# How to Do PCR: A Clear Bench Workflow

Source: https://www.verbalexperiment.com/blog/how-to-do-pcr
Published by Verbex. How these guides are written: https://www.verbalexperiment.com/how-we-write

A gel shows no band. A real-time run shows no signal. The first impulse is often to change the enzyme, primers and temperature at once. That response hides the cause. Learning how to do PCR means running a controlled test, reading the pattern of those controls, and changing one thing at a time. It also means recording what happened while the details are still visible.

## Table of Contents

-   [Starting Settings for a First PCR Program](#starting-settings-for-a-first-pcr-program)
    -   [A practical starting card](#a-practical-starting-card)
    -   [Why more rounds are not always better](#why-more-rounds-are-not-always-better)
-   [Check the Template Before You Pipette](#check-the-template-before-you-pipette)
-   [Choose Equipment That Fits the Question](#choose-equipment-that-fits-the-question)
-   [How to Run PCR from Setup to Readout](#how-to-run-pcr-from-setup-to-readout)
    -   [Prepare the reaction](#prepare-the-reaction)
    -   [Load and run the machine](#load-and-run-the-machine)
    -   [Read the result](#read-the-result)
-   [Controls and Failed-Run Diagnosis](#controls-and-failed-run-diagnosis)
    -   [Controls that belong in every run](#controls-that-belong-in-every-run)
    -   [Read the pattern before changing reagents](#read-the-pattern-before-changing-reagents)
-   [PCR Types and What a Good Result Shows](#pcr-types-and-what-a-good-result-shows)
    -   [What each readout means](#what-each-readout-means)
-   [Record PCR Observations Before Memory Fades](#record-pcr-observations-before-memory-fades)
    -   [Fields to finish at the bench](#fields-to-finish-at-the-bench)
    -   [Capture the run as it happens](#capture-the-run-as-it-happens)

## Starting Settings for a First PCR Program

**PCR** means polymerase chain reaction. It makes many copies of one chosen stretch of **DNA**, the molecule that carries genetic instructions. The tube is heated so the two DNA strands separate. It is cooled so short matching pieces called **primers** can attach to the target. It is warmed again so an **enzyme**, a protein that speeds a chemical reaction, builds a new matching strand. With each round, every finished copy becomes a template for the next one, so the target amount multiplies while reagents last.

These three stages have technical names. **Denaturation** is the heating step that separates the strands. **Annealing** is the cooling step where primers attach. **Extension** is the step where the enzyme builds new DNA. The copied stretch is called the **amplicon**.

### A practical starting card

The [Thermo Fisher Phusion PCR protocol](https://www.thermofisher.com/us/en/home/references/protocols/nucleic-acid-amplification-and-expression-profiling/pcr-protocol/phusion-pcr-protocol.html) gives **25 to 35 amplification cycles**, denaturation at **approximately 98°C**, annealing at **about 60°C**, and extension at **72°C**. Its program starts with **98°C for 30 seconds**, then uses **5 to 10 seconds** of denaturation, **10 seconds** of annealing, and **15 to 30 seconds of extension per kilobase** in each cycle, followed by **72°C for about 5 minutes**.

A **kilobase** is 1,000 DNA bases, which are the letterlike units in DNA. A 25 microliter reaction is a practical default volume. A commercial mix already contains the enzyme, **buffer** and DNA building blocks. Buffer is the liquid that keeps chemical conditions stable. The label still controls the final amounts.

The card below uses that protocol for its starting column. The adjustment column gives typical ranges for standard enzyme kits, where labels commonly specify lower denaturation temperatures and more cycles.

| Setting | Typical start | When it changes |
| --- | --- | --- |
| Reaction volume | 25 microliters | Keep concentrations the same if the volume is scaled |
| Each primer | 0.5 micromolar, typical range 0.1 to 1.0 | Lower end of the range for targets longer than 5 kilobases or several targets in one tube |
| First heating step | 98°C for 30 seconds | Often 95°C for standard enzyme kits |
| Heating step in each round | 98°C for 5 to 10 seconds | Often 95°C for standard enzyme kits |
| Primer binding in each round | About 60°C for 10 seconds | Typical range 55 to 65°C, about 5°C below the primer melting temperature |
| Copying step in each round | 72°C for 15 to 30 seconds per kilobase | Longer amplicons need longer copying time, about 15 seconds per kilobase for simple DNA |
| Number of rounds | 25 to 35 | Typical ranges are 20 to 25 for simple DNA, 30 to 35 for genomic DNA, and 35 to 40 for standard enzyme kits |
| Final copying step | 72°C for 5 minutes | Run before cooling, gel checks, or short storage |

**Micromolar** is a concentration unit. **Genomic DNA** is the full DNA set from an organism. **Melting temperature** is the temperature where half the primer molecules are bound and half are free.

### Why more rounds are not always better

Extra rounds can help when template DNA is scarce. Template is the starting DNA being copied. Extra rounds can also create unwanted products and copying errors called **artifacts**. A weak result calls for reading the controls first. The pattern across the positive control, the negative control and the sample points to whether the problem sits in the reagents, the template or the program, and only then should a setting change.

> **Starting rule:** Use the card as a first test, then adjust from what the controls show.

The final program depends on enzyme chemistry, primer melting temperature, amplicon length, template quality and machine calibration. Calibration means checking that a machine reaches the temperatures it displays. Add DNA last. Fill the remaining volume with **nuclease-free water**, which is water certified free of enzymes that break DNA. Record the program exactly as it ran, including any change from the card, because those notes are part of whether the result can be trusted later.

## Check the Template Before You Pipette

Many failed reactions trace back to the DNA or **RNA** added to the tube. RNA is a related molecule that carries copies of genetic instructions. Checking the template first separates a sample problem from a setup problem.

Record the sample source, extraction method and date. Extraction is the process that removes DNA or RNA from cells or tissue. Measure concentration with a **spectrophotometer**, an instrument that measures light absorbed by a sample, or with a fluorescent dye reader. A fluorescent dye glows when it binds the target molecule.

Then check purity using absorbance ratios at 260 and 280 nanometers and at 260 and 230 nanometers. A nanometer is a unit of light wavelength. Compare the ratios with the values listed by the kit or reader manufacturer, since unusual ratios can signal leftover protein or extraction chemicals.

-   **DNA or RNA:** RNA must first be copied into DNA before PCR.
-   **Storage:** Keep a small working portion refrigerated, freeze the rest, and avoid repeated freezing and thawing.
-   **Amount:** Start near the middle of the range printed on the enzyme or kit label.
-   **Dilution test:** Prepare a series of weaker sample dilutions to test whether something in the sample blocks the enzyme.

Primer choice also depends on knowing the intended target, and this explanation of [what a PCR primer is](https://www.verbalexperiment.com/blog/what-is-a-pcr-primer) clarifies that matching step. Carry every measurement into the lab record. A later reviewer can then tell whether a weak band reflects poor template or faulty chemistry.

## Choose Equipment That Fits the Question

The right machine depends on the question. Presence, size, amount and exact molecule count aren't the same measurement. Modern equipment helps, but it doesn't replace controls.

A **real-time PCR instrument** reads fluorescence during every round. Fluorescence is light emitted by a dye or probe. This format suits amount-based questions and often replaces a gel for those questions. A gel is a slab that separates DNA pieces by size.

A **gradient thermal cycler** runs several annealing temperatures across one plate. A thermal cycler is the temperature-controlled machine that runs the PCR program. The gradient speeds primer testing because several temperatures can be compared in one run.

A heated lid reduces condensation inside tubes. Sealed plates or caps limit evaporation and leakage between wells. A **digital PCR system** splits a sample into thousands of tiny partitions and counts the positive partitions, which suits very small amounts and samples that interfere with standard reactions.

> The equipment choice should follow the question and the controls, not replace either one.

Modern setup also uses ready-made **master mixes** and **hot-start enzymes**. A master mix combines shared ingredients in one tube. A hot-start enzyme stays inactive until heated, which reduces copying before the program begins.

Calibrated pipettes, a PCR-only hood or clean bench, and ultraviolet light or chemical cleaning reduce avoidable errors. Save the plate map with the run file. Record baseline and threshold settings because those software choices change reported values.

## How to Run PCR from Setup to Readout

The diagram below condenses the workflow into five stages. The numbered steps that follow place each bench action in context.

![A five-step instructional diagram illustrating the workflow of conducting a PCR experiment from setup to final analysis.](https://cdnimg.co/cab61d54-2a92-4be8-97b1-fd84150a185b/61a281f1-8717-41fb-9e0d-66009e1f91e5/how-to-do-pcr-pcr-process.jpg)

### Prepare the reaction

1.  **Write the question first.** Decide whether the run must show presence and size or estimate the amount of DNA or RNA. That choice determines the readout and controls.
2.  **Choose primers for one product.** Check the intended match in a primer-design tool before ordering. A primer that matches several places can produce several products.
3.  **Thaw reagents on ice.** Mix each tube by gentle flicking and spin droplets down. Cold handling protects the enzyme before the run starts.
4.  **Make one master mix.** Include enzyme, buffer, **magnesium**, DNA building blocks, primers and water for every tube, including controls. Magnesium is a mineral ion that helps the enzyme work.
5.  **Portion the mix before any template tube is opened.** Dividing the shared mix first keeps one sample from reaching every tube. Filtered tips reduce the tiny droplets carried between tubes.

Volumes for the shared mix can be checked with a [PCR master mix calculator](https://www.verbalexperiment.com/tools/pcr-master-mix-calculator) before pipetting begins. A deterministic calculator checks its own arithmetic, yet it cannot know whether the entered values suit the experiment. The scientist still checks the units, stock concentrations and final concentrations.

### Load and run the machine

6.  **Label the plate map. Check the thermal cycler.** Cap or seal every tube or plate before loading. A real-time run needs an optically clear plate or seal so the instrument can read fluorescence.
7.  **Run the chosen program.** Keep every tube closed while the program runs, since opening one releases copied DNA into the setup area. Watch for instrument warnings or a stalled cycle, and write down anything unusual as it happens. A note made during the run later explains missing signal far better than a guess made after it.

The short video below shows the physical rhythm of setup, loading and readout before the interpretation rules are applied.

### Read the result

A standard reaction is usually checked on a gel for a band of the expected size. A band is a visible group of DNA pieces of similar length. Compare it with a **ladder**, a set of DNA pieces of known sizes.

A real-time reaction is read from its amplification curve, which shows signal rising across cycles. A **melt curve** shows how the copied product separates as temperature rises. One clean melt peak supports one main product when a dye-based method is used.

The common slips leave readable patterns. A missing ingredient in the shared mix weakens every tube, including the positive control. Template carried into the blank produces signal where none should appear. A plate sealed after warming can leak or evaporate unevenly, so some wells hold less volume than the rest.

> **Usable first result:** The known positive works, the blank stays blank, and the sample matches the expected size or curve.

That pattern does not prove every biological claim. It shows the run worked well enough for the controls to guide interpretation, and a different pattern across the same controls shows where to look first.

## Controls and Failed-Run Diagnosis

A run finishes with a faint band in the blank and nothing in two of the samples. Before any reagent gets swapped, the controls have already told a story. Controls supply the internal evidence that shows whether a sample result can be believed, and their pattern is the first diagnostic tool a bench scientist should read.

![An infographic titled Controls and Failed-Run Diagnosis illustrating laboratory best practices for PCR experiment quality control.](https://cdnimg.co/cab61d54-2a92-4be8-97b1-fd84150a185b/6adf4445-cfc2-466d-ac93-aa72baabdf8e/how-to-do-pcr-laboratory-controls.jpg)

### Controls that belong in every run

A **positive control** contains the target. A **no-template control**, also called a blank, contains every reagent except template. Load the blank last so sample DNA isn't carried into it during setup.

For RNA work, a **no-reverse-transcription control** skips the step that copies RNA into DNA. Signal in that tube points toward leftover genomic DNA rather than RNA.

For a 40-cycle assay, the [CDC real-time PCR identification guidance](https://www.cdc.gov/candida-auris/hcp/laboratories/real-time-pcr-identification.html) gives an operational benchmark: a strong positive control around **Cq 22 to 28** and a lower positive control around **Cq 30 to 36**. **Cq** is the cycle where signal rises above a set background line. The same guidance treats the no-template control as a mechanism for detecting cross-contamination, recommends positive and negative controls for each reaction set, and advises cleaning benches and pipettes with a DNA-removal product followed by **70% ethanol**. Single-use aliquots, which are small portions taken from a stock tube, keep one contaminated portion from reaching the rest.

### Read the pattern before changing reagents

The [Merck PCR troubleshooting guide](https://www.merckmillipore.com/IL/en/technical-documents/technical-article/genomics/pcr/troubleshooting) recommends dilution series or a SPUD assay for detecting inhibitors and lists amplification plots, standard curves, melt plots, and raw-data views as complementary diagnostics. An inhibitor is a substance that blocks the enzyme. A SPUD assay is a test designed to reveal that blocking effect.

Match what you see to a likely cause before touching the recipe. The dilution ratios and cycle numbers below are common starting points, not fixed cutoffs.

| What you see | Likely cause | First check |
| --- | --- | --- |
| No product in sample or positive | Shared mix, missing reagent, or machine | Remake the master mix and confirm the program on the machine |
| Positive works and sample does not | Weak DNA or a substance blocking the enzyme | Dilute the sample 1 to 5 or 1 to 10 and rerun |
| Blank makes a band or curve | Contamination | Stop, discard suspect aliquots, clean, and repeat |
| Extra bands or a smear | Loose primer binding or too many rounds | Sharp extra bands point to binding temperature, a continuous smear points to excess cycles. Raise the binding temperature a few degrees first |
| Second melt peak on a dye assay | Primers binding each other | Review the melt curve and check a gel |
| Blank rises only very late | Trace contamination | Do not call samples positive; clean and rerun |
| Sample rises before cycle 10 | Too much target | Dilute and repeat |

Late signal in a blank points to trace contamination rather than a true positive, so any sample call near that signal stays suspect. A positive that lands inside its expected window and a blank that stays flat mean the controls have done their job.

Change one variable, record it and repeat the controls. Changing primer amount, temperature and template amount together may produce a band, but it won't reveal which change mattered, and the next failed run will be just as hard to read.

## PCR Types and What a Good Result Shows

Four PCR formats answer different questions. Confusing them leads to claims that the readout can't support.

| Method | Question it answers | Main readout | Evidence that makes it credible |
| --- | --- | --- | --- |
| End-point PCR | Is the DNA target present, and about what size is it | A gel band at the expected size | A working positive and a blank that stays blank |
| Quantitative PCR | How much DNA was in the tube | Amplification curve and cycle threshold | Dilution curve, replicates, efficiency near 100 percent, slope near -3.32 |
| RT-qPCR | How much RNA target was present after it is copied into DNA | Curve and cycle threshold after reverse transcription | A no-reverse-transcription control plus the qPCR checks |
| Digital PCR | How many target molecules are present | Count of positive tiny partitions | Enough partitions and a blank that stays negative |

### What each readout means

**End-point PCR** is read after the run ends. A band at the expected size supports presence and approximate size. It doesn't give a precise starting amount because the reaction slows as reagents are used.

**Quantitative PCR**, or qPCR, watches signal during the run. It can estimate starting amount when dilutions, replicates and efficiency are reported. Replicates are repeated tubes or wells prepared to check consistency.

The [MIQE guidance summarized in a quantitative PCR methods paper](https://link.springer.com/content/pdf/10.1186/1471-2199-11-74.pdf?error=cookies_not_supported&code=059f4d1e-db55-4e8a-b670-e6ebffd65ccf) recommends a serial dilution series, technical replicates, and reporting the **slope, intercept, linear dynamic range, limit of detection, and amplification efficiency**. The ideal doubling efficiency is **100%, corresponding to a standard-curve slope of approximately -3.32**.

In plain terms, 100 percent efficiency means the amount doubles each round. Slope is the steepness of the line drawn through dilution results. Intercept is where that line crosses the vertical axis. A slope far from minus 3.32 can point to primer problems, inhibitors, inaccurate dilution or uneven pipetting.

**RT-qPCR** starts with RNA, copies it into DNA and then measures that DNA. It can suggest gene activity only when RNA quality is documented and the no-reverse-transcription control stays quiet. A gene is a DNA instruction for making a product such as a protein.

**Digital PCR** counts partitions that contain the target. That count supports an estimate of molecule number without relying on a standard curve in the same way. It still needs blanks, suitable dilution and enough partitions.

A structured [qPCR template](https://www.verbalexperiment.com/blog/qpcr-template) helps keep those validation fields together. Even then, raw Cq values from different assays shouldn't be compared unless efficiencies and thresholds are standardized.

> A correct run can support identity, size, relative or absolute amount, and evidence about contamination or inhibition. One unrepeated tube can't support a broad biological claim by itself.

## Record PCR Observations Before Memory Fades

A band or Cq value alone isn't a result another scientist can evaluate. The record must show whether controls passed, whether template was diluted and whether the actual run matched the plan. Small timing and temperature changes can alter which products form, and those details are hard to rebuild after tubes and gels are discarded.

![A lab technician reviewing PCR test results on a computer monitor and writing in a lab notebook.](https://cdnimg.co/cab61d54-2a92-4be8-97b1-fd84150a185b/bc7eb6f4-71e6-489c-944f-4e339a4d0833/how-to-do-pcr-lab-technician.jpg)

### Fields to finish at the bench

The [Bio-Rad MIQE 2.0 bulletin](https://www.bio-rad.com/sites/default/files/2026-07/Bulletin_3556-Ver-B.pdf) specifies reporting the **reaction volume; amount of DNA or cDNA; primer or probe, magnesium, and dNTP concentrations; polymerase identity and concentration; buffer or kit identity and manufacturer; buffer composition; additives such as DMSO; and manufacturer product codes for tubes or plates**. It also calls for documenting template dilution and explaining how inhibition or matrix effects were evaluated.

**cDNA** is DNA copied from RNA. **dNTPs** are the building blocks used to make new DNA. **DMSO** is an additive sometimes used when DNA is difficult to copy.

| Field | What to record |
| --- | --- |
| Objective | The question, target, and expected product size or curve |
| Materials | Enzyme or kit name, lot, expiry date, primer and magnesium amounts, additives |
| Procedure | Actual machine program, any change from the plan, and the timing of each step |
| Observations | Delays, dropped tips, condensation, swapped wells, and how inhibition was checked |
| Conclusion | Control results, the decision taken, and whether the run is accepted or rejected |

Reagent lot and expiry date matter because a new lot can behave differently from an older one. A swapped well matters because the plate map may no longer match the file. Condensation matters because liquid on the lid isn't participating fully in the reaction. Notes written later often preserve the final number but lose those small events, and the small events are often the clue that everyone misses.

### Capture the run as it happens

For broader habits around contemporaneous notes, [AIDictation's guide to documenting research](https://aidictation.com/blog/research-documentation) is a useful companion resource. The principle stays the same across tools: source notes come first, interpretation comes after review.

Verbex, free to download for iPhone and iPad, is a private, on-device **lab notes app** for capturing bench reality through voice notes while both hands stay busy. Notes are timestamped and organized into Objective, Materials, Procedure, Observations and Conclusion for review, and a photo of a reagent label can fill in catalog, lot and expiry for the Materials entry.

A run recorded in the moment might note that the polymerase came from a new lot received that morning, or that condensation formed on the tube caps during the final hold. When the same assay fails two weeks later, that note tells the scientist which variable to check first rather than swapping reagents at random, and it supports the same control-pattern reasoning used to diagnose the failed run.

* * *

For the next PCR run, capture the program, control layout, delays and unexpected bench events as they happen, then review and export the completed record for the existing lab notebook. Visit [Verbal Experiment or Verbex](https://www.verbalexperiment.com) to use Verbex for iPhone and iPad, free to download.