Exome & targeted panels
A whole-genome library is hybridised to biotinylated baits, target fragments are pulled down on streptavidin beads, washed, amplified and sequenced; used for exomes and large panels.
Our ruleSimulations prepare you for the bench but don't replace hands-on practice.
Hybrid capture
A whole-genome library is hybridised to biotinylated baits, target fragments are pulled down on streptavidin beads, washed, amplified and sequenced; used for exomes and large panels.
How it differs from the standard method
- Scales to tens of megabases and tolerates new variants under the baits.
- Needs blockers, the bait's design temperature and the full hybridisation time.
- Libraries are often pooled before capture, so pooling accuracy sets each sample's share.
- Judged by on-target rate, duplicates, fold-80 and the fraction of target at ≥ 20×.
1Sequence only what you need
About 1–2% of a human genome codes for protein, and most clinically interpretable variants sit there. Targeted sequencing enriches those regions before sequencing, so the same number of reads gives one or two orders of magnitude more depth on them.
Enrichment efficiency
The share of reads that land in the target. Capture typically reaches 60–85%; amplicons are close to 100% but can spend reads on primer dimers.
Evenness
How much more sequencing would bring 80% of the target up to the current mean. Near 1.3–1.6 is a good capture; it rises with GC bias and poor stringency.
Complexity
Depth only counts if it comes from different original molecules. Duplicates (capture) or the number of UMI families (amplicons) say how many there were.
2Why molecules, not reads, set the limit
Every read is a copy of some original DNA molecule. Ten nanograms of human DNA holds about 3,000 copies of each locus, and in FFPE material perhaps a third of them still amplify across 150 bp.
3How it fails
- Low input: duplicates climb and sequencing deeper stops helping.
- Missing blockers or cool, low-stringency washes: adapters and repeats drag off-target DNA along and on-target rate falls.
- Washes too hot, or too many post-capture cycles: AT-rich or GC-rich exons drop out and fold-80 rises.
- A common SNP under an amplicon primer: one allele fails to amplify (allelic dropout) and a heterozygous site reads as homozygous.
- No UMIs: low-fraction variants cannot be told from damage and polymerase errors at the same site.
- Carry-over of amplified product: the no-template control shows target reads, and no low-fraction result from that run can be trusted.
Key terms
Practise these as flashcards →A biotinylated DNA or RNA oligonucleotide, typically 80–120 nt, complementary to a target region and pulled down with streptavidin beads.
Fraction of reads or bases falling in the targeted regions (sometimes counted with a small flank).
Mean target depth divided by the depth reached by 80% of target bases; a measure of capture evenness.
Share of reads that are copies of an already-seen molecule; it rises as depth approaches library complexity.
A short random barcode added to each original molecule so its reads can be grouped and collapsed into a consensus.
The lowest allele fraction detected with a stated probability (often 95%); set by molecules counted and by background error.
Failure of one allele to amplify, usually because a variant lies under a primer, so a heterozygous site looks homozygous.
Adapter-complementary oligos and repeat-rich DNA added to a hybridisation to stop off-target fragments being carried along.
When to use it
- Clinical exomes and gene panels where only coding regions or a defined gene list matter.
- Tumour hotspot panels from small or degraded FFPE biopsies, where only an amplicon approach yields enough data.
- Low-fraction mutations (subclones, residual disease, cell-free DNA) that need deep, molecule-counted data with UMIs.
- Large cohorts where genome sequencing would cost too much per sample.
Compared with alternatives
More even coverage, no enrichment bias, and non-coding, structural and copy-number variants; prefer it when the variant could be anywhere.
Absolute counts of one or a few known mutations at very low fractions; prefer it for monitoring a variant you already know.
Confirms a single variant above ~15–20% allele fraction; not sensitive enough for low-fraction tumour mutations.
Capture starts from an ordinary library, so its input, fragmentation and PCR decisions carry through.
Sub-techniques
Choosing one here switches the whole page, including the protocol and the lab, to it.
A whole-genome library is hybridised to biotinylated baits, target fragments are pulled down on streptavidin beads, washed, amplified and sequenced; used for exomes and large panels.
- Scales to tens of megabases and tolerates new variants under the baits.
- Needs blockers, the bait's design temperature and the full hybridisation time.
- Libraries are often pooled before capture, so pooling accuracy sets each sample's share.
- Judged by on-target rate, duplicates, fold-80 and the fraction of target at ≥ 20×.