Figure 3. Performance comparison of Balaur and other read mapping tools.

(a) Accuracy results on simulated 150 bp long reads with 1 and 2% sequencing error rates. (b) Runtime comparison and breakdown by alignment phase on 150 and 350 bp simulated reads. The reported ‘encryption' time includes both secure kmer hashing and repeat masking, while the ‘reporting' time represents the post-processing of the voting results on the client (that is, selecting the best alignment positions and SAM output file I/O). (c) Evaluation on real read data. (d) Evaluation on simulated long read data with varying sequencing error rates showing the runtime and the percentage of Q10 reads that were mapped correctly (within 20 bp of true alignment position).