Written and maintained by CASRAI Editorial Board
Last updated
Short answer
In silico PCR (also written PCR in silico, and often called virtual PCR) is a computer search that takes a pair of primer sequences and reports where, in a chosen DNA or RNA sequence database, they could anneal and which product they would amplify. It predicts target location and amplicon size without touching a thermal cycler. A predicted amplicon is a hypothesis to test at the bench, not a validated assay.
Definition and scope: what in silico PCR is and is not
A 2024 review in Frontiers in Bioinformatics describes the method this way: “In silico PCR tests the target location and amplicon size in one or more DNA templates.” The same review lists uses wherever primer or probe specificity matters, including gene discovery, molecular diagnostics, DNA profiling and repeat identification.
It is a search and prediction step. It does not model reaction chemistry, enzyme performance or inhibitors, and it does not replace running the reaction. It sits before ordering oligonucleotides (to see whether a pair hits only the intended locus) and after a failed or messy gel (to see whether a second predicted product explains an extra band). For the design side, see the guide to choosing primers for PCR.
How a run works: inputs and outputs
Most tools take the same core inputs: a forward primer, a reverse primer, a target database, and limits on how imperfect a match may be and how large a product may be. They return the predicted product positions and sequences.
The UCSC In-Silico PCR page states that it “searches a sequence database with a pair of PCR primers, using the BLAT index for fast performance.” Its output is a FASTA file containing the sequence between and including the primer pair. The header gives the coordinates with a strand sign (for example chr22:31000551+31001000), and the body is uppercase where the primer sequence matches the database and lowercase elsewhere. NCBI Primer-BLAST screens primers “against user-selected database” using BLAST and a global alignment algorithm, and checks all primer-pair combinations: forward-reverse, forward-forward and reverse-reverse.
To check a spliced transcript, UCSC lets you change the target to a gene transcript set (human and mouse assemblies only), which is how an RT-PCR design across an intron-exon boundary is tested against the mRNA rather than the genome. For a worked Primer-BLAST walkthrough, see using NCBI BLAST for primers; this page does not repeat those steps.
UCSC In-Silico PCR vs NCBI Primer-BLAST: what each run asks for
Each row below comes from the tool’s own page, checked 2026-10-29.
| Setting | UCSC In-Silico PCR | NCBI Primer-BLAST |
|---|---|---|
| Search space | Genome and assembly menu; transcript set for human and mouse RT-PCR checks | Database menu (RefSeq mRNA, RefSeq reference genomes, core_nt, nt, custom and others) |
| Product size limit | Maximum product size: “Maximum size of amplified region” | Max target amplicon size: the largest amplicon size for a PCR target to be detected |
| Mismatch rule | Min perfect match: bases that must match exactly at the 3′ end (minimum 15). Min good match: at the 3′ end, at least 2 of every 3 bases must match | Stringency: minimum total mismatches to unintended targets plus minimum mismatches within the last 1 to 22 bases at the 3′ end |
| Reverse primer entry | “Flip Reverse Primer” inverts and complements the sequence | Checks reverse-reverse and forward-forward combinations as well as forward-reverse |
One database detail matters for genome work: the Primer-BLAST page says core_nt “is the same as nt except that it does not have the eukaryotic chromosomal sequences”, so a core_nt search is not a whole-genome specificity check for a eukaryote.
What a hit does and does not mean
Two statements in the primary sources set the limits. UCSC says its tool “is not guaranteed to find absolutely all off-target locations”, and suggests extra validation with tools such as Primer-BLAST, especially for repetitive regions. The Frontiers review notes that BLAST-based approaches cannot search for pairs of queries separated by arbitrary sequences of variable length, and that “parameters sensitive enough to find an acceptable number of matches tend to generate large numbers of false positives”. In the authors’ own test with retrotransposon-derived primers, the amplicons identified in the human genome and in one other tested genome were described as most likely false positives.
Read together, a clean result (one expected product) lowers risk but does not prove specificity, and a long list of hits may reflect search sensitivity rather than real priming. Confirming specificity in the wet lab, for example with a melt curve, is still the test; see qPCR melt curve analysis.
Checklist: in silico PCR before ordering primers
- Pick a search space that matches the template: a genome for genomic DNA, a transcript set or RefSeq mRNA for cDNA from spliced RNA (UCSC and Primer-BLAST both offer these).
- Check which database you selected is actually covering your organism’s chromosomal sequence (see the core_nt note above).
- Set the maximum product size near what you expect, so a distant coincidental pair is not reported as a product.
- Decide the 3′ end rule on purpose: UCSC’s min perfect match and good match, or Primer-BLAST’s mismatches in the last 1 to 22 bases, are the settings that most change what is reported.
- Run a second tool on repetitive targets, as UCSC itself suggests.
- Keep a note of the tool, database or assembly, settings and run date alongside the primer order (a practice suggestion, not a requirement stated by either tool).
Common misunderstandings
“In silico PCR proves my primers are specific.”
It predicts specificity against one database, under the settings you chose. Both UCSC and the review quoted above state limits on completeness and false positives.
“The tool simulates the reaction.”
The tools described here match primer sequences to a database and report product location and size. Annealing temperature, enzyme, buffer and template quality are outside what they report. A thermal cycler run is the reaction itself.
“All in silico PCR tools give the same answer.”
They use different engines (BLAT index for UCSC, BLAST plus global alignment for Primer-BLAST) and different mismatch settings, so the same pair can produce different hit lists.
FAQ
What is in silico PCR?
A software search that predicts where a primer pair would anneal in a sequence database and what product it would give, without running a wet-lab reaction.
Is “PCR in silico” the same thing as virtual PCR?
Yes. “In silico PCR” and “PCR in silico” are two word orders for the same idea: computing the amplification result from sequences. “Virtual PCR” is commonly used for it too. The 2024 review names Electronic PCR, UCSC In-Silico PCR, Primer-BLAST, WebPCR and FastPCR among the tools in this space.
Which tool should I start with?
UCSC In-Silico PCR and NCBI Primer-BLAST are both free web tools with their own documentation. The table above shows what each asks for; UCSC itself suggests a second tool for primer design in repetitive regions.
Does it replace running a gel or melt curve?
No. A predicted product is a prediction; specificity is confirmed on the bench.
Sources
All pages checked 2026-10-29.
- UCSC Genome Browser, In-Silico PCR: https://genome.ucsc.edu/cgi-bin/hgPcr
- NCBI, Primer-BLAST: https://www.ncbi.nlm.nih.gov/tools/primer-blast/
- NCBI, Primer-BLAST information page: https://www.ncbi.nlm.nih.gov/tools/primer-blast/primerinfo.html
- Kalendar R, Shevtsov A, Otarbay Z, Ismailova A. In silico PCR analysis: a comprehensive bioinformatics tool for enhancing nucleic acid amplification assays. Frontiers in Bioinformatics. 2024;4:1464197. https://pmc.ncbi.nlm.nih.gov/articles/PMC11491563/
