To perform whole exome sequencing, a laboratory first extracts DNA from a blood or saliva sample and breaks it into millions of small fragments. Because the coding exome makes up only a tiny slice of the total genetic code, the laboratory then uses a step called exome capture, in which chemical probes are used to fish out specifically the fragments that come from coding regions, leaving the much larger non-coding portion of the genetic code behind. Only these captured fragments are then read by a sequencing machine, which determines their exact sequence, and a computer reassembles and compares this sequence to a standard reference to flag places where a person’s genetic code differs from what is typically expected. Those differences, called variants, are then filtered and reviewed by geneticists to decide which, if any, explain the person’s symptoms.

This capture step is what sets whole exome sequencing apart from its two closest relatives. A gene panel uses a similar capture approach, but targets only a predefined, much smaller list of genes already known to cause a particular condition or group of related conditions; because the laboratory is reading far less material, panels are generally faster, less expensive, and produce fewer findings unrelated to the original question, though they cannot help if the responsible gene turns out to lie outside the chosen list. Whole genome sequencing, by contrast, skips the capture step entirely and reads almost the complete genetic code, coding and non-coding alike; this allows it to detect certain structural changes and non-coding variants that whole exome sequencing is not designed to see, generally with simpler sample preparation, but it produces a much larger volume of data to interpret and remains costlier in most settings.

In practice, whole exome sequencing sits between these two options: broader than a gene panel, so it does not depend on correctly guessing the responsible gene in advance, but more focused than whole genome sequencing, keeping the volume of findings, including findings unrelated to the original question, more manageable. A geneticist chooses among these three based on how confident the clinical picture is and how broadly the search needs to be cast.