Skip to main content

Table 2 Model performance

From: A machine learning model to determine the accuracy of variant calls in capture-based next generation sequencing

  Present Not present
High confidence variant 6622 / 6622 (100.0%)
True positive prediction
0 / 6622 (0%)
False positive prediction
Low confidence variant 44 / 557 (7.9%)
False negative prediction
513 / 557 (92.1%)
True negative prediction
  1. Results of a 10-fold cross-validation of the model on all 7179 variants. Variants called in the NGS pipeline were tested by Sanger sequencing. Those that were confirmed by Sanger sequencing are reported here as “Present”, and those that did not confirm are reported here as “Not Present”. All the variants were evaluated by the machine learning model and categorized as a “High confidence variant” or a “Low confidence variant”