Optimization of Filter Metrics for Mitochondrial DNA Sequence Analysis.

Quality metrics translate sequence information into numerical values which allows a software program to filter through data without human intervention. Primer specific settings for the trace score and contiguous read length in Sequence Scanner Software v1.0 (Applied Biosystems, Foster City, CA) were established using a calibration dataset of 2,817 sequence traces and validated using a second dataset of 5,617 sequence traces. Prior to optimization 51.7% of the samples required manual intervention while 28.4% require review after optimization. An evaluation of signal intensity and signal to noise ratio variables was performed and no trend was recognized for predictive modeling. Use of quality values per peak to ascertain confidence in the base call was evaluated and found to be a feasible parameter for sample quality assessment and confident base calling.