Foundations

A local alignment score is evaluated against an explicit random-sequence null and search space.

HaplotypeCaller assembles candidate haplotypes in active regions and compares reads with them through PairHMM likelihoods before genotype inference.

Phylogenetic substitution models differ in equilibrium frequencies, transition rates and among-site mixtures.

Bayesian phylogenetic analysis combines an alignment likelihood with substitution, clock and tree priors and samples the posterior through MCMC.

Branch lengths in substitutions must be linked to calendar time by a clock model and calibrating information.

Under a multispecies coalescent, gene lineages can fail to coalesce between species divergences.

The 2025 study uses long-read sequencing and linear/graph analyses across 1,019 sampled humans to expand structural-variant and tandem-repeat characterization.

Flux balance imposes steady-state stoichiometric constraints and bounds on reaction rates, then selects fluxes using an objective or additional constraints.

Isotope-label distributions constrain fluxes through a specified metabolic network and atom-mapping model.

A model is structurally unidentifiable when distinct parameter values cannot be distinguished even with ideal observations; practical unidentifiability can remain because actual observations are limited.

Guide-linked single-cell perturbation screens measure transcriptomic responses in surviving sampled cells.

Comparing single and joint perturbations tests interaction relative to an explicitly chosen no-interaction expectation.

A delayed response can be explained by feedback, slow transport, state heterogeneity or an unobserved intermediate.

Jinek and colleagues demonstrated RNA-programmable Cas9 double-stranded DNA cleavage in vitro and a fused guide-RNA architecture.

Cytidine base editors couple a deaminase with DNA-targeting Cas9 machinery; UGI and nicking modifications bias processing toward desired substitutions.

Prime editing combines a Cas9 nickase–reverse-transcriptase fusion with a pegRNA carrying the desired edit.

An intended amplicon edit fraction describes the assayed surviving population and locus.

WHO’s human-genome-editing framework distinguishes somatic, germline and heritable applications and calls for oversight, transparency and responsible governance.

MINFLUX uses spatially positioned excitation minima to infer a fluorescent emitter’s location from detected photons.

A fitted relation between metabolism and body mass depends on basal-state criteria, temperature, species sampling and regression treatment.

Mechanistic niche models combine microclimate with organismal heat, water and nutritional budgets across a life cycle.

Cells sampled from one animal or culture share experimental history.

CITE-seq joins antibody-derived epitope tags with transcriptome profiling in the same sampled cell.

Integration attempts to align datasets while preserving biological differences.

Spatial transcriptomics retains coordinates but its observable depends on platform: capture-based spots can mix cells, while imaging-based panels sample selected transcripts.

Ordering cells along a transcriptional trajectory describes similarities among sampled states.

RNA velocity uses spliced/unspliced RNA relationships to estimate a local expression-change direction under a transcription/splicing/degradation model.

Spatial Perturb-Seq combines AAV-delivered genetic perturbation barcodes with spatial transcriptomic readouts in the studied Cas9 mouse hippocampus.

Whether a perturbation-derived regulatory edge persists across tissue, developmental stage or environment is a separate question from detecting a response in one experiment.

Conformational selection and induced fit describe competing temporal paths between conformation change and ligand binding.