Semiconductor Sequencing: Principle, Workflow, and Applications

  • Semiconductor sequencing is a next-generation sequencing (NGS) technology that determines DNA sequences by detecting changes in hydrogen ion concentration during DNA synthesis. Unlike conventional fluorescent sequencing technologies, which identify nucleotides through optical signals, semiconductor sequencing detects the chemical changes produced when DNA polymerase incorporates nucleotides into a growing DNA strand. This approach is commonly associated with the Ion Torrent sequencing technology, developed to provide a relatively simple and rapid method of high-throughput DNA sequencing.
  • The fundamental principle of semiconductor sequencing is based on the release of hydrogen ions during nucleotide incorporation. DNA polymerase adds complementary nucleotides to a growing DNA strand by forming phosphodiester bonds. Each successful nucleotide incorporation releases a hydrogen ion (H⁺), causing a small change in the pH of the surrounding solution. A semiconductor sensor detects this change and converts it into an electrical signal. By identifying the electrical signals generated during successive nucleotide flows, the sequencing instrument determines the DNA sequence.
  • Semiconductor sequencing differs from Illumina sequencing because it does not require fluorescently labelled nucleotides or optical imaging to identify incorporated bases. Illumina technology detects fluorescent signals from labelled reversible-terminator nucleotides, whereas semiconductor sequencing measures the chemical signal associated with hydrogen ion release. The absence of fluorescent labelling and optical detection simplifies some aspects of the sequencing system and allows sequencing signals to be detected directly through semiconductor sensors.
  • The most widely recognized implementation of this technology is Ion Torrent sequencing. It uses a semiconductor chip containing a large number of microscopic wells, with each well designed to accommodate a DNA template associated with a sequencing bead. The chip contains sensitive sensors that detect changes in hydrogen ion concentration during nucleotide incorporation. These electrical measurements are processed by the instrument software to determine the sequence of each DNA fragment.
  • The semiconductor sequencing workflow begins with the extraction of DNA or RNA from a biological sample. The starting material may come from human tissue, blood, plants, animals, microorganisms, cultured cells, or environmental samples. For DNA sequencing, genomic DNA is isolated and assessed for concentration, purity, and integrity. For RNA sequencing, RNA is extracted and generally converted into complementary DNA (cDNA) before library preparation. The quality of the starting material is important because contamination, degradation, or insufficient nucleic acid quantity can affect library construction and sequencing performance.
  • The next step is library preparation, during which DNA is converted into fragments suitable for sequencing. Genomic DNA is fragmented into smaller pieces using mechanical or enzymatic methods. The resulting fragments are modified by adding adapter sequences to their ends. These adapters contain sequences required for subsequent library processing, amplification, and sequencing. Sample-specific barcodes may also be incorporated so that multiple samples can be pooled and distinguished during bioinformatics analysis.
  • After adapter addition, the library may undergo amplification and purification. The precise preparation procedure depends on the sequencing system and library kit. The resulting DNA fragments are then used to prepare template-positive beads. In Ion Torrent workflows, this involves attaching DNA library molecules to microscopic beads and amplifying them through emulsion PCR. The amplification process generates many copies of an individual DNA template on each bead.
  • Emulsion PCR is a form of polymerase chain reaction performed within microscopic water-in-oil droplets. Each droplet acts as an isolated reaction compartment and ideally contains a bead associated with a single DNA template molecule, together with the reagents required for amplification. The DNA fragment is amplified within the droplet, producing many copies of the same template on the bead. This amplification generates sufficient DNA material for subsequent detection during sequencing.
  • Following emulsion PCR, the beads are recovered and processed to remove unwanted material. Beads carrying amplified DNA templates are enriched and prepared for loading onto the semiconductor chip. The enrichment step helps remove empty beads that do not carry useful amplified DNA. The prepared beads are then distributed into the microscopic wells of the sequencing chip. Ideally, each well contains one template-positive bead, although loading efficiency and the presence of empty or multiple-bead wells can influence sequencing performance.
  • The semiconductor chip is a central component of the technology. It contains a large array of wells, each associated with a sensitive ion-detection sensor. These sensors detect changes in the chemical environment when nucleotides are incorporated into DNA. Unlike conventional optical sequencing systems, the chip does not require a camera to detect fluorescent emissions from each DNA cluster. Instead, it converts chemical changes into measurable electrical signals.
  • During sequencing, the instrument introduces one type of nucleotide at a time into the chip. For example, the first flow may contain adenine nucleotides, followed by cytosine, guanine, and thymine in a programmed sequence. If the nucleotide supplied during a particular flow is complementary to the next unpaired base in the template, DNA polymerase incorporates it into the growing strand. This incorporation releases hydrogen ions, producing a local pH change that is detected by the semiconductor sensor.
  • If the supplied nucleotide is not complementary to the next template base, incorporation does not occur and no corresponding incorporation signal is generated. The instrument records the signal from each well during the nucleotide flow and then washes away the remaining nucleotides before introducing the next nucleotide type. This cycle is repeated until the sequencing reaction is completed.
  • The intensity of the electrical signal provides information about how many identical nucleotides were incorporated during a particular flow. For example, if the template requires three consecutive adenine bases, the introduction of adenine may result in three nucleotide incorporations and a stronger signal than a flow involving only one adenine incorporation. This allows the instrument to identify homopolymer sequences, which are stretches containing repeated copies of the same nucleotide.
  • However, estimating the number of bases in a homopolymer can be challenging. The electrical signal does not always increase in a perfectly linear manner with the number of incorporated nucleotides. Signal variability, chemical conditions, sensor characteristics, and incomplete or uneven incorporation can affect the measured intensity. Consequently, semiconductor sequencing systems may have difficulty distinguishing between homopolymer lengths, particularly when several identical bases occur consecutively. This is one of the important technical limitations of the technology.
  • The electrical signals generated during nucleotide incorporation are processed by instrument software. Signal processing includes correcting background measurements, estimating incorporation levels, and converting electrical responses into nucleotide sequences. This process is known as base calling. The resulting sequences are stored as reads, generally in FASTQ format, together with quality scores that estimate the reliability of individual base calls.
  • A simplified FASTQ record is shown below:

@READ_001

ATGCTAGCTAGCTACGATCG

+

IIIIIIIIIIIIIIIIIIII

  • The first line contains the read identifier, the second contains the nucleotide sequence, the third is a separator, and the fourth contains encoded quality values. Actual read identifiers and quality strings depend on the sequencing platform and software. Quality scores are commonly expressed using the Phred scale, calculated as Q = −10 log10(P), where P is the estimated probability of an incorrect base call. A Q20 score corresponds to an estimated error probability of 1%, while Q30 corresponds to an estimated error probability of 0.1%.
  • After sequencing, raw reads undergo bioinformatics quality control. The quality-control process examines read length, base quality, GC content, adapter contamination, sequence composition, and other characteristics. FastQC can generate reports for individual FASTQ files, while MultiQC can combine reports from multiple samples. fastp can perform adapter trimming and quality filtering for supported data types. Additional platform-specific processing may be required depending on the instrument and the format of the raw data.
  • An example of quality assessment using FastQC is:

fastqc sample.fastq.gz \

  –outdir fastqc_results

MultiQC can summarize the resulting reports:

multiqc fastqc_results \

  –outdir multiqc_report

  • When necessary, adapter sequences and low-quality regions can be removed before downstream analysis. However, trimming parameters should be selected carefully because excessive trimming may shorten reads, reduce coverage, and remove useful sequence information. The preprocessing strategy should reflect the quality characteristics of the dataset and the requirements of the intended analysis.
  • Following quality control, semiconductor sequencing reads can be aligned to a reference genome or used for de novo assembly. Reference-based alignment identifies the genomic locations corresponding to individual reads. Tools such as BWA-MEM2 and Bowtie2 can be used for suitable DNA read datasets, although aligner selection should consider the read characteristics and the sequencing technology. For RNA sequencing, splice-aware aligners such as STAR and HISAT2 may be appropriate.
  • An illustrative alignment command using BWA-MEM2 is:

bwa-mem2 mem \

  reference_genome.fa \

  sample.fastq.gz \

  > sample.sam

  • The resulting SAM file contains information about read alignments, including genomic positions, alignment flags, mapping quality, and other attributes. SAMtools can convert the SAM file into BAM format, sort the alignments, and create an index:

samtools view -bS sample.sam > sample.bam

samtools sort sample.bam -o sample.sorted.bam

samtools index sample.sorted.bam

  • The sorted and indexed alignment file can be used for coverage assessment, variant calling, and other downstream analyses. Reference selection is important because differences in genome assemblies and annotation versions can affect read mapping and the interpretation of genomic coordinates.
  • Variant calling is one of the major applications of semiconductor sequencing. It involves identifying differences between the sequenced sample and a reference genome. These differences may include single-nucleotide variants (SNVs), small insertions and deletions (indels), and certain other forms of genetic variation. Variant callers examine evidence from aligned reads, including base quality, mapping quality, read depth, and allele balance, to estimate whether a difference represents a genuine variant.
  • Tools such as Torrent Variant Caller, bcftools, GATK, and FreeBayes may be used in suitable workflows, depending on the sequencing platform, data characteristics, and analysis objective. Platform-specific variant-calling software can incorporate information about the signal characteristics and error profiles of the technology. The resulting variant information is commonly stored in VCF format. Annotation tools such as Ensembl VEP, SnpEff, and ANNOVAR can help identify the possible effects of variants on genes, transcripts, and proteins.
  • Variant calling requires particular care when analysing homopolymer regions because errors in estimating the number of repeated bases can produce false insertions or deletions. Quality filtering, platform-aware variant-calling methods, read-level inspection, and validation using an independent method may be appropriate for important findings. A detected sequence difference should not automatically be interpreted as a disease-causing mutation or a functionally important change without supporting biological evidence.
  • Semiconductor sequencing has been used in targeted gene sequencing, microbial genome analysis, infectious disease research, cancer genomics, and other applications. Targeted sequencing is particularly suitable because a selected group of genes or genomic regions can be investigated at relatively high coverage. In clinical research, targeted panels may be used to examine genes associated with inherited disorders or cancer-related pathways. The clinical suitability of a test depends on its validated performance, the types of variants it can detect, and the relevant regulatory and laboratory requirements.
  • In microbiology, semiconductor sequencing can be used to investigate bacterial and viral genomes, identify genetic differences, and study antimicrobial resistance-associated mutations. It can also support pathogen characterization and molecular epidemiology. However, the interpretation of pathogen sequence data requires appropriate reference databases, quality control, and epidemiological information. Genetic similarity alone does not establish direct transmission between individuals or locations.
  • The technology has also been used in transcriptomic research. In RNA sequencing applications, extracted RNA is converted into cDNA and prepared as a sequencing library. The resulting reads can be used to estimate gene expression, investigate transcript composition, and identify selected RNA sequence features. As with other short-read RNA-sequencing approaches, transcript abundance estimates depend on library preparation, sequencing depth, mapping accuracy, and appropriate statistical analysis.
  • Semiconductor sequencing offers several advantages. It detects nucleotide incorporation through direct electrical measurements rather than fluorescent labelling and optical imaging. This reduces the need for complex optical detection components and supports a relatively compact sequencing system. The technology can provide relatively rapid sequencing workflows and has been used for targeted sequencing and selected applications requiring moderate to high sequencing throughput. Its library preparation and sequencing procedures can be adapted to different research and diagnostic applications, depending on the platform and assay.
  • Despite these advantages, semiconductor sequencing has important limitations. The most recognized limitation is reduced accuracy in homopolymer regions, where repeated identical nucleotides can generate signals that are difficult to distinguish precisely. This can lead to insertion and deletion errors. The technology may also be affected by uneven signal responses, sequencing depth variation, library preparation bias, and other platform-specific error patterns.
  • Another limitation is that semiconductor sequencing systems generally produce short reads compared with modern long-read technologies. Short reads can make it difficult to resolve long repetitive regions, complex structural variations, and highly similar genomic regions. Although paired-end information and appropriate computational methods can improve some analyses, they do not completely overcome the limitations of short-read sequencing. Long-read technologies such as PacBio HiFi and Oxford Nanopore sequencing may be more suitable when long-range genomic information is required.
  • Semiconductor sequencing should also be distinguished from other sequencing technologies. Illumina sequencing uses fluorescently labelled reversible-terminator nucleotides and optical detection, while semiconductor sequencing measures pH changes associated with nucleotide incorporation. Sanger sequencing uses dideoxynucleotide-mediated chain termination and fragment separation. PacBio sequencing detects nucleotide incorporation in individual DNA molecules through changes in fluorescence, while Oxford Nanopore sequencing identifies DNA or RNA molecules through changes in ionic current as they pass through nanopores. These technologies differ in their sequencing chemistry, signal detection, read lengths, throughput, error profiles, and applications.
  • For reproducible semiconductor sequencing analysis, researchers should record the sample source, DNA or RNA extraction method, library preparation protocol, adapter and barcode design, sequencing instrument, chip type, read configuration, and sequencing depth. They should retain raw sequencing files, quality-control reports, reference genome accession and version, software versions, and analysis parameters. Because sequencing platforms can have different error profiles, the analytical pipeline should be appropriate for the technology used. Important variants, especially those located in difficult sequence contexts, may require confirmation using an independent method.
  • Semiconductor sequencing represents an important development in next-generation sequencing because it uses the chemical properties of DNA synthesis to generate electrical signals that can be converted into nucleotide sequences. Its reliance on hydrogen ion detection rather than fluorescent labelling distinguishes it from optical sequencing technologies such as Illumina. Although challenges such as homopolymer-associated insertion and deletion errors and short read lengths limit some applications, semiconductor sequencing has contributed to targeted genomics, microbial research, molecular diagnostics development, and other areas of bioinformatics. Understanding its principles, workflow, signal detection, data analysis, applications, and limitations provides a broader understanding of the different technologies used in modern DNA sequencing.
Author:

Leave a Reply

Your email address will not be published. Required fields are marked *