Summary of the impact of DNA source and eukaryotic DNA enrichment on the accuracy of genetic variant detection from whole-genome sequencing data.
| Variant type | Sensitivity | False positive rate |
| Blood versus non-enriched saliva or buccal | ||
| SNVs | Little or no difference | Blood |
| Indels | Little or no difference | Little or no difference |
| CNVs (deletions) | Blood | Little or no difference |
| CNVs (duplications) | Blood | Blood |
| Enriched versus non-enriched saliva or buccal | ||
| SNVs | Little or no difference | Enriched |
| Indels | Little or no difference | Little or no difference |
| CNVs (deletions) | Non-enriched | Little or no difference |
| CNVs (duplications) | Non-enriched | Non-enriched |
For each comparison, the better sample type (ie, the one having higher sensitivity or a lower false positive rate) is indicated. Blood and enriched saliva and buccal samples tended to have lower false positive rates for SNVs than non-enriched saliva and buccal samples, but the magnitude of the differences were small except when considering rare SNVs (see text) and exhibited variability across individuals.
CNV, copy number variant; SNV, single nucleotide variant.