Definitions of frequently used terms
| Term | Definition |
|---|---|
| Sequence libraries | A sample of DNA that has been processed to be sequenced. |
| Reads | Fragmented and overlapping pieces of a DNA strand that are sequenced. |
| Phred quality scores | A scaled probability that a given inference (usually a called base) is incorrect. The probabilty of error, P(e), scaled by -10 log P(e). |
| Short reads | Reads from first and second generation sequencing such as Sanger, Illumina, IonTorrent, etc. Short reads can range from 30-1000bp long. |
| Read pair | Many short read sequencing technologies sequence from both ends of a DNA fragment, resulting in a pair of sequenced reads that come from said fragment. |
| Adapter | A short piece of DNA that is ligated to the short fragment to be sequenced. The adapter allows the fragment to be affixed to a physical medium (such as a flow cell) to facilitate amplification and sequencing. |
| Insert size | The size of the DNA fragment between the adapter sequences. |
| Mate pairs | Long-insert paired end reads prepared by circularizing longer DNA fragments. |
| Jumping libraries | Junction-fragment libraries. Mate pair libraries. |
| Long reads | Reads from single-molecule sequencing technology such as PacBio SMRT and Oxford Nanopore. Long reads can range from 1000-100000+bp long. |
| Genome assembly, Assembly, de novo Assembly | 1. The process by which small overlapping parts of the genome are reconstructed into longer contiguous sequences, 2. A sequence that has undergone the assembly process. |
| Contigs | Assembled reads. Contig assembly is usually done with a graph-based representation (i.e. de Bruijn graphs) of overlapping sequence reads. |
| Scaffolds | Contigs that have been joined together to form longer sequences. Scaffolding is usually done using read pair information or long reads. |
| Reference genome | An already assembled genome to which you can compare newly sequenced reads or genomes. |
| Read mapping | The process of aligning reads from a newly sequenced genome to a reference genome |
| Mapping quality | A usually Phred scaled probability that a given read has mapped incorrectly. |
| Reference-guided assembly | 1. The process of using read mapping to reconstruct the genome from a set of reads, 2. A sequence that has undergone the reference-guided assembly process. |
| Reference bias | The phenomenon of a set of mapped reads appearing to resembe (through lower divergence) the reference genome more closely than they actually do because reads containing the most variation were not mapped. |
| Iterative mapping | The process of mapping reads to a reference genome, generating a reference-guided assembly, and then repeating the process this time mapping to the new reference guided assembly. Done to reduce reference bias. |
File formats
| Format | Use | Link | Specs |
|---|---|---|---|
| FASTA | Stores sequence data. | Wikipedia | NA |
| FASTQ | Stores sequence data and quality scores. | Wikipedia | Link |
| SAM | Sequence Alignment Map format. Stores information about reads mapped to a reference genome. | Wikipedia | Link |
| BAM | Binary Alignment Map format. The compressed binary version of SAM format. | Wikipedia | Link |
| CRAM | Another compressed format to store read mapping information. | Wikipedia | Link |
| VCF | Variant Call Format. Used to store information about variants inferred for a given sample(s). | Wikipedia | Link |
| BCF | Binary variant Call Format. The binary compressed verion of a VCF. | Wikipedia | Link |
| BED | Stores coordinates of regions of interest | Wikipedia | Link |