Basiskalling

Basiskalling, also known as base calling, is the computational process of converting raw signals produced by DNA sequencing instruments into a string of nucleotide bases (A, C, G, T, and sometimes N for unknown). It is a central step in most sequencing workflows, including widely used platforms such as Illumina, Ion Torrent, PacBio, and Oxford Nanopore.

The process typically involves interpreting instrument-specific signals—such as fluorescence intensities or ionic currents—at each sequencing cycle and assigning the most probable base for that position. Modern base-calling methods rely on statistical models, probabilistic frameworks, or machine learning approaches to estimate base identity and to attach a quality score that reflects confidence in the call. Preprocessing steps may include signal normalization, background subtraction, and correction for systematic biases.

The primary output of basiskalling is reads in a standard format (often accompanied by per-base quality scores, as in FASTQ files). These reads are then used in downstream analyses such as alignment to a reference genome, variant discovery, transcriptome profiling, and genome assembly.

Accuracy in basiskalling depends on factors like signal-to-noise ratio, instrument calibration, chemistry, read length, and platform-specific biases or errors (for example, difficulties with homopolymers on some platforms). Quality control and error correction are common follow-up steps to improve downstream results.

See also: base calling in sequencing, sequencing quality scores, next-generation sequencing, sequencing platforms, FASTQ formats.