GC Content Distribution Plot
Plot the GC percentage across a DNA sequence with a sliding window to find GC-rich and AT-rich regions.
Learn more
What it does
The GC content plot scans a DNA sequence with a sliding window and draws the GC percentage of each window against its position, revealing the local base composition rather than a single average for the whole molecule. A summary below the chart reports the overall GC content, the window mean, and the minimum and maximum window GC. Because GC content is very unevenly distributed in real genomes, an average on its own discards a great deal: two fragments that both average 45% GC may look nothing alike, one being uniform throughout and the other carrying a 70% GC-rich island next to a 25% AT-rich stretch. Finding GC-rich regions such as CpG islands in promoters, and AT-rich regions such as replication origins, is the most common use.
How it works
For a window of size w starting at position i, the plotted value is GC percent = (count of G and C inside the window) / w x 100. The window advances one base at a time, so a sequence of length L yields L - w + 1 windows, each plotted at its midpoint. A 1000 bp sequence with a 100 bp window therefore produces 901 data points. Window size trades resolution against noise. A small window is sensitive to short features but fluctuates sharply — at w = 20 a single base entering or leaving the window shifts the value by five percentage points, and the curve looks like noise. A large window is smooth but averages away features only a couple of hundred bases long. The default 100 bp is a reasonable compromise at gene scale: it lets features of a few hundred bases show up without jitter from individual bases. When judging whether a region stands out, compare it with the overall GC content of the sequence rather than against a fixed 50% — in a sequence that averages 35% GC, a local 50% is already a pronounced GC-rich region. Established definitions help with interpretation: a CpG island is usually defined as a region longer than 200 bp with GC above 50% and an observed-to-expected CpG ratio above 0.6, while windows above 70% GC often correspond to strong secondary structure that stalls polymerases.
Worked example
Take a 1000 bp sequence spanning a promoter and the start of a gene. With the default 100 bp window it yields 1000 - 100 + 1 = 901 data points. Suppose the overall GC content is 48%: the curve may still climb to 72% in the roughly 200 bp upstream of the transcription start site — meaning 72 of the 100 bases in that window are G or C — while dropping to 32% in an AT-rich stretch downstream. That 40 percentage point gap between the window minimum and maximum is exactly the information a single 48% average hides. A second, diagnostic case: a sequence whose overall GC is 60% looks unremarkable, but its maximum window reaches 85%. Such a window usually marks a region that is hard to amplify, so if a PCR fails precisely there, the plot has already supplied the reason. Conversely, if the whole curve oscillates only between 45% and 55%, the composition is uniform and GC-driven amplification or sequencing problems are unlikely.
When to use it
Before designing bisulphite sequencing primers or a methylation assay, plot the region around the promoter to locate CpG islands and place primers to either side of a GC-rich block rather than inside it — a primer sitting in high GC amplifies poorly. When a PCR consistently fails at one point in a template, a GC plot often reveals a window above 70% where the polymerase stalls, pointing to DMSO, betaine or a higher denaturation temperature as the fix. When checking a newly assembled contig, a segment whose GC content deviates sharply from the rest may indicate a horizontally transferred island, a mobile element, or contamination from another organism. In whole-genome coverage analysis, overlaying low-coverage regions on the GC curve separates GC bias in the sequencing platform from genuine copy-number change. And when choosing sgRNAs or hybridisation probes, avoiding extreme GC and targeting the 40% to 60% range is standard practice for improving efficiency.
FAQ
- How is a GC content plot different from the overall GC content?
- The overall GC content gives one average for the whole sequence; the plot gives the GC percentage at every position. The two can diverge sharply — two fragments averaging 45% GC might have one uniform and the other carrying a 70% GC-rich island beside a 25% AT-rich stretch. When you need to locate a specific region, such as a promoter, a CpG island or a segment that fails to amplify, only the plot provides position information.
- What window size should I use?
- It depends on the scale of the feature you are looking for. Locating CpG islands, which run from roughly 200 to 1000 bp, works well with a 100 to 200 bp window; scanning a whole genome for compositional skew suits 1000 bp or more; inspecting features only tens of bases long needs 20 to 30 bp, at the cost of a much noisier curve — at a 20 bp window a single base entering or leaving shifts the value by five percentage points. The 100 bp default is a good compromise for gene-scale analysis.
- How do I identify a CpG island from the plot?
- Look for a stretch that satisfies three conditions at once: longer than 200 bp, GC content above 50%, and an observed-to-expected CpG ratio above 0.6. This tool reports GC content, so use it to shortlist candidate regions, meaning stretches where the curve stays above 50%, then check the CpG ratio for those candidates separately. Regions of this kind sitting near a gene promoter are the most interesting, because they are commonly tied to transcriptional regulation and DNA methylation.
- What does a GC content above 70% mean?
- It usually indicates strong secondary structure or unusually stable duplex in that stretch, where the polymerase stalls. In practice that shows up as premature termination, low product yield, or non-specific bands, and sequencing coverage often drops in the same place. The usual remedies are adding DMSO at around 5% or betaine, raising the denaturation temperature, switching to a polymerase that tolerates high GC, or shortening the amplicon so the region can be avoided altogether.
- How does the GC plot relate to sequencing depth?
- The two are often read together. Sequencing platforms show a marked GC bias: regions of very high or very low GC tend to get lower depth and may even show coverage gaps. Overlaying a coverage track on the GC plot separates two very different explanations. If the low-coverage regions coincide with windows of extreme GC, the cause is technical bias; if coverage is low where GC is normal, a genuine deletion, duplication or copy-number change is more likely. For whole-genome and exome work, plotting GC content first is a necessary step before judging coverage anomalies.
- What input does the plot accept?
- It accepts single-letter DNA, in either case, and ignores whitespace and line breaks. Characters outside A, C, G and T — such as N or IUPAC degenerate codes — do not count towards the GC statistics but still occupy window positions, so an assembly with long N runs will show artificially low and jittery GC across those windows. If the goal is the compositional skew of a finished chromosome, trim regions with a high N fraction before plotting.