. Here's a friendly reference to the terms you'll meet in Bio-Scry — no prior background assumed. New to a result? Pair this with the tool guide.
A single stretch of DNA sequence off the machine. Assemblies are built by piecing many overlapping reads together.
A read tens of thousands of bases long (Oxford Nanopore, PacBio). Long reads span repeats, which is what makes complete assemblies possible.
A read a few hundred bases long (Illumina). Very accurate per base, but harder to assemble across repeats.
How many reads overlap a given position, on average. Higher coverage means a more confident consensus.
The sequencing platform and version (e.g. Nanopore R10, PacBio HiFi). It sets a read's length and error profile.
A substring of length k. Counting k-mers underpins genome-size estimation, contamination checks and finding read overlaps.
A contiguous stretch of assembled sequence built from overlapping reads. A finished bacterial chromosome is ideally one circular contig.
Contigs ordered and oriented (sometimes with gaps between them) against a reference or by linking information.
Reconstructing a genome from the reads alone, with no reference to lean on.
Mapping reads to a known reference genome and calling the differences, rather than assembling from scratch.
A contiguity metric: the length such that half the assembly sits in contigs at least that long. Bigger is more contiguous.
The number of contigs needed to cover half the assembly. Smaller is more contiguous.
The long-read assembly strategy: find how reads overlap, lay them into a graph, then vote out a consensus. Bio-Scry's Spine uses OLC.
The short-read assembly strategy: break reads into k-mers and thread them through a graph.
The agreed sequence at each position once many reads vote — averaging out individual read errors.
Detecting that a contig's two ends overlap and joining them into a true circle — as bacterial chromosomes and plasmids are.
A mis-assembly that stitches together sequences that don't belong together. A trustworthy assembler avoids fabricating these.
Using only as many reads as the genome needs — longest first — so assembly stays fast and repeats stay spannable.
Finding genes and features on an assembled genome and assigning them likely functions.
A coding sequence — a stretch of DNA translated into a protein.
Open reading frame: a run of codons with no stop, marking a candidate protein-coding region.
Structural RNAs. Ribosomal RNA genes like 16S are conserved enough to identify a species.
The fraction of G and C bases. A sharp shift in GC often marks a plasmid, island or foreign DNA.
The imbalance of G versus C along a strand; it usually flips sign at the replication origin and terminus.
Chemical marks on DNA bases (e.g. 5mC, 6mA) laid down by the cell's methyltransferases; readable from the Nanopore signal.
Promoters, ribosome-binding sites, terminators and operons — the switches that control when genes are expressed.
A predicted gene with no confident functional assignment yet.
Average nucleotide identity: the mean identity across two genomes' shared regions. ≥95% indicates the same species.
A conserved ribosomal gene used as a molecular barcode for identifying bacteria.
Multilocus sequence typing: an isolate's type read from the alleles of ~7 housekeeping genes — shorthand for its lineage.
A single-base difference from a reference — the currency of outbreak comparison.
Antimicrobial resistance: genes or mutations that let a microbe survive a drug.
A gene product that helps a microbe cause disease — toxins, adhesins, secretion systems.
A plasmid is an extra-chromosomal DNA circle; its replicon (Inc group) is how it's typed. Plasmids often carry resistance.
DNA that can move within or between genomes — prophages, transposons/IS, integrons, genomic islands.
The full gene repertoire across a set of genomes: the core (in all) plus the accessory (variable).
The ratio of amino-acid-changing to silent substitutions in a gene. Above 1 signals positive selection; below 1, purifying.
Reconstructing evolutionary relationships as a tree from sequence differences.
An estimate of how whole an assembly is, from a census of universal single-copy genes that should appear exactly once.
Sequence from a second organism mixed into the sample; flagged by odd composition or duplicated single-copy genes.
Plain text for sequences: a > header line then the bases. Used for genomes and contigs.
Like FASTA, but each read also carries a per-base quality score. The usual raw-read format.
A rich annotated-sequence format carrying features and metadata alongside the bases.
Variant Call Format: a table of SNPs and indels relative to a reference.
A table of genome features (genes, CDS, RNA) with their coordinates.
Compressed aligned reads. A modBAM also carries base-modification (methylation) tags.